Research network on capital markets and financial integration in

RESEARCH NETWORK ON CAP ITAL MARKETSAND F INANC IAL INTEGRAT ION IN EUROPE

DECEMBER 2004

CENTER FOR

FINANCIALSTUDIES

RESULTS ANDEXPERIENCE AFTER TWO YEARS

RESE

ARCH

NET

WO

RK O

N C

APIT

AL M

ARKE

TS A

ND

FIN

ANCI

AL I

NTE

GRA

TIO

ND

ECEM

BER

2004

EURO

PEAN

CEN

TRAL

BAN

K

Titel_CFS 21.12.2004 11:03 Uhr Seite 1

In 2004 all ECB publications will feature

a motif taken from the

€100 banknote.

RESEARCH NETWORK ON CAPITAL MARKETS AND FINANCIAL INTEGRATION IN EUROPE

RESULTS AND EXPERIENCE AFTER TWO YEARS

DECEMBER 2004

CENTER FOR

FINANCIALSTUDIES

© European Central Bank, 2004

AddressKaiserstrasse 2960311 Frankfurt am Main, Germany

Postal addressPostfach 16 03 1960066 Frankfurt am Main, Germany

Telephone+49 69 1344 0

Websitehttp://www.ecb.int

Fax+49 69 1344 6000

Telex411 144 ecb d

All rights reserved.Reproduction for educational andnon-commercial purposes is permittedprovided that the source is acknowledged.

As at December 2004.

ISBN 92-9181-527-6 (print)ISBN 92-9181-528-4 (online)

THE JOINT ECB-CFS Research Network on “Capital Market andFinancial Integration in Europe” aims at promoting high qualityresearch. The Network as such does not express any views, nortakes any positions. Therefore any opinions expressed indocuments made available through the Network (including thisreport and the website) or during its workshops and conferencesare the respective authors’ own and do not necessarily reflectviews of the ECB, the Eurosystem or CFS.

3ECB

Capital markets and financial integration in EuropeDecember 2004

CONTENTS

EX E CU T I V E S UMMARY 4

1 I N T R O D U C T I O N 7

2 PURPO S E O F T H E N E TWORK ANDMA I N P R I O R I T Y A R E A S 8

3 A C T I V I T I E S O F T H E N E TWORK 10

3.1 Conferences 10

3.2 Fellowships 11

3.3 Publications 13

3.4 Website 14

3.5 Related initiatives 14

4 R E SU LT S O F T H E N E TWORK 16

4.1 Results in priority areas 16

4.1.1 Bank competition and thegeographical scope ofbanking 16

4.1.2 European bond markets 17

4.1.3 European securitiessettlement systems 20

4.1.4 The emergence and evolutionof new markets in Europe(in particular start-upfinancing markets) 22

4.1.5 International portfoliochoices and asset marketlinkages between Europe,the United States and Japan 23

4.2 Results in other areas 24

5 FU TUR E S T E P S : C ON T I NUAT I ON O F T H ENE TWORK AND O F T H E L AM FA LU S S YR E S E A R CH F E L LOWSH I P P ROGRAM 27

5.1 The relationship between financialintegration and financial stability 27

5.2 EU accession, financialdevelopment and financialintegration 28

5.3 Financial system modernisationand economic growth in Europe 28

C O N T E N T S6 ORGAN I S AT I ON O F T H E N E TWORK 30

6.1 A “network of people” 30

6.2 Structure 30

A N N E X E S 32

A The roadmap 32

B Program and summary of thelaunching workshop 52

C Affiliations of participants tonetwork events 73

D Program and summary of thesecond workshop 74

E Program and summary of thethird workshop 95

F Program and summary of thesymposium 118

G Papers of the Lamfalussyfellows 147

H Outline of the special issue ofthe “Oxford Review of EconomicPolicy” on “European FinancialIntegration” 152

I List of contributors to theECB-CFS researchnetwork (2002-2004) 153

J Composition of the SteeringCommittee (2002-2004) 157

K Composition of the ECB internaland NCB contact groups 158

4ECBCapital markets and financial integration in EuropeDecember 2004

In April 2002 the European Central Bank(ECB) and the Center for Financial Studies(CFS) launched the ECB-CFS ResearchNetwork to promote research on “CapitalMarkets and Financial Integration in Europe”.The ECB-CFS research network aims atstimulating top-level and policy-relevantresearch, significantly contributing to theunderstanding of the current and futurestructure and integration of the financialsystem in Europe and its international linkageswith the United States and Japan. This reportsummarises the work done under the networkafter two years.

Over time the network formed a coherent andgrowing group of researchers interested in theintegration of European financial markets,while using light organisational structures andbudgets. The members of this evolving groupmet repeatedly at the events organised by thenetwork to present the latest results of theirresearch and to share views on policy options.In this sense, the “network of people” intendedat the start was created. Overall, the networkaroused great interest, as leading academicresearchers, researchers from the main policyinstitutions and high-level policy makersparticipated actively in it by presentingresearch results, through speeches and inpolicy panels. It also stimulated a new researchfield on securities settlement systems, an areaof high policy relevance and interest to theECB that had not attracted much interest in theresearch community beforehand. Also, thenetwork seems to have triggered several relatedoutside initiatives by international institutions,such as the IMF or the OECD.

During its first two years the network wasorganised around three workshops and a finalsymposium on 10-11 May 2004. To focusresearch resources and to ensure medium-termpolicy relevance, a limited number of areashave been given top priority: bank competitionand the geographical scope of banking;international portfolio choices and assetmarket linkages between Europe, the UnitedStates and Japan; European bond markets;

European securities settlement systems; andthe emergence and evolution of new markets inEurope (in particular start-up financingmarkets).

In order to stimulate further research focusedon the priority fields of the network, the ECBLamfalussy research fellowships wereestablished. These fellowships sponsorprojects proposed by young researchers, bothadvanced doctoral students and youngerprofessors. Five Lamfalussy fellowships weregranted in 2003 and five more in 2004. The firstpapers from this program have already beenissued in the ECB working paper series or areforthcoming. One of them won the prize for thebest paper written by a Ph.D. student at the2004 European Finance Association Meetingsin Maastricht.

Results of the network in the five top priorityareas can be summarised as follows:

Bank competition and the geographical scopeof banking. First, integration does not appear tobe very advanced in many retail bankingmarkets. Second, some of the inherentcharacteristics of traditional loan and depositbusiness constrain the cross-border expansionof commercial banking, even in a commoncurrency area. Hence, the implementation ofsome policies to foster cross-border integrationin retail banking may be ineffective. Third,theoretical research suggests that supervisorystructures may not be neutral towards furtherEuropean banking integration. Finally, astronger role of area-wide competition policiescould be beneficial for further bankingintegration. This would also stimulateeconomic growth, as more competition in thebanking sector induces financially dependentfirms to grow more.

European bond markets. While the governmentbond market has integrated rapidly with theEMU convergence process, its full integrationhas not yet been achieved. The introduction of acommon electronic trading platform reducedtransaction costs substantially, but yield

EX E CU T I V E S UMMARY

5ECB


EXECUTIVESUMMARY

spreads of long-term sovereign bonds of theeuro area are still heterogeneous. This islargely explained by different sensitivities toan international risk factor, whereas liquiditydifferentials only play a role in conjunctionwith this latter factor. Somewhat surprisinglyin this context, the dynamically developingcorporate bond market exhibits a relativelyhigh level of integration. There is alsoincreasing evidence that the introduction of theeuro has contributed to a reduction in the costof capital in the euro area, in particular throughthe reduction of corporate bond underwritingfees. As a result, firms may wish to increasebond financing relative to equity financing.The development of a larger corporate bondmarket is also important for monetary policy.For example, US evidence suggests that therating of corporate bonds may contribute to thepersistence of recessions, as rating agencies’policies affect firms asymmetrically in theiraccess to the bond market over the businesscycle. US evidence also suggests that liquidityconditions in stock and bond markets tend to bepositively correlated.

European securities settlement systems.European securities settlement infrastructuresare highly fragmented and further integrationand/or consolidation would exploit economiesof scale that could greatly benefit investors. Itis not clear, however, whether direct publicintervention in favour of consolidation wouldlead to the highest level of efficiency, forexample because of the existence of strongvertical integration between trading andsecurities platforms (“silos”). In contrast,promoting open access to clearing andsettlement systems could lead to consolidationand the highest level of efficiency. Finally,regarding concerns about unfair practices byCentral Securities Depositories (CSDs) towardcustodian banks, regulatory interventionsfavouring custodian banks should bediscouraged, as long as CSDs are not allowedto price discriminate between custodian banksand investor banks.

The emergence and evolution of new markets inEurope (in particular start-up financingmarkets). While fairly well integrated, “newmarkets” and start-up financing are lessdeveloped and integrated in Europe than in theUnited States. However, new markets andventure capitalists are the most importantintermediaries for the financing of projectswith high risk but with potentially very highreturn. The analysis carried out within thenetwork reveals that European start-upfinanciers are mostly institutional investors,while US venture capitalists are mostly richindividuals. Also, new markets are essential forthe development of start-up finance in Europe,as they provide an exit strategy for start-upfinanciers who can then sell new successfulprojects using initial public offerings. Finally,the legal framework affects the development ofventure capital firms. For example, very strictpersonal bankruptcy laws constrain early stageentrepreneurs, reducing demand for venturecapital finance.

International portfolio choices and assetmarket linkages between Europe, the UnitedStates and Japan. At a global scale, assetmarket linkages have increased recently. Forexample, major economies such as the UnitedStates and the euro area have become morefinancially interdependent. This phenomenoncan be observed in stock and bond markets aswell as in money markets, where the maindirection of spillovers has recently been fromthe US to the euro area. Country-specificshocks now play a smaller role in explainingstock return variations of firms whose sales areinternationally diversified. Increases in firm-by-firm market linkages are a globalphenomenon, but they are stronger within theeuro area than in the rest of the world. Variousother phenomena also increase market linkagesand therefore the likelihood that financialshocks spread across countries. One example isthe use of global bonds. Finally, the nowadaysmore direct access of unsophisticated investorsto financial markets may increase volatility.


Other areas. Financial integration affectsfinancial structures, but it does not need to leadto their convergence across countries.Financial structures matter for growth, asmarket-oriented financial systems benefit allsectors and firms, whereas bank-based systemsprimarily benefit younger firms that depend onexternal finance. Moreover, good corporategovernance increases firms’ value. Inparticular, the dual board system, where themonitoring and advising roles of the board ofdirectors are separated, is found to dominatethe single board structure. Therefore, thefurther development of the European singlemarket should strongly require good corporategovernance. In general, well designedinstitutions foster entrepreneurial activity,partly by relaxing capital constraints.

The results of the network clearly illustratedthe substantial effects the introduction of theeuro had on euro area financial markets. Inaddition to the effects on bond markets, stockmarkets and the cost of capital summarisedabove, research produced showed that thesingle currency had its strongest effects onmoney markets, whose unsecured segment isnow completely integrated. Without any doubtthe euro generally enhanced the liquidity andefficiency of euro area financial markets, andongoing initiatives such as the EuropeanUnion’s Financial Services Action Plan willhelp to continue this process.

In sum, in the first two years the network hasestablished itself as the hub for the researchdebate on European financial integration.Some of the best papers produced by thenetwork, leading to the conclusions mentionedabove, are currently being considered forpublication in two special issues of academicjournals. An issue of the Oxford Review ofEconomic Policy on “European financialintegration” is published contemporaneouslywith this report, and an issue of the Review ofFinance is planned for next year. The currentpolicy context, the gradual progress ofintegration as well as the creation of otherrelated non-ECB or non-CFS initiatives on

financial integration suggest that this topic willremain high on the agendas of policy makersand academics for the years to come.

Therefore, the ECB Executive Board and theCFS decided to continue the network,refocusing its priorities. Three priority areashave been added: 1) The relationship betweenfinancial integration and financial stability,2) EU accession, financial development andfinancial integration, and 3) financial systemmodernisation and economic growth in Europe.These three areas have become particularlyimportant at the current juncture, but have notreceived particularly strong attention in thefirst two years of the network. For example, thearea of financial stability research washighlighted by the ECB research evaluators asan area deserving further development.Moreover, despite the results found in the firsttwo years of the network, new developmentsremain to be further explored in the earlierpriority areas.

A three-year extension is envisaged, runningfrom after the May 2004 symposium until 2007,with two events to be held per year. The three-year period is long enough to consider the firsteffects of the Financial Services Action Plan. Italso constitutes a realistic horizon for theambitious agenda implied by the three newpriorities. The generally light organisationalstructure and working of the network will notbe changed. In addition, given the value of theLamfalussy fellowship research program ininducing further research in the areas of thenetwork, the program has also been extendedfor all the research topics in the area of thenetwork.

7ECB


INTRODUCTION

Financial markets integration can contribute tohigher European economic growth. A study ofthe European Commission by Giannetti, Guiso,Padula and Pagano (2002)1 shows that financialintegration, defined as a convergence forEuropean financial development toward the USstandard, would imply a one percentage pointgain in EU GDP growth. This important gain isonly a lower bound, as the report did notconsider integration of service sectors. Toquickly reap all the benefits from furtherEuropean financial integration, theCommission initiated the Financial ServicesAction Plan in 1999 to remove the regulatoryand market barriers that exist to the cross-border provision of financial services and toencourage the free flow of finance withinEurope.

In this context, the European Central Bank(ECB) and the Center for Financial Studies(CFS) launched, in April 2002, the ECB-CFSresearch network to promote research on“Capital Markets and Financial Integration inEurope”. The network was launched in aneffort to understand better the state ofEuropean integration in each financial marketand where improvements are possible. Thenetwork was first established for a period oftwo years, which ended in May 2004. Thisreport summarises the work done under thenetwork.

Section 2 recalls the purpose of the network andits main priority areas. The activities of thenetwork are summarised in Section 3. Section 4presents the main results, Section 5 makes thecase for the continuation of the network andpresents additional areas where further work isadvisable. Finally, Section 6 explains theorganisational structure.

1 I N T RO D U C T I O N

1 Giannetti, M., L. Guiso, T. Japelli, M. Padua and M. Pagano(2002), “Financial integration, corporate financing andeconomic growth”, CEPR, f inal report to the EuropeanCommission, 22 November.


In 1999, the Commission presented aframework for action in the financial servicesindustry to help achieve the benefits of theSingle Market in financial services. Theformulated objectives of this FinancialServices Action Plan (FSAP) are to ensure asingle EU market for wholesale financialservices and to guarantee open, secure retailmarkets and modern, prudential rules andsupervision. The Commission proposed 42measures and a timetable for their adoptionwith a deadline of 2004. In 2002 theCommission reported that “recent progress inthe Council and the European Parliament on anumber of proposals demonstrate that thepolitical commitment to implement the FSAPon time is beginning to be translated intofirm political agreements … Even if not allbarriers have been removed, significantand irreversible progress towards a strongintegrated European financial sector by 2005 isachievable – it is a prize that is now within ourgrasp”. In this context the ECB-CFS researchnetwork on “Capital Markets and FinancialIntegration in Europe” was created.

The ECB-CFS research network aims atstimulating top-level and policy-relevantresearch, significantly contributing to theunderstanding of the current and futurestructure and integration of the financialsystem in Europe and its international linkageswith the United States and Japan. The work ofthe network focuses on three distinct, butnot unrelated, main broad research areas:(i) European financial integration, (ii) financialsystem structures in Europe and (iii) financiallinkages between the euro area/EuropeanUnion (EU), the United States and Japan. Whenthe network was created, these areas wereunder-researched and knowledge about theminsufficient.

A detailed description of key research areaswas developed under the supervision of theSteering Committee (SC) and made publiclyavailable in the network “roadmap”. While the“roadmap” is geared towards applied andpolicy-relevant questions, both empirical and

2 PURPO S E O F T H E N E TWORK AND MA I NPR I O R I T Y A R E A S

theoretical research are important and wereencouraged. Annex A contains this “roadmap”.

To concentrate research resources and ensurepolicy focus, a limited number of areas withinthe three main broad research fields have beengiven top priority: Bank competition and thegeographical scope of banking activities;international portfolio choices and assetmarket linkages between Europe, the UnitedStates and Japan; European bond markets;European securities settlement systems; andthe emergence and evolution of new markets inEurope (in particular start-up financingmarkets).

– Work on bank competition and thegeographical scope of banking activitieshas been given priority, because – despitethe adoption of the “single passport”principle – the euro area is stillexperiencing relatively few cross-bordermergers compared with domesticconsolidation and relatively limited cross-border corporate lending. Also, supervisorystructures and regulatory approachespertaining to the banking sector underwentprofound reforms whose effects should beanalysed.

– Priority was also given to internationalportfolio choices and asset market linkagesbetween Europe, the United States andJapan, as the past few decades have broughtan enormous expansion of internationalcapital flows. As a consequence, globalfinancial linkages have strengthened.While their impact on economies isnowadays far larger than traditional tradelinkages, for example, knowledge aboutthe driving factors behind internationalfinancial flows is still relatively limited.

– European bond markets have undergonerapid changes in the past few years,including the development of euro area-wide electronic secondary market tradingplatforms, as well as the development of amore significant corporate bond market.

9ECB


2 PURPOSE OF THENETWORK ANDMAIN PRIORITY

AREASThese changes are so recent that their mainsources and their wider implications are notwidely understood. For example, bondmarkets constitute one key market for theconduct of monetary policy by centralbanks.

– Work on European securities settlementsystems has also received priority as thefragmentation of this industry in Europe,resulting in high cross-border securitiestrading costs, may well constitute thesingle most important obstacle to furthersecurities market integration.2 The rapidstructural change in the European securitiessettlement industry, and the very limitedresearch available on these topics when thenetwork was created, made work in this areaparticularly important and relevant topolicy-makers.

– The area of new markets also received avery high priority. In 1998, the EuropeanCommission reckoned that “fewertechnology-based enterprises are createdin Europe and their prospects for growthare inhibited. Also, venture capital isunderdeveloped in many Europeancountries compared with the United States,in particular in the field of seed andearly stage finance”.3 This highlightedthe importance of the availability of awide range of funding and investmentpossibilities for innovations and risksharing – and hence ultimately for growthand welfare.

2 See Giovannini Group (2001), Cross border clearing andsettlement arrangements in the European Union, Brussels,November; and Giovannini Group (2003), Second report onEU clearing and settlement arrangements, Brussels, April.

3 European Commission (1998), Fostering entrepreneurship inEurope: Priorities for the future. COM (98) 222 f inal.


3.1 CONFERENCES

In its first two years the network was organisedaround three workshops and a finalsymposium. Workshops systematically workeddown the list of main priority areas, coveringeach priority area on at least two occasions.The symposium covered all areas of thenetwork. Each workshop and the symposiumfeatured parallel sessions where research in therelevant area of the network was presented andended with a policy panel. This section brieflysummarises the structure of each workshop andof the symposium.

The network was launched with a workshop atthe ECB on 29-30 April 2002. It featured“agenda setting” talks by prominentresearchers and members of the networkSteering Committee, pure research paperpresentations and the key policy speeches“Monetary policy in an environment of globalfinancial markets” by Otmar Issing (ECB),“Competition, co-operation and public action:Three necessary drivers for European financialintegration” by Tommaso Padoa-Schioppa(ECB) and “Consolidation in the Europeansecurities infrastructure – What is needed?” bySirkka Hämäläinen (ECB).

The sessions of the launching workshopconcentrated on the priority areas of Europeandebt market structures and internationalfinancial linkages. In the closing plenarysession, Jesper Berg (ECB), AlbertoGiovannini (Unifortune Asset Management)and David Wright (European Commission)gave their views about the status of theEuropean financial system and the way forwardto complete integration. Vítor Gaspar (ECB)acted as moderator. Annex B contains theprogram and a detailed summary of theworkshop. Aside from the research progressreported in the parallel sessions, the mainresult of the launching workshop was thedevelopment of the “roadmap” to help guidework under the network (see Annex A). Thisroadmap has been developed by ECB staff

3 A C T I V I T I E S O F T H E N E TWORKunder the auspices of the SC on the basis of the“agenda setting” talks and the policy speechesat the workshop, in collaboration with the ECBInternal Contact Group, the NCB ContactGroup (see section 6 below for those groups)and other workshop participants.

The response to the workshop announcementwas very encouraging. Representatives of allinvited institutions expressed interest inparticipating (including Eurosystems NCBs,the BIS, IMF, World Bank, Board of Governorsof the Federal Reserve System, and theEuropean Commission). Also, the response tothe first Call for Papers was satisfactory with68 submissions received. This positiveresponse led to a decision to combine the twoworkshops originally planned for the first year(2002) into the larger launching workshop.There were 114 participants, of which 38 werefrom academia, 35 from the ECB and 41 fromother official institutions. Affiliations ofnetwork event participants are detailed inAnnex C.

In deciding to establish the network, the ECBwished to hold subsequent workshops in otherEuropean cities to present the network as atruly European initiative. As a consequence,two more workshops in 2003 were hosted byeuro area national central banks (NCBs).

The Bank of Finland hosted the network’ssecond workshop in Helsinki on 11-12 March2003. The main priority areas analysed in thecourse of the workshop were (1) bankcompetition and the geographical scope ofbanking activities; (2) international portfoliochoices and asset market linkages betweenEurope, the US, and Japan and, as a specialtopic requested by several launching workshopparticipants, European equity markets. Theworkshop combined research key lectures,research paper presentations and a plenarypanel discussion on “The future of exchanges,consolidation or competition”. Panel speakerswere Niall Bohan (European Commission),Peter Gomber (Deutsche Boerse AG), Hannu

11ECB


3 A C T I V I T I E SO F T H E

N E T WO R KHalttunen (Nordea) and André Went(Euronext). Annex D contains the program anda summary of this workshop.

The Call for Papers elicited 59 submissions,from which 15 papers were selected. Theworkshop was intended to gather a rathersmaller number of participants relative to thelaunching workshop to allow for discussionson each selected paper. There were 87participants, of which 39 were from academia,15 from the Bank of Finland (the hostinginstitution), 15 from the ECB and 18 from otherofficial institutions.

The Bank of Greece hosted the network’s thirdworkshop in Athens on 20-21 November 2003.The main priority areas analysed in the courseof the workshop were (1) European bondmarkets; (2) European securities settlementsystems; and (3) start-up financing and newmarkets. This third workshop also combinedresearch key lectures, research paperpresentations and a plenary panel discussion on“European Securities Settlement Systems”,which included Kenneth D. Garbade (FederalReserve Bank of New York), Randy Kroszner(University of Chicago), Anso Thiré(Euroclear France), and Gertrude Tumpel-Gugerell (ECB). Annex E contains the programand a summary of the network’s thirdworkshop.

The Call for Papers elicited 77 papers, againconfirming the research community’s interestin the network initiative. 18 papers wereselected for presentation in Athens. There were77 participants, of which 37 came fromacademia, 3 from the Bank of Greece (thehosting institution), 16 from the ECB and 21from other official institutions.

The network’s first two years were concludedwith a symposium in Frankfurt on 10-11 May2004 (see Annex F for the program and asummary of the symposium). The Call forPapers brought in 118 submissions. Thesymposium was an opportunity to summarisethe progress that the network has accomplished

in terms of research and to outline itscontinuing efforts and new priority areas. Apolicy panel focusing on “Drivers of Europeanfinancial integration – Markets or policy?”concluded the symposium. It was chairedby Tommaso Padoa-Schioppa (ECB) andincluded high-level representatives of thepublic and the private sectors: Mario Draghi(Goldman Sachs), Alexander Schaub(European Commission) and Jens Thomsen(Danmarks Nationalbank). Furthermore,Robert Flood (International Monetary Fund)delivered a key lecture on new approaches toassessing financial integration; Otmar Issing(ECB) gave a speech on “Asset prices andmonetary policy”; Alexandre Lamfalussy(Institut d’études européennes, UniversitéCatholique de Louvain) expressed his views onEuropean bond markets; and Gertrude Tumpel-Gugerell (ECB) closed the symposium bylaying down the priorities for work within thenetwork for the coming three years. There were149 participants, of which 55 came fromacademia, 51 from the ECB (the hostinginstitution) and 43 from other officialinstitutions.

3.2 FELLOWSHIPS

The SC believed that the main priority areasselected for both workshops in 2003 wereunder-researched areas that were highlyrelevant to the work of the ECB and theEurosystem. In order to stimulate networkparticipants further to undertake research in theabove-mentioned fields, the ECB members ofthe SC proposed to fund some focused researchactivity in the context of the network in theform of fellowships. These fellowships areaimed at young researchers, mainly veryadvanced doctoral students and youngprofessors. This target group was felt to beparticularly responsive to suggestions forresearch areas of high policy importance. Thefellowships were named after AlexandreLamfalussy, the first President of the EuropeanMonetary Institute.


Following a Call for Projects highlighting thetop priorities of the network, five fellowshipswere granted in 2003 and five more weregranted in 2004. Each fellowship is endowedwith €10,000. The 2003 Lamfalussy fellows’papers are available in Annex G.

In 2003, the five Lamfalussy fellowships weregranted to:

– Rui Albuquerque (Assistant Professor,University of Rochester), “Asymmetricinformation and the persistence ofinternational equity flows”, ECB WorkingPaper No. 310. This project falls under mainpriority area “International portfoliochoices and asset market linkages betweenEurope, the US and Japan”.

– Giulia Iori (Lecturer, King’s College,University of London), “An investigationof the efficiency and stability of alternativedesigns for securities clearing andsettlement infrastructures”, ECB WorkingPaper No. 404. This project is acontribution to main priority area“European securities settlement systems”.

– Leo Kaas (Assistant Professor, Universityof Vienna), “Bank competition andfinancial market integration”, ECBWorking Paper No. 403. This contributes tomain priority area “Bank competition andthe geographical scope of bankingactivities”.

– Albert J. Menkveld (Assistant Professor,Vrije Universiteit Amsterdam),“Substitutability, fragmentation and pricediscovery in the European government bondmarket: An empirical study based onEuroMTS data”, ECB Working Paper No.385. This project falls under the mainpriority area “European bond markets”.

– Yigal S. Newman (Ph. D. student, StanfordGraduate School of Business), “The volumeof new issuance and its impact on market-wide credit spreads” forthcoming in the

ECB Working Paper Series. This projectcontributes to the main priority area“European bond markets”. Newmanreceived the 2004 European FinanceAssociation award for best Ph. D. paper forhis contribution.

In 2004, the five Lamfalussy fellowships weregranted to:

– Tomas Dvorak (Assistant Professor, UnionCollege, Schenectady, New York),“European Union enlargement and equitymarkets in candidate countries”. Followingthe European Commission report outliningthe timing and countries involved inenlargement, stock prices in the ten easternEuropean candidate countries went up onaverage by over 46%. The project willinvestigate whether the rise in stock pricesin the accession countries was related to EUenlargement and in particular, whether therise in stock prices was a result of repricingof systematic risk due to integration of localstock markets into the world market.

– Mariassunta Giannetti (AssociateProfessor, Stockholm School ofEconomics), “International banks: Effectson firm growth and financial stability”.Using micro-level data for listed andunlisted companies and information onforeign bank entry, the project will analysewhether local firms in emerging marketsreally benefit from the entry of foreignbanks in terms of growth and investmentpolicies. Moreover, it will study whetheryoung firms with little collateral benefit asmuch as listed companies.

– Miklós Koren (Ph. D. student, HarvardUniversity and Hungarian Academy ofSciences), “Financial integration andincome volatility”. The purpose of thisproject is to investigate empirically therelationship between financial integrationand income volatility and to provide atheoretical model explaining the findings.The envisaged key mechanism is that

13ECB



N E T WO R Kfinancial integration helps countries adoptless risky technologies, which results inlong-term income stability.

– Michael Kollo (Ph. D. student, LondonSchool of Economics), “The pricing ofunderwriting services by European, U.S.and Japanese underwriters in the Eurobondmarket”. Underwriter fees are an importantpart of the cost of capital for firms andtherefore may act as a barrier to entry intothe Eurobond market. The project will testthe relative importance of differentdeterminants of underwriter fees across thedifferent European and other nationalitiesof underwriters in the Eurobond market andwill also provide a currency breakdown.

– Philip Lane (Associate Professor, TrinityCollege), “Global bond holdings and theEurozone”. The project is based on theobservation that the pattern of foreignownership is not totally globalised, in thesense that the nationality of investors stillmatters. This heterogeneity in the investorbase has consequences for the stability ofthe international demand for the assetsissued by a given region as well as for thetransmission of financial shocks and for thechoice of exchange rate regime. The projectwill investigate these issues by analysingbilateral patterns in global bond holdingswith a particular emphasis on the impact ofEMU. The compositions of the selectioncommittees in 2003 and 2004 are detailedon the network website (http://www.eu-financial-system.org/fellowships.html).

3.3 PUBLICATIONS

Dissemination of the work undertaken withinthe network is through the normal academicpublication process via peer-reviewedjournals. However, in order to disseminateresults from the network at an early stage, theSC decided to suggest the possibility ofreleasing papers presented in the workshops orthe symposium in the ECB Working Paper

Series. It is expected that 10 to 20 of the bestpapers presented in the workshops or submittedto the symposium will be released in this way.

In addition to the publication in the ECBWorking Paper Series, two academic journalsare planning special issues. First, the “OxfordReview of Economic Policy” is preparing aspecial issue on European financialintegration, building on the work of a selectionof network contributors. The authors wereasked to prepare broad papers, discussing themost important developments in the main partsof the European financial system. The editorsXavier Freixas (University of Pompeu Fabra),Philipp Hartmann (ECB) and Colin Mayer(University of Oxford) will prepare a generalassessment of the papers published and discusspolicy options. The outline of this specialissue, which will be published in December2004, is in Annex H.

Second, some of the symposium papers will besolicited for a special issue of “Review ofFinance”, the journal of the European FinanceAssociation, devoted to capital markets andfinancial integration in Europe. All the papersin this issue will be screened through thejournal’s regular refereeing process. FranklinAllen (University of Pennsylvania) and MarcoPagano (University of Naples Federico II) fromthe network SC will edit this special issue.

Finally, there are plans to combine the papersof the “Oxford Review of Economic Policy”special issue with about 20 furthercommissioned papers for a book on “FinancialMarkets and Institutions: A EuropeanPerspective”. This volume would be edited bythe same three persons for Oxford UniversityPress. It will cover sections on “Financialsystems and financial integration”, “Financialsystems and the corporate sector”, “Financialinstitutions”, “Financial markets”, and“Financial regulation and macroeconomicpolicy”. The time horizon for the preparation ofthis book is longer, extending to the end of2005.


3.4 WEBSITE

A website for the network has been establishedas a source of information for researchers andpractitioners interested in European financialintegration. It can be accessed at the followingaddress http://www.eu-financial-system.org.It contains important information on Europeanfinancial integration that can be accessed eitherthrough links or through a library of researchpapers and official reports. As such, itfacilitates the dissemination of the informationrelated to network events and the work donewithin the network. Over time, the website hasbeen developed into a source of information forresearchers worldwide interested in Europeancapital markets and financial integration. TheIT infrastructure of the network website ishosted by the CFS. It is maintained by thenetwork secretariat as a service to the networkcontributors and to the general public.

The website contains inter alia: A descriptionof the network’s goals and main research areas;announcements of upcoming workshops ndsummaries of past events; information aboutfunding of research in the field of the network,notably the Lamfalussy Fellowship program;available statistical resources; and a library ofresearch papers and policy documents in thearea of the network.

The library has a search function that allowsusers to search for research areas, keywords,authors, etc. To a large extent, research papershave been chosen among unpublished work.However, the library also includes seminalpublished papers and books of special interest.

Now that the ECB-CFS research networkwebsite has been up for almost 24 months,some usage statistics are available. The ECB-CFS website had 209 “unique visitors”4 in 2002(covering a period of 3 months), 10,188 uniquevisitors in 2003 and 11,545 unique visitors in2004 (covering a period of 9 months). Therewere 1.7 visits per visitor in 2002, 1.4 in 2003and 1.4 in 2004. A visit is here defined as eachnew incoming visitor who was not connected to

the site during the last hour. A visitor viewed ordownloaded a page, image or file from the site1.1 times per visit in 2002, 4.8 in 2003 and 4.1in 2004.

Another relevant statistic is the page viewed bythose visitors. We restrict this measure to 2003and 2004 given the short period in which thewebsite was active in 2002. In 2003, 35% ofvisits viewed the home page (35% in 2004),11% (8% in 2004) viewed the library and 6%(8% in 2004) viewed the page related toupcoming events. To give an idea of who usesthe website, we report the origin of visitors in2003 only. National central banks viewed ordownloaded a page, image or file 4.9 times pervisit,5 followed by the European Commission(1.1 times/visit) and the Federal Reserve Board(0.5 times/visit).

3.5 RELATED INITIATIVES

This section describes other initiatives that fallinto the scope of the network and have been, forthe most part, initiated by participants to thenetwork, but do not fall into the aboveactivities.

The Second ECB Central Banking Conferenceon the “Transformation of the EuropeanFinancial System” took place in Frankfurt inOctober 2002. The purpose of the conferencewas to convey the interest of the ECB in thetransformation of the financial system via,among other things, financial integration. Theconference was based on four papers. JeanDermine (INSEAD) addressed the lessons thatcan be learnt for banking system integrationfrom the corporate structures adopted bybanking firms. Raghu Rajan and Luigi Zingales(both University of Chicago) describedsubstantial development of market finance inEurope, as compared with more traditional

4 A unique visitor is def ined as a host (IP address) who came tovisit the site and viewed at least one page.

5 Identif ied NCBs were the Bank of Finland, theOesterreichische Nationalbank, the Banco de Portugal,Sveriges Riksbank and the Banco de España.

15ECB



N E T WO R Kbank credit financing. Kpate Adjaouté (HSBCRepublic Bank) and Jean-Pierre Danthine(Lausanne University) looked at EMU’simplications for the pricing of governmentbonds and equities, arguing that theintroduction of the euro had a larger effect onequity returns than usually perceived. Finally,Bruce Carnegie-Brown and Matt King (both JPMorgan) studied corporate bond markets,advancing the hypothesis that the Europeanissuance boom following the introduction ofthe euro may not be explained by the singlecurrency. The proceedings of the Second ECBCentral Banking Conference were published inMay 2003.6

In October 2002, the ECB published the“Report on Financial Structures” incollaboration with the national central banks ofthe Eurosystem. The report aims at serving as areference work for researchers, policy makersand the general public. It describessystematically the structure of the financialsystems of the countries constituting the eurozone. The report shows that the euro areafinancial system is both deep and highlydiversified. However, it is changing rapidly.While credit institutions still play a pivotal rolein the euro area economy, non-financial sectorshave been redirecting their funds increasinglytowards new forms of financial intermediation.The report also notes the increased scope fornon-financial corporations to use debtsecurities for their financing.

The policy relevance of financial integrationand financial linkages that the network set as amain priority area at its inception two years agohas been confirmed by various initiativesrecently taken by other institutions. Forexample, the IMF research department hasstarted a strongly related “Global Linkages”initiative. It organised a conference on “GlobalLinkages” in Washington, D.C. on 30-31January 2003 to explore how economiclinkages across countries have changed inrecent years and what implications thesechanges have for policy makers in developedand emerging markets. This initiative is

6 Gaspar, V., P. Hartmann and O. Sleijpen (eds., 2003), TheTransformation of the European Financial System,Proceedings of the 2nd ECB Central Banking Conference,Frankfurt: European Central Bank.

strongly related to the network priority oninternational financial linkages.

The OECD also appears to be planning a relatedinitiative on financial integration and theability of financial systems to absorb shocks.

Finally, a purely academic network initiativeby the Center for Economic Policy Research(CEPR) on “International financialintegration”, aiming to improve understandingof the link between financial integration andfinancial efficiency, and between financialefficiency and economic performance, hasrecently been submitted for funding to theEuropean Commission. This proposedinitiative is quite similar to the presentnetwork. The Commission decided to put it onhold.


This section reviews the main results of thenetwork, going through the five priority areas,as formulated in the network “roadmap”. Whatfollows is a selection of particularly relevantresults, as it is impossible to summarise allpapers that were contributed to the networkevents. A summary of all contributed papers isavailable in Annexes B, D, E and F.

4.1 RESULTS IN PRIORITY AREAS

4.1.1 BANK COMPETITION AND THEGEOGRAPHICAL SCOPE OF BANKING

Results reveal that, first, integration is not veryadvanced in many retail banking markets.Second, some of the inherent characteristics ofthe traditional loan and deposit businessconstrain the cross-border expansion ofcommercial banking, even in a commoncurrency area. Hence, the implementationof some policies to foster cross-borderintegration of retail banking may beineffective. Third, theoretical researchsuggests that supervisory structures may not beneutral towards further European bankingintegration. Finally, a stronger role of area-wide competition policies could be beneficialfor further banking integration. This wouldalso stimulate economic growth, as morecompetition in the banking sector inducesfinancially dependent firms to grow more.

The ECB Occasional Paper on measuringfinancial integration in the euro area by LievenBaele (Ghent University), Annalisa Ferrando,Peter Hördahl, Elizaveta Krylova and CyrilMonnet (all ECB; see the symposium) findsthat the degree of integration varies in thedifferent segments of the retail bankingmarkets. In the corporate lending market,short-term and medium- and long-term lendingmarkets are still segmented. Short-termcorporate lending is least integrated.Household mortgage loan rates appear to havebecome more uniform across countries, whilethe consumer credit segment remains highlyfragmented. This price-based evidence is

4 R E SU LT S O F T H E N E TWORKconfirmed by quantity-based evidence showingthat cross-border activities within the euro areaare still limited in the retail-banking markets.In this sense, there are also clear signs ofpersistence in strong home biases in lendingand borrowing to non-financial corporationsand households.

Findings presented by Steven Ongena (TilburgUniversity) at the launching workshop and thesecond workshop confirm the home bias’persistence. Empirical evidence shows thatlarge multinational corporations still prefersmall local institutions to global financialinstitutions for their local cash management –i.e. short-term lending, liquidity management,etc. This is surprising, as one would expectmultinational corporations to be the firstbeneficiaries of global banks. His studies alsopoint to informational and political barriersthat limit mergers and acquisition in banking.In particular, there is evidence that banks“over-invest” domestically. Furthermore LuigiGuiso (University of Sassari), Paola Sapienza(Northwestern University) and Luigi Zingales(University of Chicago) reported evidence inthe launching workshop that even if financialmarkets become increasingly integrated,domestic financial institutions do not becomeredundant. These results suggest that localfinancial development and therefore localbanks are an important determinant of aregion’s economic success, even in anenvironment where there are no frictionsimpeding capital movements. All in all,traditional retail loan and deposit businessappears to solve economic frictions in a waythat is difficult to reconcile with cross-borderexpansions comparable to the one observed forwholesale business.

Simple measures such as facilitating domesticcompetition have been put forward in order tofacilitate mergers and acquisitions. Leo Kaas(University of Konstanz and Lamfalussyfellow) has addressed theoretically theconsequences that policies favouring foreignbanks’ entry into a domestic banking sectorwould have on competition, bank stability and

17ECB


4 RESULTSOF THE

NETWORKeconomic welfare. His results suggest thatpolicies favouring foreign banks’ competitionwould improve welfare if incumbent bankswere somehow less efficient. In addition, twopapers have studied the effects of competitionin banking. In the symposium, Hans Degryseand Steven Ongena (both Tilburg University)reported empirical results suggesting that whenbank branches face stiff local competition, theyengage relatively more in relationship-basedlending. The effect of competition on industryspecialisation is much less pronounced. Theauthors concluded that branches of a bankengage somewhat fewer borrowers in the sameindustry if local market concentrationdecreases. Finally, they also reported that theprobability of observing relationship bankingdecreases significantly with the distanceseparating the borrower from the lender. StijnClaessens (University of Amsterdam) and LucLaeven (World Bank) tested empiricallywhether competition in the banking sector isbeneficial to economic growth. Their resultsdepend upon the degree of financialdevelopment. In less developed countries,sectors that are financially dependent growslower when the financial system is morecompetitive, while in developed countriesmore competition is associated with highergrowth. More precisely, financially dependentfirms will grow by 1.5% per annum more if thecountry’s financial sector is more competitive.These findings support the view that marketpower in banking systems might be beneficialto less developed countries but detrimental toindustrial countries.

The regulatory and supervisory framework wasfound to affect the retail credit market. In thelaunching workshop, Gayle DeLong (BaruchCollege) and Claudia Buch (Kiel Institute ofWorld Economics) found that regulation is adriving factor behind international mergers:Banks operating in a more regulatedenvironment are less likely to be the target ofinternational bank mergers. Hence, regulatorybarriers are an impediment to further mergersand acquisitions. Also in the launchingworkshop, Harry Huizinga (Tilburg University

and European Commission) and GaetanNicodeme (European Commission) found thatinternational non-bank depositors appear tofavour banking systems covered by explicitdeposit insurance and they are attracted tosystems with co-insurance, a privateadministration, and a low deposit insurancepremium. The sensitivity of non-bank depositsto deposit insurance policies opens up thepossibility of international regulatorycompetition in this area. A theoretical analysisby Giovanni dell’Ariccia (IMF) and RobertMarquez (University of Maryland), alsopresented at the launching workshop, showsthat a centralised regulator would increaseefficiency at the cost of flexibility in applyingdifferent regulations to countries with differentfinancial systems. The benefits of a singleregulatory framework therefore heavily dependon the symmetry in the financial systems of therelevant countries.

Finally, the joint consequences of bank sectorconsolidation on reserves holdings and inter-bank market liquidity have been analysedtheoretically by Elena Carletti (CFS), PhilippHartmann (ECB) and Giancarlo Spagnolo(Sveriges Riksbank and University ofMannheim), who presented their results at thelaunching workshop. They conclude that amerger wave leading to greater banking systemheterogeneity is more likely to generate anadverse outcome in terms of the aggregateliquidity needed. In contrast, a mergermovement that leaves behind relatively littleheterogeneity in banks’ balance sheets mayleave inter-bank market liquidity unaffected oreven improve it.

4.1.2 EUROPEAN BOND MARKETS

While the government bond market hasintegrated rapidly with the EMU convergenceprocess, its full integration has not yet beenachieved. The introduction of a commonelectronic trading platform reducedtransaction costs substantially, but yieldspreads of long-term sovereign bonds of theeuro area are still heterogeneous. This is


largely explained by different sensitivities toan international risk factor, whereas liquiditydifferentials only play a role in conjunctionwith this latter factor. Somewhat surprisinglyin this context, the dynamically developingcorporate bond market exhibits a relativelyhigh level of integration. There is alsoincreasing evidence that the introduction of theeuro has contributed to a reduction in the costof capital in the euro area, in particularthrough the reduction of corporate bondunderwriting fees. As a result, firms may wishto increase bond financing relative to equityfinancing. The development of a largercorporate bond market is also important formonetary policy. For example, US evidencesuggests that the rating of corporate bonds maycontribute to the persistence of recessions, asrating agencies’ policies affect firmsasymmetrically in their access to the bondmarket over the business cycle. US evidencealso suggests that liquidity conditions in stockand bond markets tend to be positivelycorrelated.

The government bond market integratedsignificantly with the EMU convergenceprocess and with some efforts to harmoniseissuing procedures and conventions. Forexample, there is also a longer trend in thegradual reduction in the home bias ofgovernment bond investments. However,empirical results from the ECB OccasionalPaper on measures of integration presentedat the symposium also indicate that fullintegration of the government bond market hasnot yet been achieved. For example, while thelevel of convergence in yields is impressive,yields of government bonds with similar, or insome cases identical, credit risk and maturityhave not entirely converged. Typically, yieldson 10-year euro area government bonds maydiffer by around 15-20 basis points betweendifferent countries.

To explain the persistent yield differentialsobserved between long-term sovereign bondsin the euro area, Carlo Favero (BocconiUniversity), Marco Pagano (University of

Salerno) and Ernst-Ludwig von Thadden(University of Mannheim) identify the relativeroles of an international risk factor (measuredas the differential between high-risk UScorporate bonds and US government bonds)and liquidity differentials. At the symposium,they argued that liquidity differentials alone donot explain much of the yield differentials. Thedifferentials are largely explained by varyingsensitivities of local yields to the internationalrisk factor. Liquidity only plays a role wheninteracting with the international risk factor.

Most European government bonds are nowtraded on MTS, the single internationalplatform. Lamfalussy fellow Albert Menkveld(Vrije Universiteit Amsterdam), Yiu Cheungand Frank de Jong (both University ofAmsterdam) conducted a microstructure studyof MTS, showing that national orderimbalances appear to be diversifiable across allmarket participants. More precisely, none ofthe national order imbalances in thegovernment bond markets of Germany, France,Italy or Belgium affects benchmark yield (i.e.German yield) innovations. Governments havealso a lot to gain from integration and welldesigned markets. The design of markets iscrucial for their integration and the prevailingprices. Specifically, differences in themicrostructure of the European treasury billsmarket across countries and through time affectshort-term yields in government treasuryauctions, and thus the cost of funds forgovernments. In particular, as shown in thepaper by Bruno Biais, Antoine Renucci andGilles Saint-Paul (all University of Toulouse)presented at the symposium, regularly issuingbills significantly reduces yields. Also, whenbills are traded on a centralised, transparentelectronic limit order book, such as MTS, theirliquidity rises and the yields declinesignificantly. Governments could thereforeenhance liquidity and reduce yields and thecosts of their funds by efficiently designingTreasury securities and issuing procedures, aswell as by promoting modern trading systems.Biais et al. estimate, for example, that bydesigning Treasury auctions more efficiently

19ECB


4 RESULTSOF THE

NETWORKgovernments could raise up to €350.19 millionmore than they do presently.

The ECB Occasional Paper on measures ofintegration also reports price-based integrationmeasures for corporate bond markets in asubset of six euro area countries. Thesemeasures suggest that, somewhat surprisingly,the level of integration is reasonably high inthis young and rapidly expanding market.Specifically, the country where a bond is issuedhas only marginal explanatory power for thecross-section of corporate bond yield spreads,once a number of systematic risk factors arecorrected for. Quantity-based indicators tendto support this conclusion. Considering theproportion of assets invested in bond marketfunds with a European-wide investmenthorizon between 1998 and 2002 the marketshare of European-wide managed bond funds(both government and corporate) increaseddramatically, from below 20% to above 60%(on average), indicating a drastic reduction inthe home bias of bond portfolios in the euroarea. This is even more striking whencomparing it with the home bias in cross-Atlantic portfolios of corporate loans. At thesymposium, Mark Carey and Greg Nini (bothFederal Reserve Board) provided empiricalevidence that interest rate spreads onsyndicated corporate loans are on average 30basis points lower in Europe than in the US.The authors conclude that the syndicated loanmarkets in Europe and the US might not befully integrated and presume that theexplanation for the home bias they observe inthe data may be an important factor forunderstanding this pricing difference.

There is increasing evidence that the euro’sintroduction has contributed to a reduction inthe cost of capital in the euro area. Arturo Bris(Yale University), Yrjo Koskinen (StockholmSchool of Economics) and Mattias Nilsson(Stockholm Institute for Financial Research)showed at the second workshop that theintroduction of the euro has lowered firms’ costof capital by further increasing capital marketintegration in Europe and by eliminating

currency risks among the countries that joinedEMU. More precisely, they showed that thevaluations of large firms in euro land (asmeasured by Tobin’s Q) in the period 1998-2000 increased by 7.9% per year relative tofirms in non-EMU countries, after controllingfor firm-, country- and time-specific effects.

Also, at the launching workshop, KostasTsatsaronis (BIS) and João Santos (FederalReserve Bank of New York) showed that theeuro led to a reduction in the underwriting feesof international corporate bonds issued in thenew currency and that this reduction is largelyexplained by greater contestability of theinvestment banking business in the post-EMUEuropean market. There was a clear globaldownward trend in fees over the 1994-2001period, with value-weighted average fees for2001 standing 86 basis points below their 1994levels, the equivalent of a 37% reduction,which is largely attributable to a sharp drop inthe euro-denominated segment. As a result,underwriting fees for euro-denominatedcorporate bonds are now at the same level as inthe dollar segment of this market.

As a consequence of the lower borrowing costsfor firms, bond markets are expected to developfurther. This is important, given that GoizuetaTaurn Chordia (Emory University), ArsaniSarkar (Federal Reserve Bank of New York)and Avanidhar Subrahmanyam (University ofCalifornia, Los Angeles) showed that theliquidity of US Treasury bond markets issignificantly and positively correlated with themonetary policy stance during crises. Theyfound that the ratio of net borrowed reservesover total reserves in crisis times explainsabout 4.5% of the variation in bond spread after2 months.7

Another issue in firm financing is the issuanceof corporate bonds. In the third workshop,Yigal Newman (Stanford University and

7 The authors consider three crises: The bond market crisis of1994, the Asian f inancial crisis of 1997 and the Russiandefault/LTCM crisis of 1998.


Lamfalussy fellow) and Michael Rierson(Citigroup) documented that a very large firm’snew issues of corporate bonds can temporarilyraise the yield spreads of bonds issued by otherfirms.8 They also show that the increase in yieldspreads begins well before the issuance date.The temporary aspect of the effect and the factthat it happens relatively close to the date ofissuance rather than to the announcementsuggests that this effect is not due to newfundamental information about the issuer thatmay be revealed during the issuance process.

Finally, João Santos argued in the thirdworkshop that rating agencies’ policies affectfirms asymmetrically in their access to thebond market over the business cycle. As aconsequence, the impact of recessions acrossfirms is not uniform. The cost of capitalincreases most for mid-credit quality firms.As ratings become increasingly uncertainin recessions, information becomes moreasymmetric. Uncertainty about the quality ofthose firms then raises the cost of access tocapital. As a consequence, recessions will tendto last longer in the presence of asymmetricinformation than in its absence.

4.1.3 EUROPEAN SECURITIES SETTLEMENTSYSTEMS

European securities settlement infrastructuresare highly fragmented and further integrationand/or consolidation would exploit economiesof scale that could greatly benefit investors. Itis not clear, however, whether direct publicintervention in favour of consolidation wouldlead to the highest level of efficiency, forexample because of the existence of strongvertical integration between trading andsecurities platforms (“silos”). In contrast,promoting open access to clearing andsettlement systems could lead to consolidationand the highest level of efficiency. Finally,regarding concerns about unfair practices byCentral Securities Depositories (CSDs)toward custodian banks, regulatoryinterventions favouring custodian banksshould be discouraged, as long as CSDs are not

allowed to price discriminate betweencustodian banks and investor banks.

The two Giovannini reports (2001, 2003)stressed the need to improve significantly thestructure of securities clearing and settlementsystems in Europe, especially regarding cross-border settlement. Legal and technical barriersto further integration were highlighted. Giventhe Giovannini group’s already verycomprehensive work in analysing barriers toconsolidation, the network focused onsomewhat complementary issues. Two pointswere particularly addressed. First, there is a lotto gain in Europe from further consolidation ofsecurities settlement systems. Second, inaddition to the barriers highlighted in theGiovannini reports, there may be intrinsicfeatures of the securities trading and settlementindustry that prevent consolidation.

Regarding the first point, Markku Malkamaki(Bank of Finland), Heiko Schmiedel (ECB) andJuha Tarkka (Bank of Finland) investigated theexistence and extent of economies of scale indepository and settlement systems. Theypresented their results at the third workshop.They showed that settlement in Europe is 33%more costly than in the US, as the average costper settled transaction is $3.86 in Europe andonly $2.90 in the US domestic market. Thisdifference is partly explained by segmentationin the European market, as the average cost foroperating an international CSD in Europe is$40.54, relative to $3.11 for a domestic one.However, looking at the exploitation ofeconomies of scale in Europe and the US,they also showed that systems on the otherside of the Atlantic are operating at a muchmore efficient level. The European settlementinfrastructures show a strong potential for costsaving: Costs will increase by a factor of only0.7 when the number of instructions increasesby 1 – whereas the analogous factor for the USis 0.9. Hence, Europe has a lot to gain from

8 For this paper, Newman received the 2004 European FinanceAssociation award for the best Ph. D. paper.

21ECB


4 RESULTSOF THE

NETWORKfurther consolidation. This has also beenshown in the symposium by Jens Tapking(ECB) and Jing Yang (Bank of England), whotheoretically analysed, in a two-country model,welfare implications of vertical and horizontalintegration. They show that both verticalintegration of trading platforms and settlementsystems and horizontal integration ofsettlement systems are welfare improving.However, given the level of complexity in EUinternational securities settlement systems,the effectiveness of settlement industryinfrastructure may benefit from furtherinitiatives simplifying the procedure for cross-border settlement, as for instance advocated inthe second Giovannini report (2002).

Cyril Monnet (ECB) and Thorsten Koeppl(ECB) argued at the third workshop that liftingall legal and technical barriers to consolidationmay not suffice to ensure that the best forms ofconsolidation take place. They analysedtheoretically the role of vertical “silos” insecurities market organisation for efficienthorizontal consolidation between componentsof the silo, i.e., exchanges and back-officeoperations such as clearing and settlement. Anefficient merger is characterised by the lowestcost of clearing and settlement. The papershows that it is impossible to achieve such amerger when silos are in place because the lackof information about the competitor’s coststructure raises the costs of achieving anefficient merger. These additional costs cannever be covered with the revenues of themerger, as they increase with the revenues. Theauthors then showed that exchanges canachieve an efficient merger by outsourcingtheir own settlement operation, respectively, aslong as each settlement system competes forsettling all trades of the merged exchange.Hence, they argue that fostering competitionand open access to securities settlementsystems may be required.

Regarding the efficient design of the structureof securities settlement systems, CorneliaHolthausen and Jens Tapking (both ECB)tackled the issue of the pricing strategy of

Central Securities Depositories relative tocustodian banks at the third workshop. BothCSDs and custodian banks provide the sameservice, but custodian banks often need toresort to CSDs. In an environment whereCSDs’ pricing strategies do not discriminatebetween custodian banks and usual investorbanks, Holthausen and Tapking showed thatCSDs can increase the costs of custodian banksby increasing the variable part of their priceschedule. The equilibrium market share of theCSD is relatively high and the equilibrium isusually not efficient. However, whether theefficient market share of the CSD is higher orsmaller than the equilibrium market sharedepends on the parameters. The authorsconcluded that regarding concerns about unfairpractices by CSDs toward custodian banks,regulatory interventions favouring custodianbanks should be discouraged, as long as CSDsare not allowed to price discriminate betweencustodian banks and investor banks.

Also related to the efficient design of thisindustry, Giulia Iori (University CollegeLondon and Lamfalussy fellow) presented atthe symposium a study on the efficiency andstability of alternative designs for securitiesclearing and settlement infrastructures. Sheassumes that settlement takes place in batchesthroughout the day and that settlement can bedelayed. She found that increasing thefrequency of settlement (and thereforeapproximating real-time settlement) increasesthe likelihood of failure but reduces thesystemic effects from a failure. As aconsequence, the shorter the interval betweensettlement batches, the more stable grosssettlement systems are compared with netsettlement systems, and vice-versa.


4.1.4 THE EMERGENCE AND EVOLUTION OFNEW MARKETS IN EUROPE (INPARTICULAR START-UP FINANCINGMARKETS)

Relative to the United States, European “newmarkets” and start-up financing are relativelylittle developed and integrated. However, newmarkets and venture capitalists are the mostimportant intermediaries for the financing ofprojects with high risk and with potentiallyvery high return. The analysis carried outwithin the network reveals that European start-up financiers are mostly institutionalinvestors, while US venture capitalists aremostly rich individuals. Also, new equitymarkets are essential for the development ofstart-up finance in Europe, as they provide anexit strategy for start-up financiers, who canthen sell new successful firms using initialpublic offerings. Finally the legal frameworkaffects the development of venture capitalfirms. For example, very strict personalbankruptcy laws constrain early stageentrepreneurs, reducing demand for venturecapital finance.

Marco Da Rin (University of Turin) describedat the third workshop key characteristics andissues in the European venture capital industry.During the 1990s, the European venture capital(VC) industry invested 50% less than its UScounterpart, showing the relativeunderdevelopment of this industry in Europe.Also, its characteristics are different. TheEuropean landscape of VC firms is highlycaptive, with an important presence of banksubsidiaries, corporate VC firms and publicVCs. Typically, these types of VC firms investless in early stage and high-technology projectsand provide less soft support to privatecompanies compared with individual venturecapital. There is also evidence that the goal ofEuropean VCs is to sell their firms. As shownby Giovanna Nicodano, Marco Da Rin andAlessandro Sembenelli (all University ofTurin) at the third workshop, the creation ofNew Markets could therefore foster thecreation of VC firms, as additional exit

opportunities create further incentives for VCsto invest. However, Da Rin reported findingssuggesting that European VCs might actuallynot make a difference for the companies theyare financing. While VC-backed companiesraise more capital with IPOs than mature firms,they do not tend to grow faster than these otherfirms. At the third workshop, Tereza Tykvová(Center for European Economic Research,Mannheim) and Uwe Walz (University ofFrankfurt and CFS) brought additionalevidence that firms backed by bank-dependentand public VCs have significantly lowermarket value relative to firms backed byindependent VCs. In general, firms backed byindependent VCs perform better and display alower return volatility than firms of other VCsor non venture-backed firms.

Marco Da Rin (University of Turin) reportedthat VC firms are investing very locally andusually less than 10% of the partners come fromabroad. Also, cross-border investments of VCfirms represents less than 2% of totalinvestments. Finally, less than a third of fundsoriginate from foreign investors. Most foreigninvestors are from the US and are concentratedin only a small number of firms. This shows avery low degree of integration in this sector. Asfor characteristics limiting integration in thecredit markets, this may be explained by thenature of the VC industry, which is quitedifferent from that of other financialintermediaries: VC makes localised andundiversified investments; is based on humanrather than financial capital and has a smallnumber of investors. However, financialintegration could indirectly restructure the VCindustry through its affect on the allocation offunds and the changes in the EU economicstructure, notably on the listing ability of newfirms. In this regard, financial integration mayimprove exit channels for VC and reallocatetalent and human capital.

Finally, the legal framework surrounding earlyfinance can also greatly affect the VC industryand the development of New Markets. At thethird workshop, Douglas Cumming (University

23ECB


4 RESULTSOF THE

NETWORKof Alberta) and John Armour (CambridgeUniversity) reported that very strict personalbankruptcy laws discourage early stageentrepreneurs and therefore significantlyreduce the demand for VC finance.

4.1.5 INTERNATIONAL PORTFOLIO CHOICESAND ASSET MARKET LINKAGESBETWEEN EUROPE, THE UNITED STATESAND JAPAN

At a global scale, asset market linkages haveincreased recently. For example, majoreconomies such as the United States and theeuro area have become more financiallyinterdependent. This phenomenon can beobserved in stock and bond markets as well as inmoney markets, where the main direction ofspillovers has recently been from the US to theeuro area. Country-specific shocks now play asmaller role in explaining stock returnvariations of firms whose sales areinternationally diversified. Increases in firm-by-firm market linkages are a globalphenomenon, but they are stronger within theeuro area than in the rest of the world. Variousother phenomena also increase market linkagesand therefore the likelihood that financialshocks spread across countries. One example isthe use of global bonds. Finally, the nowadaysmore direct access of unsophisticated investorsto financial markets may increase volatility.

Robin Brooks (IMF) and Marco Del Negro(Federal Reserve Bank of Atlanta) reported atthe launching workshop that the degree of co-movement across national stock markets hasincreased dramatically in recent years. Theyfound that the ability of country-specificeffects to explain international variation inasset and sales growth and return fellsignificantly during the late 1990s, while theexplanatory power of global industry effectsincreased and in some cases surpassed that ofcountry effects. Yet, this question is notsettled. Although Geert Rouwenhorst (YaleUniversity) reckoned at the second workshopthat there are some grounds to believe thatinternational linkages are becoming stronger,

he presented evidence that country effects arestill large. Where firms are located still appearsto matter more than what they actually produce,although there is evidence that at the Europeanlevel industry effects are gaining furtherimportance.

Stronger international linkages may beexplained by growing cross-listings. MichaelHalling (University of Vienna), Marco Pagano(University of Salerno), Otto Randl and JosefZechner (both University of Vienna) reportedat the symposium that more companies arelisting their shares, not only on their domesticstock exchange, but also on foreign exchanges.They found that cross listing initially raisestrading volume in foreign markets, but adeclining trend then follows. Although thiswould suggest a return to the dominance of thedomestic market, the decline in foreign tradingis quite slow for certain companies. Foreigntrading volume turns out to be higher forexport-oriented companies and for companiesthat cross-list on foreign exchanges with lowertrading costs and better insider tradingprotection. Also, small, high-growth and high-technology firms tend to have relatively highforeign trading activity.

Investors, as firms, are seeking ways to exploitfinancing capacities of all markets by creatinginstruments, such as global bonds, that can besimultaneously traded in multiple markets.Darius Miller (Indiana University) and JohnPuthenpurackal (Ohio State University)reported at the third workshop that globalbonds are likely to be an expanding form offinance, as this instrument reduces the cost ofdebt capital. According to their study, theborrowing costs for global bonds are 15 basispoints lower than on those of comparable USdomestic bonds. Moreover, issuing costs ofglobal bonds are 13 basis points lower thanthose of US domestic bonds. Making thesetypes of instruments more attractive willundoubtedly bring additional linkages betweenEurope, the US and Japan, thus increasing therisk of volatility spillovers among the differentmarkets.


Examining the effects of monetary policyannouncements and macroeconomic news oninterest rates in the money markets, MichaelEhrmann and Marcel Fratzscher (both ECB) atthe second workshop already reported evidencethat the spill-over effects are stronger from theUS to the euro area than vice versa. They alsofound that since the introduction of the euro thecross-Atlantic interdependence of moneymarkets has steadily increased. In a similarvein, Michael Fleming (Federal Reserve Bankof New York) and Jose Lopez (Federal ReserveBank of San Francisco) examined at the thirdworkshop whether information from othertrading centres affect intra-market variancesfor US treasury bonds in London, New Yorkand Tokyo. They find strong evidence thatvolatility spills over from New York to Londonand Tokyo but not the reverse.

Increased volatility spillovers may becomeworrisome with the “democratisation” ofaccess to financial markets. Michael Haliassos(University of Cyprus), Luigi Guiso(University of Sassari) and Tullio Jappelli(University of Salerno) presented evidence atthe second workshop that lower access costsbrings less sophisticated investors into stockmarkets, with the potential consequence ofinducing greater volatility. For example,unsophisticated investors can react excessivelyto market signals, e.g., because of smallinvestors’ limited ability to withstand financialpressure. This suggests that it would be wise toconsider policies that can reduce volatility,such as improving the flow of accuratefinancial information.

Cross-border asset holdings are alsoencouraged by the harmonisation of securitiesregulation, as Jonas Vlachos (University ofChicago) showed at the symposium.Institutional or cultural differences havenegative effects on bilateral asset holdings, andresults for regulatory differences are robusteven after taking these effects into account.Hence, with increased cross-border assetholdings, local markets will be populated withinvestors that have more diverse information

and portfolios. The presence of heterogeneousinvestors in stock markets has been analysedby Rui Albuquerque (University of Rochesterand Lamfalussy fellow), Gregory Bauer(University of Rochester) and MartinSchneider (New York University), who showedthat this heterogeneity is crucial for explaininginternational portfolio choices. They propose amodel of international portfolio choice whereinvestors are heterogeneous, both within acountry and across countries. Bringing themodel to the data, their main finding isthat domestic heterogeneity of investors ismuch more important than cross-countryheterogeneity.

4.2 RESULTS IN OTHER AREAS

Financial integration affects financialstructures, but it does not need to lead to theirconvergence across countries. Financialstructures matter for growth, as market-oriented financial systems benefit all sectorsand firms, whereas bank-based systemsprimarily benefit younger firms that depend onexternal finance. Moreover, good corporategovernance increases firms’ value. Inparticular, the dual board system, where themonitoring and advising roles of the board ofdirectors are separated, is found to dominatethe single board structure. Therefore, thefurther development of the European singlemarket should strongly require good corporategovernance. In general, well designedinstitutions foster entrepreneurial activity,partly by relaxing capital constraints.

In “The euro-area financial system: Structure,integration and policy initiatives”, PhilippHartmann, Angela Maddaloni and SimoneManganelli (all ECB) provide an overview ofthe current structure and integration of theeuro-area financial systems and related policyinitiatives. They document how the euro-areafinancial structure is placed somewhat inbetween those of the US and Japan, withfinancial institutions playing an important role,but with market-based instruments developing

25ECB


4 RESULTSOF THE

NETWORKfurther. They find that financial integration hasprogressed quickly in some of the main euro-area financial segments, such as the unsecuredmoney market, but rather slowly in others, suchas retail banking. Despite further financialintegration, they also show evidence that invarious dimensions the financial structures ofeuro-area countries seem to become morediverse over time. Apparently, progress infinancial integration does not implyconvergence of financial structures.

At the symposium, Solomon Tadesse(University of South Carolina) providedempirical evidence of the impact of financialstructure (bank-based or market-based) ontechnological innovation. Market-basedfinancial systems have an overall positiveeffect on technological progress. However,financial structure appears to affect industriesheterogeneously. In particular, industrieswhose small and young firms are relativelymore dependent on external finance fare betterin bank-based financial systems. Hence,financial structure plays an important role inshaping a country’s industrial structure.

Related to financial structure is the question ofwhy banks adopt multiple lending, i.e., whybanks share financing of a single firm withother banks. Elena Carletti (CFS), VittoriaCerasi (University of Milan) and Sonja Daltung(Sveriges Riksbank) tackled the effects ofmultiple-bank lending at the symposium. Firstit improves banks’ monitoring incentives byallowing them to finance more projects andtherefore achieve greater diversification.Second, it entails free-riding problems andduplication of efforts, thus reducing banks’incentives. Multiple lending is optimalwhenever the first effect dominates the second.The model predicts a greater use of multiple-bank lending when banks are small relative tothe projects they finance, when firms are lessprofitable and when poor financial integration,strict regulation and inefficient judicialsystems make monitoring more costly.

Finally, three noteworthy papers on firmfinancing and corporate governance werepresented at the symposium. First, Mihir Desai,Paul Gompers and Josh Lerner (all HarvardUniversity) investigated the importance of theinstitutional framework for entrepreneurialactivity in Western and Central and EasternEuropean (CEE) regions. The authorsidentified a particular sensitivity ofentrepreneurial activity to institutional factors(corruption/fairness, protection of propertyrights, well-functioning legal system) incountries of the CEE region. In particular, lesscorruption and better protection of propertyrights increase entry and reduce exit of firms.This supports the view that well-designedinstitutions foster entrepreneurial activity inemerging countries, partly through the positiveimpact on relaxing capital constraints.

Second, Renée Adams and Daniel Ferreira(both Stockholm School of Economics)addressed the question of why – unlike in theUS – European governance structures rely on a“dual” board system, which separates themonitoring and advising roles of the board ofdirectors. In their model revealing informationand getting advice enables the manager to makebetter decisions, but this might increase hisprobability of getting fired when thisinformation changes the board’s opinion abouthis ability. This trade-off provides a rationalefor the board to reduce its monitoring activityup-front whenever it is not too costly to inducethe manager to reveal his information. Theauthors show further that the first-best level ofmonitoring can be attained when a “dual” boardsystem is used. Hence, when given the choicebetween a single and a dual board system, as inthe new European company statute, firms maybe advised to rather adopt the dual boardsystem.

Finally, Mariassunta Giannetti and AndreiSimonov (both Stockholm School ofEconomics) found at the second workshop thatinvestors are more likely to buy a firm’s stockwhen the ratio of control to cash-flow rights ofthe principal shareholder – expected to be


positively correlated with the extraction ofprivate benefits in a firm, and therefore usedas a proxy for bad governance – is lower.Improved corporate governance has a strongereffect on sophisticated investors, like financialinstitutions and foreign investors, while largedomestic investors and individuals who areboard members do not base their investmentdecisions on corporate governance grounds.Corporate governance is therefore an importantaspect in understanding portfolio choicesacross countries.

27ECB


5 FUTURE STEPS:CONTINUATION

OF THE NETWORKAND OF

THE LAMFALUSSYRESEARCH

FELLOWSHIPPROGRAM

In its two years, the network has achieved thegoals it formulated at the start. It provided aforum in which researchers from academia andpolicy institutions interested in the evolutionof the European financial system regularlymeet to discuss their findings. It deepened ourunderstanding of European financialintegration by mapping the current Europeanfinancial market landscape. It also providedtheoretical underpinnings for many of the factsobserved. Finally, the network stimulatedresearch on securities settlement systems, aresearch area that hardly existed before.

Based on these achievements, the network willcontinue for three further years, refocusingits priorities. Three priority areas havebeen added: 1) The relationship betweenfinancial integration and financial stability;2) EU accession, financial development andfinancial integration; and 3) financial systemmodernisation and economic growth in Europe.These three areas have become particularlyimportant at the current juncture, but have notreceived particularly strong attention in thepast two years of the network. The three sub-sections in this section show the motivation forchoosing the three additional priorities and listspecific questions that could be addressedunder them. Moreover, despite the resultsfound in the first two years of the network,there are new developments in the earlierpriority areas that merit further exploration.

The Lamfalussy research fellowship programhas also been reconfirmed as a tool forfostering research in the network’s priorityareas. Accordingly, five new fellowships havebeen attributed for projects pertaining to thenew areas of the network described below (seeSection 3.2).

5 FU TUR E S T E P S : C ON T I NUAT I ON O F T H ENE TWORK AND O F T H E L AM FA LU S S YR E S E A R CH F E L LOWSH I P P ROGRAM

5.1 THE RELATIONSHIP BETWEENFINANCIAL INTEGRATION ANDFINANCIAL STABILITY

While an integrated area offers moreopportunities to share risk and to allocatecapital, it might be less resilient to unexpectedand uninsurable shocks, as they may propagatewider and faster. Moreover, it is of interest toknow whether greater integration couldincrease the risk of cross-border contagion in afinancial crisis. With the integration ofEuropean financial markets going forward, it isimportant to understand what type of integratedfinancial structure is the most resilient. TheECB has an obvious interest in this, as inaccordance with Article 105(5) of the Treaty,the European System of Central Banks shall“contribute to the smooth conduct of policiespursued by the competent authorities relatingto the prudential supervision of creditinstitutions and the stability of the financialsystem”. Furthermore, the ECB researchevaluators highlighted financial stabilityresearch as an area deserving furtherdevelopment.9

As explained in the October 2003 article of theECB Monthly Bulletin on the integration ofEurope’s financial markets, a comprehensivereview of the EU arrangements for financialregulation, supervision and stability wasunderway. The review, which shouldcontribute to the further integration of the EU’sinstitutional financial architecture, wasinitially triggered by the report of theLamfalussy Committee on the regulation ofEuropean securities markets in 2001.10

The work of the research network on therelationships between financial integration andfinancial stability could therefore help shape aview of what this new financial architectureshould be. Particularly important questions

9 Goodfriend, M., R. Koenig and R. Repullo (2004), Externalevaluation of the economic research activities of the EuropeanCentral Bank, Frankfurt, ECB.

10 Committee of Wise Men (2001), Final report of the committeeof wise men on the regulation of European securities markets,Brussels, 15 February.


include: Are market-based or banked-basedfinancial structures more resilient to shocks?How can financial regulations be designed soas to be conducive to integration while stillensuring financial stability? What is the linkbetween financial integration and contagionrisk? What are the mechanisms that can triggercontagion and financial fragility? How and towhat extent do financial crises affect economicactivity? Finally, how does competition in thefinancial services industry affect the ability ofthe financial system to withstand shocks?

5.2 EU ACCESSION, FINANCIALDEVELOPMENT AND FINANCIALINTEGRATION

As expressed by the President of the ECB,“Assessing the impact of EU enlargement onthe European economy is a complex question.Despite the considerable complexitysurrounding this issue, one thing that is clear isthat EU enlargement will provide newopportunities to trade and investment flows.Many of these effects are visible because of thehigh degree of economic integration alreadyreached between the present Member Statesand the acceding countries.”11 Further financialintegration is likely to follow swiftly thepolitical and economic integration of thesecountries in the European Union. There aremany ways in which this financial integrationcan take place. One way, which currentlyseems to be a dominant one, is that foreignfinancial institutions acquire financialinstitutions of new member states.12 Anotherway would be the cross-border provision offinancial services. Clearly the first is likely tobe faster than the second and may developaccession country financial systems faster aswell. The example of the “old” EU memberstates suggests that there are many obstacles tothe direct provision of financial servicesabroad, particularly in retail markets. Theheavy presence of foreign financial institutionsmay, however, pose other challenges for thenew member countries, e.g., in the area offinancial stability and supervision. All these

developments will determine financialstructures, development and competition infinancial services across the enlargedEuropean Union.

Important questions for further work withinthe network are therefore the following:What factors explain differences in financialstructures across new and old EU memberstates? What role is financial integration takingin this development? What is the relationbetween financial integration and financialdevelopment? Who benefits from financialdevelopment in the process of integration?Are there any risks to financial stability duringfast changes in financial structures andinstitutional arrangements? How will financialintegration among accession countries advancerelative to integration between accessioncountries and the previous EU countries?

5.3 FINANCIAL SYSTEM MODERNISATIONAND ECONOMIC GROWTH IN EUROPE

The capacity of financial systems to promoteeconomic growth depends not only on theirlevel of integration, but also on their qualityand the efficiency with which they channelsavings into investment. In other words, whentackling the issue of economic growth, it wouldbe too narrow to place the focus entirely on thelevel of integration. While several financialsystems can be very financially integrated,they may not be developed in such a wayas to achieve a greater volume or efficiencyof financial intermediation, and thereforemay not improve the growth performance ofthe economy. Financial modernisation andfinancial development, however, improve theefficiency of financial intermediation, and theprocess is influenced by many factors otherthan financial integration. As the network hasso far contributed relatively little to issues

11 “The challenges for the European economy in 2004”, Speechby Jean-Claude Trichet, President of the European CentralBank, Conference organised by Foro de la Nueva Economiaand The Wall Street Journal, Madrid, 29 January 2004.

12 See “Financial Sectors in EU Accession Countries” (Ed.Christian Thimann), ECB, 2002.

29ECB


5 FUTURE STEPS:CONTINUATION

OF THE NETWORKAND OF

THE LAMFALUSSYRESEARCH

FELLOWSHIPPROGRAM

regarding the link between finance and growth,this topic has been added as a separate priorityarea.

Work could focus, in particular, on thefollowing issues: How can one further improvethe structure of highly developed financialsystems? How can one measure a financialsystem’s success in performing its functions?What are the costs and the benefits ofmodernising the financial system? What is thebest form of corporate governance? What arethe implications of uniform accountingstandards? What is the best way to incorporateimprovements in the structures of the financialsystem? What is the most effective way toenforce rules?

ECBCapital markets and financial integration in EuropeDecember 200430

The ECB-CFS network is organised as a“network of people” with a particularly lightstructure. The present section describes itscharacter as a network of people and its internalstructure.

6.1 A “NETWORK OF PEOPLE”

The ECB-CFS network was conceived as a“network of people”, based on the observationthat potential synergies between researchers inNCBs, the ECB and the academic sectors werehardly exploited. It provides a co-ordinationdevice for the best researchers to work togetherin areas of special interest to the ECB. Hence,a key feature of the network is a stronginteraction among all these researchers. A“network of people” means that the goal is notan exhaustive representation of institutions,but rather a collaboration among the best mindsresearching the network’s chosen topics,irrespective of the institution or department inwhich they work.

Contributions to the ECB-CFS network aremade through continuous active participationin the its organised events. Active participationcan take several forms, be it as a panelparticipant, a paper presenter, a discussant or asession chair. The list of contributors to thenetwork during its first phase is available inAnnex I. Leading researchers from academiaand policy institutions joined the network tobenefit from the continuous interaction on atheme of common interest. Participation in thenetwork implies a two-sided commitment. Onone side, the CFS and the ECB organise andprovide funding for meetings of the SC and forworkshops and conferences, establish anetwork website and intermediate contactsbetween academic and central bank networkparticipants. On the other side, researchersproduce original, publishable research in thefields of the network, present it at theworkshops and conferences and make itgenerally available on the network website.The combination of these factors in theframework of a network and the cross-

6 ORGAN I S AT I ON O F T H E N E TWORKfertilisation between researchers fromdifferent institutions over time results insynergies that lead to productivity and outputthat goes beyond what can be expected from thesimple one-off organisation of traditionalseminars and workshops. While the purpose ofthe network is to produce and stimulateresearch in the particular fields mentionedabove, no attempt is made by the SC toprescribe concrete research projects to networkcontributors. Explicit priority areas, policypanels and speeches are, however, used tosignal to contributors issues of particularimportance.

6.2 STRUCTURE

The network initiative was initially envisagedfor period of two years and has now beenextended for another three years. It is jointlyorganised by the ECB and the CFS. This sharedorganisation and the character of a “network ofpeople” had the consequence of keeping thegeneral organisational structure “light”, e.g.,by sharing the administrative burden.

The network is headed by a SteeringCommittee (SC). The SC is composed of twoECB representatives, two CFS representativesand leading academic scholars in the mainareas of the network. The composition of theSC is reported in Annex J. Apart frommanaging the network, the role of the SC is alsoto co-ordinate the involvement of researchersfrom Eurosystem NCBs and from otherEuropean and overseas central banks, such asthe US Federal Reserve; and to liaise withresearchers in the academic sector. The SCmembers located in Frankfurt meet regularly,usually every 2 to 4 weeks. Consultation withthe outside SC members is conducted throughelectronic mail.

Also, an NCB Contact Group and an ECBInternal Contact Group were set up to ensurethe necessary information exchanges withpolicy areas of the ECB and with NCBs. Thecompositions of the two groups are detailed in

ECBCapital markets and financial integration in Europe

December 2004

6 ORGANISATIONOF THE

NETWORK

31

Annex K. Members of the two contact groupstend also to be involved in the activities of thenetwork. The existence of these groups, interalia, allows the network to benefit from astronger consideration of the relevant policycontext.


ANNEX A TH E ROADMAP

TAB L E O F CONT EN T SEXECU T I V E S UMMARY 3 3

1 I N T R O D U C T I O N 3 4

2 THE S T RU C TUR E O F T H E “ ROADMAP ” 3 6

3 S P E C I F I C I S S U E S O F T H E N E TWORK 3 8

3.1 European financial integration andfinancial linkages with the UnitedStates and Japan 38

The geographical scope of bankingand lending relationships 38

The optimal degree andimplications of bank competition 38

Determinants of bank mergers 38

Implications of bankconsolidation 39

International portfolio choices ofinstitutional investors/determinantsof international portfolio flows 39

Asset price and volatility linkagesacross countries and their changes/contagion and crisis linkages 39

Determinants of market returns/determinants of inflation, liquidityand default risk premia 40

Liquidity in secondary markets/substitutability of government bondsfrom different sovereign issuers 40

Restructuring of settlementinfrastructures 41

Consolidation of stock exchanges 41

Foreign stock listings and the roleof depository receipts 41

Integration of repo markets/substitutability of government bondsfrom different sovereign issuers 42

Capital flows and exchange rates 42

Determinants of country investmentportfolios and the home biasphenomenon 42

Globalisation and the increasing roleof capital flows/synchronisation ofinternational business cycles 43

Goods market and financialintegration 43

3.2 Financial system structures inEurope 43

Bank versus market financing 43

The resiliency of the financialsystem 44

The efficiency of European banks 44

Insolvency regimes 44

Technology and financial services 45

The importance and determinantsof the breadth and completeness offinancial markets/financialinnovation 45

The role of legal systems 45

Market microstructures/the role ofelectronic trading platforms 45

The organisation of primarymarkets/equity versus debtfinancing/private versus publicequity markets/start-up financing 46

Corporate governance and control/ownership structures of financialintermediaries 46

Transparency and accounting rules/accounting conventions and marketvolatility 47

The selection of benchmark bondsand yield curves 47

Market expectations and thepredictability of monetary policy 47

The size and development ofcorporate bond markets/whichliabilities are replaced bycorporate bond growth? 47

The market microstructure andexchange rates 48

Determinants of the cost of capitaland economic growth 48

Financial indicators of underlyingeconomic variables 48

4 MA I N P R I O R I T I E S 5 0

33ECB


ANNEX A

The ECB-CFS research network on “CapitalMarkets and Financial Integration in Europe”aims at stimulating top-level and policy-relevant research, significantly contributing tothe understanding of the current and futurestructure and integration of the financialsystem in Europe and its international linkageswith the United States and Japan. The presentdocument and associated table establish a“roadmap” to help guide work under thenetwork. They (i) define the scope of thenetwork, (ii) identify a set of research areasand (iii) highlight concrete issues where newresearch looks particularly promising and/ornecessary. From the “roadmap”, five specificareas, listed below, are considered top priority.These areas will feature as special themes innetwork workshops, and papers on these topicswill receive special consideration forpresentation in those workshops.

Work under the network has to fulfil the usualacademic standard of being publishable in apeer reviewed journal, while:

(1) dealing with European financialintegration, with financial systemstructures in Europe or with financiallinkages between the euro area/EuropeanUnion, the United States and Japan;

(2) focusing on financial intermediaries,financial markets (including the relatedsettlement infrastructures) or therelationship between finance and the rest ofthe economy; and

(3) providing a description of/measuring one ofthe items mentioned under (1) above,explaining driving factors behind theseitems, discussing obstacles to integration orderiving policy implications on the basis ofefficiency or stability considerations.

While the “roadmap” is geared towards appliedand policy-relevant questions, both empiricaland theoretical research are important andwelcomed.

EX E CU T I V E S UMMARYThe three criteria referred to above define alarge number of relevant topics, which aresummarised in the table. The five top priorityareas selected are:

(1) bank competition and the geographicalscope of banking activities;

(2) international portfolio choices and assetmarket linkages between Europe, theUnited States and Japan;

(3) European bond markets;

(4) European securities settlement systems;and

(5) the emergence and evolution of newmarkets in Europe (in particular start-upfinancing markets).


The ECB-CFS research network on “CapitalMarkets and Financial Integration” aims atstimulating top-level and policy-relevantresearch, significantly contributing to theunderstanding of the current and futurestructure and integration of the financialsystem in Europe and its international linkageswith the United States and Japan.

The present document and the related table onpage 35 establish a “roadmap” to help guide thework under the network. It has been producedunder the auspices of the network SteeringCommittee. It particularly benefited from the“agenda setting” talks given at the launchingworkshop of the network, as well as fromcomments by workshop participants and ECBstaff.

More in detail, the purpose of this “roadmap” is

(i) to define the scope of the network,(ii) to identify a set of research areas and(iii) to highlight concrete issues where new

research looks particularly promising and/or necessary.

The document is structured as follows.Section 2 summarises the structure of the“roadmap”, thereby defining the scope of thenetwork. Section 3 provides brief descriptionsof the specific research topics from the“roadmap”. Finally, section 4 defines a smallnumber of priority areas, which will feature asspecial themes in the network workshops.

1 I N T RODUC T I ON 1

1 The views expressed in this document do not necessarilyrepresent the views of the European Central Bank (ECB), theEurosystem or the Center for Financial Studies (CFS). It hasbenef ited from comments by Franklin Allen, IgnazioAngeloni, Claudia Buch, Giorgio Calcagnini, Mark Carey,Vítor Gaspar, Reint Gropp, Hanna Hempell, Heinz Herrmann,Peter Hördahl, Cornelia Holthausen, Harry Huizinga,Stephane Kerjean, Stefanie Kleimeier, Jan-Pieter Krahnen,Philip Lane, Guillaume Leclercq, Angela Maddaloni, SimoneManganelli, Arnaud Mehl, Christina Metz, Cyril Monnet, PerNymand-Andersen, Steven Ongena, Charlotte Ostergaard,Marco Pagano, Chryssa Papathanassiou, Enrico Perotti,Harald Sander, João Santos, Martin Scheicher, SandrineScheller, Giancarlo Spagnolo, Mikael Stenstrom, EvangelosTabakis, Jens Tapking, Christian Upper and Jukka Vesala.

35ECB


ANNEX A

Tabl

e Su

mm

ary

of t

he “

road

map

”

Mai

n re

sear

ch a

reas

➝F

inan

cial

sys

tem

seg

men

ts

➝

Eu

rop

ean

fin

anci

al in

tegr

atio

n a

nd

fin

anci

al li

nk

ages

wit

h t

he

Un

ited

Sta

tes

and

Jap

anF

inan

cial

sys

tem

str

uct

ure

s in

Eu

rop

e

Financialintermediaries Financial markets

Ban

ks

Sec

uri

ties

firm

s

Insu

ran

ce c

omp

anie

s

•G

eogr

aphi

cal

scop

e of

ban

king

and

len

ding

rela

tion

ship

s•

The

opt

imal

deg

ree

and

impl

icat

ions

of

bank

com

peti

tion

•D

eter

min

ants

of

bank

mer

gers

(in

part

icul

ar c

ross

-bor

der)

•Im

plic

atio

ns o

f ba

nk c

onso

lida

tion

(e.g

. fo

r cr

edit

and

int

erba

nk m

arke

ts)

•In

tern

atio

nal

port

foli

o ch

oice

s of

ins

titu

tion

alin

vest

ors

(mut

ual

fund

s, p

ensi

on f

unds

,re

-ins

uran

ce c

ompa

nies

, et

c.)

•B

ank

vs.

mar

ket

fina

ncin

g•

Res

ilie

ncy

of t

he f

inan

cial

sys

tem

•R

isk-

shar

ing

wit

h ba

nk i

nter

med

iati

on•

Spe

cial

isat

ion

vs.

univ

ersa

l ba

nkin

g•

Eff

icie

ncy

of E

urop

ean

bank

s•

Ow

ners

hip

stru

ctur

es o

f fi

nanc

ial

inte

rmed

iari

es•

Inso

lven

cy r

egim

es•

Tra

nspa

renc

y an

d ac

coun

ting

rul

es•

Tec

hnol

ogy

and

fina

ncia

l se

rvic

es(i

nter

net

bank

ing,

fin

anci

al e

-com

mer

ce,

etc.

)

Cle

arin

g an

dse

ttle

men

tin

fras

tru

ctu

re

•A

sset

pri

ce a

ndvo

lati

lity

lin

kage

sac

ross

cou

ntri

es a

ndth

eir

chan

ges

•D

eter

min

ants

of

mar

ket

retu

rns

(cou

ntry

-sp

ecif

ic v

s. e

uro

area

wid

e, d

omes

tic

vs.

exte

rnal

)•

Con

tagi

on a

nd c

risi

sli

nkag

es•

Det

erm

inan

ts o

f in

ter-

nati

onal

por

tfol

io f

low

s•

Res

truc

turi

ng o

fse

ttle

men

tin

fras

truc

ture

s(f

ragm

enta

tion

and

pote

ntia

l re

form

of

the

secu

riti

es s

ettl

emen

tin

dust

ry,

dete

rmin

ants

and

role

of

cros

s-bo

rder

sett

lem

ent

cost

s)

•C

onso

lida

tion

of

stoc

kex

chan

ges

•F

orei

gn l

isti

ngs

and

the

role

of

depo

sito

ryre

ceip

ts•

Rel

ativ

e im

port

ance

of

sect

or a

nd c

ount

ryef

fect

s

•P

riva

te v

s. p

ubli

c eq

uity

•E

quit

y vs

. de

btfi

nanc

ing

•C

orpo

rate

gov

erna

nce

and

cont

rol

(for

ms

ofco

rpor

ate

cont

rol,

boa

rdtu

rn-o

ver,

sha

reho

lder

righ

ts)

•A

ccou

ntin

g co

nven

tion

san

d m

arke

t vo

lati

lity

Eq

uit

y

•S

elec

tion

of

benc

hmar

kbo

nds/

yiel

d cu

rves

•M

arke

t ex

pect

atio

nsan

d th

e pr

edic

tabi

lity

of m

onet

ary

poli

cy•

Siz

e an

d de

velo

pmen

t of

corp

orat

e bo

nd m

arke

ts•

Whi

ch l

iabi

liti

esre

plac

ed b

y co

rpor

ate

bond

gro

wth

?

•D

eter

min

ants

of

infl

atio

n, l

iqui

dity

and

defa

ult

risk

pre

mia

•S

ubst

itut

abil

ity

ofgo

vern

men

t bo

nds

from

diff

eren

t so

vere

ign

issu

ers

•In

tegr

atio

n of

rep

om

arke

ts

Fix

ed in

com

e(m

oney

an

db

ond

mar

ket

s)

•M

arke

t m

icro

stru

ctur

ean

d ex

chan

ge r

ates

•C

apit

al f

low

s an

dex

chan

ge r

ates

For

eign

exc

han

ge

•O

ptim

al s

ecur

itie

sse

ttle

men

t sy

stem

stru

ctur

es(e

.g.

conc

entr

atio

n)

•D

eter

min

ants

of

the

cost

of

capi

tal

and

econ

omic

grow

th•

Sta

rt-u

p fi

nanc

ing

•F

inan

cial

ind

icat

ors

of e

cono

mic

gro

wth

and

oth

erun

derl

ying

eco

nom

ic v

aria

bles

•D

eter

min

ants

of

coun

try

inve

stm

ent

port

foli

os(i

nclu

ding

hou

seho

lds)

and

the

hom

e bi

asph

enom

enon

•G

oods

mar

ket

and

fina

ncia

l in

tegr

atio

n•

Glo

bali

sati

on a

nd t

he i

ncre

asin

g ro

le o

f ca

pita

l fl

ows

•S

ynch

roni

sati

on o

f in

tern

atio

nal

busi

ness

cyc

les

•Im

port

ance

and

dete

rmin

ants

of

the

brea

dth

and

com

plet

enes

s of

mar

kets

/ f

inan

cial

inno

vati

on(d

evel

opm

ent

of r

epo

and

deri

vati

ve m

arke

ts,

brea

dth

of r

isk

prof

iles

,em

erge

nce

ofco

mm

erci

al p

aper

,in

dex-

link

ed b

onds

and

mor

tgag

e-ba

cked

secu

riti

es m

arke

ts e

tc.)

•R

ole

of l

egal

sys

tem

s•

Liq

uidi

ty i

n se

cond

ary

mar

kets

•M

arke

t m

icro

-st

ruct

ures

•R

ole

of e

lect

roni

cm

arke

t tr

adin

gpl

atfo

rms

•O

rgan

isat

ion

ofpr

imar

y m

arke

ts

•D

escr

ipti

on a

nd

mea

sure

men

t

of:

–pr

ice

link

ages

,–

the

geog

raph

ical

sco

pe o

f le

ndin

g an

din

vest

men

t, a

nd–

the

rela

tive

im

port

ance

of

fund

ing

sour

ces.

Par

ticu

lar

impo

rtan

ce i

s gi

ven

to d

evel

opin

gne

w m

easu

res,

usi

ng m

icro

data

to

addr

ess

aggr

egat

e ph

enom

ena

and

unde

rtak

ing

com

pari

sons

acr

oss

EU

cou

ntri

es,

the

Uni

ted

Sta

tes

and

Japa

n

•D

rivi

ng

fact

ors

are:

–co

mpe

titi

on,

–co

-ope

rati

on a

nd–

publ

ic a

ctio

nas

wel

l as

int

erac

tion

s be

twee

n th

em.

(Wh

at is

th

e ro

le o

f th

e eu

ro?)

•O

bst

acle

s

are,

e.g

.:–

diff

eren

t le

gal

and

tax

syst

ems,

–he

tero

gene

ous

regu

lato

ry a

ppro

ache

s,–

inco

mpa

tibl

e in

dust

ry s

tand

ards

(e.g

. ac

coun

ting

and

dis

clos

ure)

and

–cu

ltur

al a

nd l

angu

age

barr

iers

.

•P

olic

y im

pli

cati

ons

Eff

ects

on

effi

cien

cy,

fina

ncia

l st

abil

ity,

grow

th a

nd w

elfa

re.

Fin

ance

an

dth

e E

con

omy


The “roadmap” is summarised in the table onpage 35, which has the following structure. Ithas three broad dimensions: (i) the columnsdescribe the main broad research areas; (ii) therows describe the main segments of a financialsystem; (iii) the large and shaded inner boxdescribes the main broad approaches andfactors that apply to all dimensions defined byresearch areas (i) and (ii) above. Finally, thecells of the table contain a selection of bulletpoints, which highlight concrete topics orissues that are of particular importance. Next,the three broad dimensions will be described ingreater detail.

There are three distinct, but not unrelated, mainbroad research areas: (i) European financialintegration, (ii) financial system structures inEurope and (iii) financial linkages between theeuro area/European Union (EU), the UnitedStates and Japan. For presentational reasons,the first and the third areas have been puttogether in the first column of the table, whilethe second area has been put in the secondcolumn. The first column describes the area ofresearch on financial integration and financiallinkages between different geographicalentities. In other words, this dimension of thework of the network deals with cross-countryor cross-regional linkages within the euro area/EU and with linkages between the euro area/EU, the United States and Japan. Financiallinkages are defined by relationships betweenprices and by relationships between quantitiesacross the relevant geographical entities,mainly countries. The second column describesthe area that deals with the structure of the euroarea/EU financial system as a whole in terms ofthe importance of the different ways offinancial intermediation and the differentsources of funding. Depending on the waysof intermediation and the sources of funding,different types of financial systems can bedistinguished. Given that the financial systemsin different euro area/EU countries are not fullyhomogeneous, this broad area also includescomparative work on the structure of differentfinancial systems in various countries.

The main segments of a financial system aretraditionally differentiated by referring tofinancial intermediaries, on the one hand, andfinancial markets, on the other. Apart fromfocusing on financial integration, linkages andstructure with respect to institutions andmarkets, the network also intends to includework on the interaction between the financialsystem (in the narrow sense) and the rest of theeconomy. Among the financial intermediaries,banks should receive special attention, but thenetwork also intends to spur more work onsecurities firms (such as mutual funds) andinsurance/re-insurance companies. In terms offinancial markets, the network covers fixedincome (bond and money markets), equity andforeign exchange markets, as well as therelated securities clearing and settlementinfrastructures. As to the relation between thefinancial system and the rest of the economy,there is a special interest in the implications offeatures of the financial system for economicactivity and the importance of goods marketintegration for financial integration, and viceversa. However, work on the non-financialsectors of the economy only, without a primaryrole for financial variables, would not beconsidered.

There are a number of factors and approaches,which are at the heart of the network and applyto all sub-areas defined by the matrix ofdimensions of research areas and financialsystem segments. They are summarised in thelarge shaded box in the centre of the table.

First, there is interest in the exact measurementof the degree of financial integration and inprecise descriptions of the type(s) of financialsystem structures observed. For example,financial integration may be measured on thebasis of the correlation or dispersion patternsof asset prices or by the regional andinternational scope of lending, financing andinvestment activities. Financial systemstructures may be described by the quantitativedocumentation of the relative importance ofdifferent sources of funding.

2 TH E S T RU C TUR E O F T H E “ ROADMAP ”

37ECB


ANNEX A

Second, there is interest in the identification ofthe factors that drive the process of financialintegration and the design of financial systemstructures. These driving factors includecompetitive market forces, co-operativeinitiatives between market participants andpolicy action by public authorities. Forexample, how do market forces shape thedegree of financial integration and how canpolicies be effective in healing marketimperfections? Needless to say that the roleStage Three of Economic and Monetary Union(EMU) and the euro play as a force in this gameis of special interest to the present initiative.

Directly related is the interest in theidentification of obstacles to the integrationprocess or to the emergence of the bestfinancial system structure. Such obstacles areoften historically grown elements of financialsectors that cannot be easily removed, e.g. forpolitical economy reasons. They include legaldifferences between countries, regulatoryapproaches, country-specific industrystandards (e.g. market conventions or ITstandards) or corporate governance systems. Itis important to understand how and to whatextent these obstacles hamper the process ofEuropean financial integration and theemergence of an efficient financial systemstructure.

The shaded box closes with issues that add amore normative perspective and, ultimately,the derivation of policy implications. Acrossall fields, the network is interested in theassessment of the efficiency and the stability ofthe structures observed. This should thenestablish the link to growth and generalwelfare. Consequently, the research conductedin the context of the network should try todeduce policy conclusions from its findings, inparticular in the areas of greatest relevance.

In summary, research papers that addressthe areas defined by the “roadmap” table, asdescribed above, fit into the work programmeset for the network. In other words, a paper canbe considered under the network, if it finds a

match in the table regarding all three broaddimensions discussed above. This means that itshould deal with at least one of the main broadresearch areas, involve one (or more) of themain financial system segments and addressone (or more) of the factors and approachesmentioned in the shaded box of the table.Papers that directly address one of the bulletpoints enumerated in the non-shaded parts ofthe table will receive special consideration forfuture network workshops and conferences.The next section describes some of thesespecial issues in greater detail, so as to betterexplain what is behind those issues. Finally,while the “roadmap” is geared towards appliedand policy-relevant questions, both empiricaland theoretical research are important andwelcomed.


The “roadmap” table described above containsa selection of bullet points in the differentcells. These bullet points each refer to an issueor a research topic of special importance to thenetwork. To clarify what is meant under thesebrief entries, a list of short descriptionsfollows.

3.1 EUROPEAN FINANCIAL INTEGRATION ANDFINANCIAL LINKAGES WITH THE UNITEDSTATES AND JAPAN

THE GEOGRAPHICAL SCOPE OF BANKING ANDLENDING RELATIONSHIPS

Banking business has become moreinternational over the past few decades.However, not all banking products are tradedglobally. Many credit and deposit marketsregional in character, whereas securities andforeign exchange trading is global. It isimportant (for the institutional design ofsupervisory structures, for instance, or forcompetition policies) to understand theregional and international scope of lending anddeposit taking. What is the “relevant market”for a specific banking service? Does it cutacross borders or not? Are banks located indifferent regions or countries competing witheach other? What are the implications formarket integration? Are long-term lendingrelationships geared to home banks or hostbanks, e.g. do they depend on the country oforigin? What are the lending and borrowinglinkages between banks? What are theimplications for contagion risk? Empiricalapplications with a strong euro area/EUcomponent are highly desirable.

THE OPTIMAL DEGREE AND IMPLICATIONS OFBANK COMPETITION

In the past, too much bank competition wasoften regarded as undesirable, or evendangerous. It was argued, for example, thatbanks not earning monopolistic profits wouldtake excessive risks and therefore endanger thestability of the financial system. Similarly, in

3 S P E C I F I C I S S U E S O F T H E N E TWORKthe past, many countries had some controls onthe credit process or on capital flows. Financialliberalisation, lighter regulatory approachesand also globalisation have recently increasedthe role of competitive forces in the bankingsector, and reductions in governmentownership and government guarantees maycontinue this process in the future. It isimportant to understand what the implicationsof this greater competition in the bankingsector are, in particular regarding risk taking bybanks and the availability of credit to firms, butalso in terms of the provision of financialservices more generally. A precondition forthis is a better understanding of the process ofcompetition in the banking sector as such, e.g.from the perspective of industrial organisationand information economics. This will have totake account of the differences between thevarious banking services. In the Europeancontext, the relevance (or not) of cross-bordercompetition in the provision of bankingservices seems to be an important issue, withthe situation potentially differing in wholesaleand retail business. The results will allowpolicy-makers to better assess whether furtherliberalisation and deregulation are advisableand, if so, in which areas, or whether thisprocess has already gone too far. Empiricalapplications to the euro area/EU areparticularly important. Possible additionalquestions include: What are the implications ofincreased integration for risks in banking, andthus for banking supervision? How shouldsupervision be designed to allow for faircompetition and further banking marketintegration within the EU?

DETERMINANTS OF BANK MERGERS

The past decade has seen a wave of mergers andacquisitions (M&As) in the banking industry inEurope and worldwide. Most of these M&Asinvolved banks established in the samecountry. What are the reasons for this “homebias” in financial consolidation? Moregenerally, research needs to clarify better whatmotivated these bank mergers and whichpolicies are likely to increase or reduce merger

39ECB


ANNEX A

activity in the banking sector. Whatcharacteristics are shared by the few banks thatare successful in entering foreign markets? Ofspecial importance are theoretical andempirical analyses explaining why there havethus far been only a smaller number of cross-border bank mergers in the euro area/EU andwhether this is likely to change.

IMPLICATIONS OF BANK CONSOLIDATION

Bank consolidation tends to increaseconcentration in lending and deposit markets,but it may also foster integration. Would morecross-border consolidation be desirable for thefinancial landscape, in particular bankingintegration, in the euro area/EU? Would cross-border consolidation be better for regionalcompetition than national consolidation? If theanswers to both these questions are yes, whichpolicies would facilitate cross-borderconsolidation in European banking sectors? Dothe current forms of consolidation have adverseeffects on competition in loan and depositmarkets in the euro area/EU? Do potential costor profit efficiency advantages offset increasesin market power? Given that there are mainlynational bank mergers, special emphasisshould be placed on the effects of consolidationon regional and national loan and depositmarkets, e.g. small business finance. Howimportant is bank consolidation for credit anddeposit market integration? How doesconsolidation affect the liquidity insuranceprovided by interbank markets and theliquidity-providing function of banks? Is therea risk that, beyond a certain level, they maybecome less liquid and less able to smooth outliquidity shocks to the banking system? Whatdo these effects imply for the liquiditymanagement of the central bank? Finally, howdoes the consolidation process affect assetdiversification, the risk-taking behaviour ofbanks and the risk of instability in the bankingsystem as a whole?

INTERNATIONAL PORTFOLIO CHOICES OFINSTITUTIONAL INVESTORS/DETERMINANTS OFINTERNATIONAL PORTFOLIO FLOWS

The largest asset portfolios tend to be managedby institutional investors. Their behaviour thushas a significant impact on international capitalflows and global asset price fluctuations.Despite the high mobility of capital in worldmarkets, the home bias in portfolios remainspersistent. It is therefore important to betterunderstand the investment decisions ofinstitutional investors. Particularly valuablewould be portfolio choice analyses on the basisof disaggregated, firm-level data, not onlywithin the euro area, but also globally (inparticular regarding “G-3” capital flows). Thiscould make it possible, for instance, to linkinstitutional investors’ choices to householdportfolios. Is the growth of the mutual andprivate pension fund business continuing?Does increasingly professional management ofportfolios reduce home bias in euro areacountries and increase financial integration?What have been the implications of the removalof currency-matching and other regulationsfrom investment portfolios? Has theimportance of cross-sector asset allocationstrategies relative to cross-country strategiesincreased through these developments, therebyreducing country effects in returns? What arethe market imperfections that lead to biases inportfolio choices? How important are differentgroups of institutional investors in countries’portfolios (mutual funds, pension funds, hedgefunds, re-insurers, etc.)? What are theimplications for asset price volatility andexchange rates?

ASSET PRICE AND VOLATILITY LINKAGESACROSS COUNTRIES AND THEIR CHANGES/CONTAGION AND CRISIS LINKAGES

There is an increasing number of researchpapers estimating asset market linkages acrossdifferent countries, both in “normal” times andin crisis situations. This literature is importantfor the network, but it needs to be extended anddeepened in various directions. First, cross-


asset linkages (e.g. between stocks, bonds andmoney market instruments) should receivegreater attention. Moreover, it is increasinglyimportant to explicitly identify the factors thatdetermine the strength of return and volatilitylinkages and their changes over time.Methodological approaches that explicitlymodel gradually shifting or regime-switchingcorrelations or other (non-linear) measures ofco-dependence are particularly interesting foradvances in those directions. For crisislinkages, further developments of multivariateextreme-value analysis are important, e.g.when combined with regression-type analysisadding explanatory variables. More conceptualor even theoretical work needs to better clarifythe relationship between price/returncorrelation and market integration. Othermeasures seem to be needed. The recentapplication of concepts such as β- and σ-convergence (borrowed from growth theory)seem (at least in part) to point in this direction(β-convergence refers to the speed ofconvergence of prices of the same asset indifferent locations, and σ-convergence to theevolution of asset price dispersion acrosslocations over time).

DETERMINANTS OF MARKET RETURNS/DETERMINANTS OF INFLATION, LIQUIDITY ANDDEFAULT RISK PREMIA

EMU implied a change from a more nationallyoriented financial market behaviour to morearea-wide approaches. Therefore, analyses ofthe factors driving financial market returns inthe euro area/EU should reassess their mainsources. Which fundamental variables drivestock returns and bond yields? Are country-specific data more important than area-widestatistical releases, or vice versa? As Europeanstock markets become increasingly integrated,thus affecting the pricing of the relevant riskfactors, how will this impact on the cost ofcapital for firms? Does Europe need US-typedata release calendars with “lock-upconditions”, etc.? Do euro area returns becomemore independent of external factors outsidethe euro area? What explains any change

identified? How important is the introductionof the euro for these changes? Recent researchhas found a diminishing role of country effectsin global equity returns and an increasing roleof sector effects. Is this a cyclical phenomenon,potentially related to a contemporaneous boomin the IT sector, or rather a persistent one? HasEMU played a role for European developmentsthat go beyond the global trend? Whicheconomic explanations can be found for it?Another task is the separation of different riskpremia for bond returns, in particular regardingpotential liquidity risk premia in contrast toinflation and default risk premia. Are theliquidity levels for different government bondissues in the euro area uniform and, if not, why?Moreover, better models of default premia incorporate bond yields are needed.

LIQUIDITY IN SECONDARY MARKETS/SUBSTITUTABILITY OF GOVERNMENT BONDSFROM DIFFERENT SOVEREIGN ISSUERS

EMU has basically unified the unsecured euromoney market and acts as a catalyst for furtherconsolidation in other euro area financialmarkets. This raises the issue of how liquidityevolves in money, bond, equity and derivativesmarkets. For example, financial marketmicrostructure analysis can be used to assessvarious dimensions of liquidity, such as thetrading volume, transaction costs and the priceimpact of trades and order flows. A relatedquestion is the implications of liquiditychanges or differentials for market returns. Towhat extent are differences in liquidity orchanges in liquidity priced as risk factors inbond markets? Are there episodes whereliquidity dries up in the secondary markets?When and why does this happen? How severeare the consequences for market efficiency andfinancial stability? Moreover, are theredifferent “liquidity pools” of assets within thesame euro area asset class (such as governmentbonds) and, if so, why? For example, despiterelatively small default risk differentials,government bonds from various euro areacountries do not seem to be fully substitutable.How is the degree of substitutability of

41ECB


ANNEX A

government bonds from different sovereignissuers affected by the existence and size ofderivatives markets tied to specific nationalbonds? Particularly new areas are the liquidityof secondary corporate bond and repo markets.Liquidity issues have recently also been raisedfor foreign exchange markets, whereconsolidation in the corporate and financialsectors as well as the introduction of the eurohave eliminated a portion of the tradingvolume.

RESTRUCTURING OF SETTLEMENTINFRASTRUCTURES

A barrier to further integration of euro areafinancial markets is the current shape of thesecurities clearing and settlementinfrastructure in Europe. Despite rapid recentchanges, including a large number ofhorizontal and vertical mergers, the highfragmentation of securities settlement systemsleads to high cross-border settlement costs,hampering cross-border securities tradingwithin the euro area/EU. Moreover, banksincreasingly compete with traditionalinstitutions of the industry by offering in-housetrading, clearing and settlement services, whilecentral counterparty clearing is offered byclearing houses. In light of this, a number ofimportant questions arise. A very importantone concerns the optimal securities settlementsystem structure in the euro area, in particularregarding the right degree of centralisation andthe right degree of horizontal and verticalintegration between firms and exchanges.Another is the political economy reasons thatmake the reform of these structures difficultand, linked thereto, whether any policy orregulatory interventions should be carried outto facilitate reform. Third, how do the differentsecurities settlement channels (internationalcentral securities depositories, national centralsecurities depositories, global custodians)compete and co-operate with each other? Here,recent theoretical work on competition withinand between networks has defined a startingpoint. Linked to this question, an interestingissue concerns the impact of the “insourcing”

of securities services by banks. Finally, there isvery little published research addressing thecost and pricing of securities settlementservices. This whole field of the network is ascomplex as it is important. Since there is hardlyany research literature available, work in thisfield could be extremely topical and important.

CONSOLIDATION OF STOCK EXCHANGES

The introduction of the euro has brought to thefore the large number of stock markets inEurope. Meanwhile, various markets have triedto merge or even staged take-over attempts. Anaccurate cost-benefit analysis of the effects ofgreater stock market consolidation is needed.This implies assessments of economies of scaleand the effects of liquidity pooling, as well aspotential political economy reasons forobstacles to greater consolidation. Is there an“optimal structure” of stock markets in Europe,and how can it be reached? Is this optimalstructure identical for all stock marketsegments? Should the consolidation of listingand trading be linked to the consolidation ofsettlement procedures, and vice versa? Areelectronic trading links and joint venturessubstitutes for or complements toconsolidation?

FOREIGN STOCK LISTINGS AND THE ROLE OFDEPOSITORY RECEIPTS

Another step towards more financialintegration is the increased listing of foreignstocks in a country’s equity market. Somemarkets seem to be more successful inattracting foreign listings than others. Whatdrives the decision to list abroad, and whatdrives the choice of where to list? What are theadvantages and disadvantages of doing so,compared with a common market? An evenmore direct way of integrating stock markets isthe issuance of depository receipts, such as theincreasingly important American DepositoryReceipts (ADRs) used by US investors at hometo trade stocks listed abroad. How advanced isthe use of depository receipts in Europe? Howdo they change the competition between


different exchanges? How do they compare toother forms of integration? How important areforeign listings/depository receipts as a vehicleof cross-border capital flows? Do firms whichenjoy access to foreign exchanges face lowercosts of equity capital? Do they face differentrelative costs of debt and equity capital? Doestheir financial structure therefore change withaccess to foreign exchanges? Does equitycapital raised abroad replace or complementdomestic equity sources? Hence, to what extentdoes a foreign listing/depository receipt issueimply competition for domestic suppliers/intermediaries of equity capital and/ordomestic suppliers/intermediaries of debtcapital?

INTEGRATION OF REPO MARKETS/SUBSTITUTABILITY OF GOVERNMENT BONDSFROM DIFFERENT SOVEREIGN ISSUERS

Whereas the unsecured euro money marketsintegrated very fast with the introduction of theeuro, the same did not happen with the repomarkets. Despite some recent growth in thissegment, various obstacles to its fullintegration remain, such as the fragmentationof securities settlement systems, the parallelexistence of different master agreements andthe imperfect substitutability of governmentdebt within the euro area. Nevertheless, withfew exceptions, this important market forcentral banks has received very little attentionin academic literature. Therefore, theoreticaland empirical papers on the role of repomarkets and on how to create a unified repomarket in the euro area are important.Moreover, do the impediments mentionedabove lead to significant arbitrageopportunities? How does the microstructure ofthe repo market work? And finally, what is theimportance of the further development andrelative size of the repo market, in comparisonwith the unsecured money market, for financialsystem stability.

CAPITAL FLOWS AND EXCHANGE RATES

The increasing size of international capitalflows has drawn the attention to their role indetermining exchange rates. This themefigured highly in the discussions on the earlydepreciation of the euro. The networkencourages further work on the influencechanges in stocks and flows of internationalequity, bond and money market investment andfinancing have on “G-3” exchange rates, inparticular on that of the euro. When doing so,however, it is important that a number ofmacroeconomic constraints are explicitlyaccounted for. Obviously, a pre-condition forsuch exchange rate research is a thoroughunderstanding of what moves capital betweenthe euro area, the United States and Japan in thefirst place. Particularly important isunderstanding how market participants formexpectations about the productivity and growthpotentials of the “G-3” economies, and aboutrelative fiscal and monetary policies.

DETERMINANTS OF COUNTRY INVESTMENTPORTFOLIOS AND THE HOME BIASPHENOMENON

This heading would include research on theobservation that individuals hold too little oftheir wealth in foreign assets, and therefore donot optimally hedge risks across countries.Related to this, macroeconomic literature hasidentified a “consumption home bias”phenomenon, meaning that output risks are notoptimally shared across countries, so thatdomestic consumption is correlated withcountry-specific shocks to domestic output. Aninteresting question is whether, from aEuropean perspective, these two forms of homebias may be related. How do the portfoliochoices of different groups of investors differand what does this imply for the causes of homebiases? Another topic that could be coveredunder this heading concerns the role of realestate in households’ wealth. This is a topic ofsome significance with respect tounderstanding investment decisions in theeconomy, given that real estate is one of the

43ECB


ANNEX A

most important ways of saving for manyindividuals. Finally, an important area is“home bias at home”, analysing to which extentbiases in international portfolio changes areany different from biases in regional portfoliochoices within countries. This issue could beparticularly relevant for assessing whetherEMU will significantly reduce portfolio biasesin the euro area.

GLOBALISATION AND THE INCREASING ROLEOF CAPITAL FLOWS/SYNCHRONISATION OFINTERNATIONAL BUSINESS CYCLES

Previous research suggests that the varioustypes of capital flows differ in their impact onthe economy in terms of their volatility, forinstance, or in terms of whether they tend tocrowd out or complement domestic savings.More research on the links between thestructure of cross-border capital flows andeconomic activity (growth and volatility) isneeded. To this end, it is important to betterunderstand the link between financial flowsand the synchronisation of internationalbusiness cycles, as well as the determinants ofthe structure of capital flows. One determinantthat has not received sufficient attention so faris the domestic institutional environment,including corporate governance rules. On theone hand, it has been argued that domesticdemand for bank credit, debt and equity capitalis determined by the characteristics ofdomestic industries and firms and by thedomestic institutional environment governingagency costs in financing decisions. Thestructure of foreign capital inflows would thenbe expected to be similar to domestic financialstructures. However, firms might also opt forforeign sources of funding with the aim oftaking advantage of corporate governance rulesthat are better at reducing the (agency) costs ofcertain types of financing. In that case, thestructure of cross-border capital flows wouldbe expected to differ markedly from domesticfinancial structures.

GOODS MARKET AND FINANCIAL INTEGRATION

This heading would include research on theinteraction of financial integration and realintegration, including the direction ofcausality. It would also cover research on therole of financial integration for business cycletransmission and, more generally, on the linkbetween financial integration and economicgrowth. In addition to other channels, could anincreased integration of European capitalmarkets have a positive impact on economicgrowth by increasing the attractiveness ofEuropean markets for companies and investors,including foreign ones? Moreover, this areacould include measurement issues of price-based and quantity-based financial and goodsmarket integration indicators.

3.2 FINANCIAL SYSTEM STRUCTURES INEUROPE

BANK VERSUS MARKET FINANCING

In principle, bank and market financing arecomplementary elements of any financialsystem. Historically, however, the Europeanand Japanese financial systems have beenbased on bank lending. The US financialsystem, by contrast, has developed financialmarkets much more rapidly and extensively. Itis of interest to the network to understand thecurrent state of bank and market financing inthe euro area/EU. Will the European financialsystem develop over time into a structure verysimilar to the one observed in the United States,or will substantial differences remain? Whatare the factors that drive the relativeimportance of bank and market financing (legalsystems, political economy factors, etc.)? Arethere differences between more bank-basedfinancial systems in terms of the efficiency ofthe resource allocation, the cost of capital andeconomic growth or in terms of financialstability? Bank-based financial systems tend tobe less well understood than market-basedfinancial systems in terms of how bankintermediation contributes to risk sharing for


instance. Is there any substantial difference inthe riskiness of the average household’s assetholdings in bank-based and market-basedsystems? Is it true that markets better sharecross-sectional risk, while banks provideintertemporal smoothing? One differencebetween a further development towards a moremarket-based financial system in Europe andthe situation in the United States could be thegreater involvement of universal banks inEurope, as opposed to specialist financialintermediaries.

THE RESILIENCY OF THE FINANCIAL SYSTEM

Experience with financial instabilities duringthe past few years emphasises the importanceof the ability of financial systems to absorbadverse shocks. This capacity of a financialsystem is particularly important in the contextof a rapidly changing financial landscape, suchas that in Europe. In the context of Europeanfinancial markets, promising areas of researchfor the network would address the followingquestions. What is the relationship between thetype of financial system (e.g. bank versusmarket-based) and the ability of the financialsystem to withstand shocks? How can thechoice of accounting rules and standardsincrease transparency, accountability and theability of marking to market, and therebystrengthen the capacity of the financial systemto absorb shocks? How is the role of Europeanbanks as providers of liquidity insurance to thecorporate sector evolving? What is the linkbetween financial fragility and contagion?How important are contagion and financialfragility in practice, and what mechanismstrigger them? How and to what extent dofinancial crises impact on economic activity?Finally, does competition in the financialservices industry affect the ability of thefinancial system to withstand shocks?

THE EFFICIENCY OF EUROPEAN BANKS

For a long time, Europe has been characterisedby a situation of “over-banking”, particularlyexcessive numbers of branches. Consolidation

has solved this problem only partly. Inparticular, previous empirical research hasshown that many larger mergers are often notefficiency-enhancing. Better knowledge ofwhich mergers are efficiency-enhancing inEurope, and which are not, would be important.More generally, empirical studies on theevolution of cost and profit efficiency in theEuropean banking sector are needed. Are theredifferences between countries? What explainsthese differences? What is the role of cross-border competition in potential efficiencyimprovements? Can efficiency also be assessedon a product-by-product basis, instead of onlyat the level of the overall firm? There is also aninterest in investigating the sources andimplications of structural differences in thebanking systems, such as differences in bankmarket concentration, the number of bank–firmrelationships, branching, ownershipconcentration and technology.

INSOLVENCY REGIMES

A special complication in the internationalfinancial landscape is the heterogeneity ofinsolvency regulations, also for financialintermediaries. The Directive on the winding-up of credit institutions has harmonised certainconflict-of-law issues in cross-borderinsolvencies of banks with several branches inthe EU. However, the treatment of bankruptaffiliates in a group of companies remains anissue. Furthermore, conflicts can arise betweeninsolvency regimes for banks located in andoutside the EU. The countries following the“single-entity” approach (e.g. the UnitedKingdom) pool all assets of a bankrupt bankand its branches for the satisfaction of allcreditors. Other countries (such as the UnitedStates) follow the “separate-entity” approachthat ring-fences local assets of a branch for thesatisfaction of the local creditors/depositors.When these two approaches meet, there is noanswer how to solve the conflict. An analysis ofthese problems could provide evidence for thenext generation of EU regulations andsolutions for EU authorities in the case offinancial crises. Contract and market-based

45ECB


ANNEX A

bankruptcies have gained momentum and theycould be a solution to ease tension created bydifferences in legal regimes. An importantissue is whether the current situation createsuncertainty for foreign equity and bondinvestors as to what the priority of their claimswould be if the respective company were to gobankrupt. Or, do differences in insolvency lawsimply distortions in international investments?

TECHNOLOGY AND FINANCIAL SERVICES

The development of new technologies isdrastically transforming the global financiallandscape. What impact will the internet haveon the provision of financial services inEurope? Will it increase bank competition andreduce over-branching? Under whichconditions will it do so? Search costs need to bemodelled and the effects of obfuscation-friction strategies. In addition, the effects oftransparency on producers and consumers areof relevance for assessing the impact of theinternet on markets for financial services. Howfast is internet banking evolving? Whichproducts are more affected by it, and which less(deposit-taking and retail payments, lending,securities trading, etc.)? Is the expansion offinancial e-commerce a concern in terms offinancial stability or consumer protection?

THE IMPORTANCE AND DETERMINANTS OF THEBREADTH AND COMPLETENESS OF FINANCIALMARKETS/FINANCIAL INNOVATION

Another essential element of the network is thenumber of financial products available in afinancial system. Which markets are missing orstill relatively illiquid in the euro area/EU?What is the status of these markets and what arethe conditions under which they would emergeand develop further? How important arespecific markets for risk sharing and financialdevelopment? Studies on the followingmarkets in Europe seem particularly important:(i) the European corporate bond market as awhole and the breadth of the risk profiles in it,(ii) the market for asset-backed securities,(iii) commercial paper markets, (iv) credit and

other derivatives markets and their links tocash markets, (v) index-linked bond markets,(vi) private equity and start-up markets, and(vii) markets for repurchase agreements(repos). In this research, the euro area-wideperspective (or at least some cross-countrydimension thereof) is particularly important.

ROLE OF LEGAL SYSTEMS

Recent literature has argued that the type oflegal system is a crucial long-term determinantof the type of financial system in an economy.What is the most adequate legal environmentfor well functioning financial markets? Isthere a need to reconcile common law andcivil law approaches? How important is theharmonisation of legal systems for thedevelopment of the euro area financial system?Is there a need to go beyond the legislativeinitiatives launched under the EU FinancialServices Action Plan? Is there a need for legalharmonisation to foster integration in financialmarkets? What is the impact of the legalsystems of European countries on thedevelopment of specific markets? To whichextent do laws influence cross-border capitalflows? Conversely, to what extent does theelimination of obstacles to cross-border capitalflows create pressures for legal reform? Towhat extent does it create room for “legalarbitrage” and what would be the implications(race to the bottom or to the top)? What is therole/importance of the legal system ascompared to the influence from politicaleconomy factors?

MARKET MICROSTRUCTURES/THE ROLE OFELECTRONIC TRADING PLATFORMS

The literature on the financial marketmicrostructure highlights the importance ofmarket design features in determining theefficiency and liquidity of a given market.More recently, microstructure-type analyseshave even been used to address ratheraggregate phenomena, such as exchange ratedevelopments, international currency uses,monetary policy implementation, etc.


Therefore, this approach makes it possible toanalyse important general issues about the stateof European financial markets in depth,including market liquidity, informationalefficiency and the effects of news (pricediscovery), transaction costs, tradingpractices, high-frequency arbitragerelationships and integration. In Europe, inparticular, electronic trading has gained anever greater market share in secondarymarkets, especially in the stock and bondmarkets, but to some extent also in the moneymarkets. This allows the functioning ofmarkets to be studied with increasing“granularity” and precision. The design andscope of electronic trading platforms is ofinterest itself, since it to a large extentdetermines the aspects enumerated above. Forexample, is it advisable also to give retailcustomers access to electronic platforms (opensystems), or should they be reserved forwholesale interbank trading? What is theoptimal degree of transparency in thosesystems? What is the role of inter-dealertrading? How important is it that the samesystem spans different countries?

THE ORGANISATION OF PRIMARY MARKETS/EQUITY VERSUS DEBT FINANCING/PRIVATEVERSUS PUBLIC EQUITY MARKETS/START-UPFINANCING

The way primary bond and stock markets areorganised has an important impact on the costof debt and equity financing. Can importantchanges be identified in the organisation ofbond underwriting business and initial publicofferings (IPOs)? What are the implications forthe relative cost of debt and equity financing?In the euro-denominated international bondmarkets, for example, a significant entry offoreign book-runners could be observed withthe introduction of the euro and consequentlyalso a substantial reduction of underwritingcosts. By eliminating exchange rate risk, acommon currency may also modify companies’choices between debt and equity, as has for thefirst time been argued in a recent paper. Verylittle is known about how private equity

markets work. How important are privateequity markets in comparison with regular andstart-up public equity markets? Whatdetermines a company’s choice betweenprivate equity and an IPO in the start-upsegment of the public stock market? What arethe important elements of venture capital andprivate equity financing? What is the best wayto provide for the financing of new firms,avoiding both under- and over-provision offunds? How is European financing ofinnovative start-up businesses developing?Which policy measures could support efficientstart-up financing in Europe? Can any lessonsbe drawn from the US experience? Whatexplains the large differences betweencountries? Which financial intermediaries andindividual investors fund venture capitalfirms? Further analysis is also needed of newEuropean public equity market segments, suchas the Neue Markt and the Nouveau Marché, inparticular a comparison with those in theUnited States (notably the Nasdaq).

CORPORATE GOVERNANCE AND CONTROL/OWNERSHIP STRUCTURES OF FINANCIALINTERMEDIARIES

Governance of industrial firms tends to bediverse in Europe. For example, generalshareholder rights, take-over regulations andobserved board turnover can vary significantlyacross countries. Activity in the market forcorporate control is not uniform either. Majordifferences exist between the United Kingdom(or the United States) and continental Europeancountries. What is the right balance betweenharmonisation and competition betweensystems? Is one system superior to the other interms of performance? What form does systemcompetition take (e.g. foreign listings versusrelocation of headquarters versus voluntaryfirm-level commitments in response topressure from foreign shareholders)? What arethe implications of different forms ofcompetition (e.g. does a migration ofheadquarters entail significant externaleffects)? To what extent is there a danger thateither harmonisation or competition might

47ECB


ANNEX A

destroy the internal consistency of a corporategovernance system and lead to a worsening ofperformance? There are also pronounceddifferences in corporate structures. Is onebetter than others in terms of innovations, forinstance? Are differences in corporategovernance more pronounced across nationalborders or across different types of firms (e.g.firms with and without access to internationalcapital markets, small firms versus largefirms)? Are different rules required fordifferent activities/types of firms? Should/willharmonisation and competition play differentroles in changing the rules for different types offirms (e.g. rely on competition forinternationally active firms, rely onharmonisation for the others)? Finally, studiesof the ownership structures of financialintermediaries are important, such as theimplications of different degrees ofconcentration in ownership.

TRANSPARENCY AND ACCOUNTING RULES/ACCOUNTING CONVENTIONS AND MARKETVOLATILITY

The international harmonisation of accountingstandards has proven to be a difficultundertaking. The importance of transparencyand accounting practices have recently beenillustrated by the Enron failure and othercorporate scandals. A largely unexplored areais the extent to which differences in accountingpractices can explain differences in stockmarket volatility across countries. Is moretransparency always better? What are theimplications of more transparency forefficiency and financial stability.

THE SELECTION OF BENCHMARK BONDS ANDYIELD CURVES

Both fiscal consolidation in the United Statesand EMU have reduced the role of thegovernment bond yield curve as a benchmarkon which to base the pricing of derivatives, forinstance. In Europe, the swap curve nowbasically plays the benchmark role fornon-government bonds. How did that happen

and what determines the choice of benchmarkcurve more generally (“benchmark-tippingphenomenon”)? How does the choice ofbenchmark affect the pricing in the market (if atall)?

MARKET EXPECTATIONS AND THEPREDICTABILITY OF MONETARY POLICY

Bond and money markets offer manyinstruments to assess market expectationsabout monetary policy. For example, potentialexpectation errors on interest rate changes canbe derived. Moreover, the degree of marketuncertainty can be measured from impliedvolatilities and risk-neutral densities. Inderiving market expectations and measures ofuncertainty, the measurement of risk premiaplays a key role, and should therefore beinvestigated further. Research in this directioncan show how predictable monetary policy isand how successful central bankcommunications are in preventing an increasein uncertainty in the market.

THE SIZE AND DEVELOPMENT OF CORPORATEBOND MARKETS/WHICH LIABILITIES AREREPLACED BY CORPORATE BOND GROWTH?

A very visible and pronounced event around theintroduction of the euro was the significantgrowth of primary corporate bond markets inthe euro area. How much of this can beassociated with the start of Stage Three ofEMU and how much is related to otherdevelopments, such as corporate restructuringand the financing of privatisation inthe telecommunications sector, etc.? Moregenerally, what are the macroeconomic andmicroeconomic variables that determine bondissuance volumes? Are they the same in thedomestic and international segments of themarket? Is the European corporate bond marketbecoming more even in terms of the issuingsectors or will banks continue to dominate thismarket? How broad and stable is the high-risk(“junk”) segment of the market? Whichliabilities are replaced by the growth ofbond issuance? Can further increases be


expected in the market size of asset-backedsecurities (e.g. collateralised debt obligations),optional credit instruments (e.g. spreadoptions, callable and convertible corporatebonds), and other credit derivatives (e.g. creditdefault swaps). How will the advancement ofsuch market segments affect the functioning offinancial markets as a whole?

THE MARKET MICROSTRUCTURE ANDEXCHANGE RATES

Recent research has underlined the importanceof order flows for the behaviour of exchangerates. Order flows have been shown to havepermanent effects on currency prices.However, apart from the dynamics of inter-dealer trading, what are the transactions behindorder flows that make them so informative? Arecentral bank foreign exchange operationslikely to be as informative as dealer order flowsand, therefore, more likely to have lastingeffects on exchange rates than previouslythought? How does the emergence of largeelectronic broking systems, through which thelargest share of trading volume is channelledtoday, affect the role of order flows? Has theconcentration process in the dealer communityand among non-financial firms had any adverseeffects on the efficiency of foreign exchangetrading?

DETERMINANTS OF THE COST OF CAPITAL ANDECONOMIC GROWTH

The real implications of changes in financialstructures and integration are very important.The overall efficiency of financial marketsdetermines the cost of capital. And the cost ofcapital heavily influences investment activity,firm growth (in terms of, inter alia, productionand employment) and, ultimately, economicgrowth. An important area of work would be toderive aggregate measures of financial marketefficiency and the cost of capital. Is there anytrend or structural change in the developmentof financial system efficiency and the cost ofcapital in the euro area/EU? Does theincreasing integration of financial markets in

the euro area lead to greater financial marketefficiency, as reflected in market spreads, forinstance? Are there signs of better access offirms to external financing, which allows themto grow faster? Based on the measuresmentioned above, how strong is the linkbetween the cost of capital and growth? Whatwill be the impact of current public reforms onthe cost of capital and which further reformscould lower the cost of capital in the euro area/EU? In particular, what are the links betweenregulatory changes, their effects on thefunctioning of various sources of funding andthe subsequent impact on the cost of capital forfirms? Finally, what is the impact ofglobalisation on the equity cost of capital?

FINANCIAL INDICATORS OF UNDERLYINGECONOMIC VARIABLES

Prices of financial assets provide a rich sourceof information for markets and policy-makerswith respect to investors’ expectations aboutfuture developments in underlying economicvariables, such as the economic growthprospects, inflation and monetary policy.Before the introduction of the euro, availableasset prices allowed information to beextracted for each individual nationaleconomy. Following the introduction of thesingle currency, some markets and instrumentshave been created that allow information forthe euro area as a whole to be obtained (e.g. theEURIBOR spot and derivatives markets, theEONIA swap market, and the euro-denominated swap and swaptions markets). Atthe same time, other prominent financialindicators are still based on national financialinstruments, but are often used as indicators forthe entire euro area (e.g. German Bund futuresand options and, until recently, the Frenchindex-linked bond market). The rapidevolution of financial markets in the euro areanecessitates research on what type ofinformation can be extracted from observedasset prices, in particular with respect to euroarea-wide economic fundamentals. This alsorequires careful analysis of factors other thanthose purely related to expectations, which also

49ECB


ANNEX A

affect the pricing of the financial instrumentsexamined. In particular, issues linked to riskpremia, market liquidity, microstructureeffects and other market-specific factors needto be considered carefully in this context.Available financial indicators should also beevaluated with respect to their ability toforecast economic fundamentals or the degreeof uncertainty about such fundamentals.Moreover, there is a need to assess theinformation of specific national financialindicators when these are used to provideinformation about the euro area as a whole.


The list of important topics in Section 3 israther long. Although hints where given in eachparagraph as to whether the respective topic isamong those with very high importance or not,it may be useful to repeat the limited number ofareas that have been given top priority. Each ofthese few selected areas is envisaged to becomea special theme in one of the forthcomingworkshops (or has already been one in thelaunching workshop of the network). Goodpapers in these fields will be given priority inthe workshops.

The launching workshop of the network at theECB in April 2002 already had two mainpriority areas, apart from the “agenda-setting”that led to the present “roadmap”, namelyEuropean debt market structures andinternational financial linkages. The formerarea featured, on the one hand, papersaddressing issues on the “Geographical scopeof banking and lending relationships”,“Determinants of bank mergers” and“Implications of bank consolidation forinterbank markets”. On the other hand, itfeatured papers addressing the effect of theeuro’s introduction on the “Determinants ofbond market yields”, the choice between“Equity and debt financing” and the“Organisation of primary bond markets”. Thearea of international financial linkages wasrepresented by papers on “Equity marketvolatility linkages across countries”, the“Determinants of equity market returns” andthe “Relative importance of country and sectoreffects in equity markets”.

Areas that will continue to be given very highpriority in the network are bank competitionand the geographical scope of banking,international portfolio choices and assetmarket linkages between Europe, the UnitedStates and Japan, as well as European bondmarkets. In addition, the new areas of Europeansecurities settlement systems and of theemergence and evolution of new markets inEurope (in particular start-up financingmarkets) will receive top priority.

4 MA I N P R I O R I T I E SWork on bank competition and thegeographical scope of banking continues tobe very important, as the euro area is stillexperiencing relatively few cross-bordermergers compared with domesticconsolidation, as supervisory structures andregulatory approaches are being reformed andas the “special” status of banks is beingquestioned in a number of respects. Such workshould focus on “The optimal degree of bankcompetition”, “Determinants of bank mergers”(in particular regarding obstacles and drivers ofcross-border bank consolidation),“Implications of bank consolidation” and the“Efficiency of European banks”. The resultscan be expected to provide important insightsregarding the development of integrated andefficient banking markets in the euro area in thecontext of global financial markets.

The past few decades have brought anenormous expansion of international capitalflows. Linkages created by capital flows maynow have a far larger impact on domesticeconomies than traditional trade linkages, forexample. They are, of course, intimately linkedto important asset prices, such as exchangerates, government bond and stock prices.However, knowledge about the driving factorsbehind international financial flows is stillrelatively limited. This is why priority was alsogiven to international portfolio choices andasset market linkages between Europe, theUnited States and Japan. Work on this areashould focus on “International portfoliochoices of institutional investors”,“Determinants of international portfolioflows”, “Asset price and volatility linkagesacross countries and their changes”,“Contagion and crisis linkages”,“Determinants of market returns”, “Capitalflows and exchange rates”, “Marketmicrostructure and exchange rates”,“Determinants of country investmentportfolios and the home-bias phenomenon”,“Globalisation and the increasing role ofcapital flows”. Work using disaggregated firmand high-frequency data sets is especiallyencouraged. At the methodological level,

51ECB


ANNEX A

approaches combining and integrating methodsfrom finance, open economy macroeconomicsand micro-macro capital-flow modelling seemparticularly promising. Work in this area islikely to be important to explain recent swingsin the exchange rates between the euro, theUS dollar and the Japanese yen.

Work on European bond markets shouldfocus on “Liquidity in secondary bondmarkets”, “Bond market microstructure”, “Therole of electronic bond market tradingplatforms”, “Substitutability of governmentbonds from different sovereign issuers”,“Determinants of inflation, liquidity anddefault risk premia”, “The size anddevelopment of corporate bond markets”(including explanations for the issuing boomthat occurred parallel to the introduction of theeuro), the “Organisation of primary bondmarkets”, “Selection of benchmark bonds andyield curves”, and bond yields as indicators formonetary policy and other economicfundamentals (“Market expectations and thepredictability of monetary policy”, “Financialindicators of underlying economic variables”).European bond markets have undergone rapidchanges in the past few years, including thedevelopment of euro area-wide secondarymarket trading platforms. Moreover, bondmarkets constitute a key area from the point ofview of central banks, and research related tothis market is therefore likely to be highlyrelevant from a policy perspective.

The fragmentation of the European securitiessettlement industry, resulting in high cross-border securities trading costs, may wellconstitute the single most important obstacle tofurther securities market integration. Work onEuropean securities settlement systems istherefore of great importance. It should focuson the optimal securities settlement systemstructures (in particular the optimal degree ofconcentration) and the “Restructuring ofsettlement infrastructures” (in particular thepolitical economy of securities settlementsystem reform, competition and co-operationbetween different settlement channels, and

costs and pricing of securities settlementservices). The rapid changes that the structureof the securities settlement industry currentlyis undergoing and the relatively limitedresearch available on these topics so far, makework in this area particularly important andrelevant to policy-makers.

An important difference between the US andEuropean financial systems is the greater“breadth” of markets in the former. In otherwords, the “financial architecture” of theUnited States builds on a wider range offinancial instruments traded in more liquidmarkets. On the other hand, new markets arealso emerging in Europe and some of them havedeveloped rapidly in the recent past. Given theimportance of the availability of a wide rangeof funding and investment possibilities forinnovations and risk-sharing – and henceultimately for growth and welfare – the area ofnew markets is also receiving a very highpriority. It would focus, in particular, on the“Importance and determinants of the breadthand completeness of financial markets”,“Start-up financing”, “Private versus publicequity markets”, the “Organisation of primarymarkets”. These topics cover “new” equitymarkets and venture capital. Moreover, thisimportant topic also includes work on thewidening and deepening of credit markets,such as asset-backed securities, creditderivatives, commercial paper, and repomarkets. Finally, work on the evolution ofother new derivatives markets in Europe isencouraged. Of specific interest are, forexample, the structure and liquidity of key newmarkets, the driving factors behind theirexpansion, the optimal number of markets andwelfare implications. Work with a euro-areawide perspective would be favoured and workon international comparisons (particularlybetween the euro area and countries like theUnited States and Japan) is specificallyencouraged.


Day 1: Building an agenda – Issues and methods

Monday 29 April, 2002

14:00-14:30 Jan Krahnen (CFS), Welcome addressKeynote speech by Otmar Issing (Member of the ECB Executive Board)

14:30-15:30 1) The Structure of Financial SystemsChair: Jan Krahnen (CFS)

Franklin Allen (University of Pennsylvania), The structure of financialsystems and financial stability

Eli Remolona (BIS), Debt securities market micro and macro structures

15:30-16:00 Coffee Break

16:00-17:00 2) European Financial IntegrationChair: Jean-Pierre Danthine (University of Lausanne)

Marco Pagano (University of Salerno), Measuring financial integration,based on “Study to analyse, compare, and apply alternative indicators andmonitoring methodologies to measure the evolution of capital marketintegration in the European Union” (January 2002), study for the InternalMarket Directorate General of the EU Commission by CSEF, University ofSalerno

Xavier Vives (INSEAD), Industrial organisation of banking, bankcompetition and bank market integration


17:30-18:30 3) Global Financial LinkagesChair: Axel Weber (University of Cologne)

Philipp Lane (Trinity College Dublin), International financial linkages

Giancarlo Corsetti (University of Rome III), Empirical stylised facts ofinternational financial linkages

18:30-19:00 Philipp Hartmann (ECB) and Jan Krahnen (CFS),ECB-CFS Research Network: a roadmap

20:00 Dinner speech by Tommaso Padoa-Schioppa (ECB)

Day 2: European debt market structures and international financial linkages –First results

Tuesday 30 April, 2002

8:30-9:00 Chair: Antonio Sáinz de Vicuña (ECB)

Colin Mayer (Oxford University), Corporate governance and legal structures

ANNEX B PROGRAM AND S UMMARY O F T H EL AUNCH I NG WORK SHOPHELD AT TH E E C B I N F R ANK FURT, A P R I L 2 9 - 3 0 , 2 0 0 2

53ECB


ANNEX B

9:00-10:30 Session 1.1: Law and Financial Markets

Harry Huizinga (Tilburg University and European Commission), Depositinsurance and international bank deposits (with Gaëtan Nicodeme, EuropeanCommission)

Tuomas Takalo (Bank of Finland), Law or finance: Evidence from Finland(with Ari Hyytinen and Ikka Kuosa, ETLA)

10:00-10:15 Discussant: Kostas Tsatsaronis (BIS)

10:15-10:30 Open discussion

Session 1.2: Nominal Convergence and Financial IntegrationChair: Harald Hau (INSEAD)

Giorgio Calcagnini (University of Urbino), Financial convergence in theEuropean Monetary Union? (with Fabio Farabullini, Banca d’Italia, andDonald Hester, University of Wisconsin)

Stefanie Kleimeier (University of Maastricht), European financial marketintegration (with Harald Sander, University of Cologne)

10:00-10:15 Discussant: W. Jos Jansen (De Nederlandsche Bank)



11:00-13:00 Session 2.1: Bank Competition, Credit Market Integration and PublicPolicyChair: Xavier Freixas (Pompeu Fabra)

Giovanni Dell’Ariccia (International Monetary Fund), Competition amongregulators and credit markets integration (with Robert Marquez, University ofMaryland)

Reint Gropp (ECB), Contestability, technology and banking (with SandrineCorvoisier, ECB)

Giancarlo Spagnolo (University of Mannheim), Bank mergers, competitionand liquidity (with Elena Carletti, University of Mannheim, and PhilippHartmann, ECB)

12:30-12:45 Discussants: First two papers Elena Carletti (University ofMannheim), last paper Cornelia Holthausen (ECB)



Session 2.2: The Effect of the Euro on Bond MarketsChair: Marina Emiris (National Bank of Belgium)

Roberto Blanco (Banco de Espana), The euro area government securitiesmarkets: Recent developments and implications for market functioning

Yrjo Koskinen (Stockholm School of Economics), Capital structure and theeuro (with Arturo Bris and Mattias Nilsson, Yale School of Management)

João A. C. Santos (Federal Reserve Bank of New York), The cost of barriersto entry: Evidence from the market for corporate euro bond underwriting(with Kostas Tsatsaronis, BIS)

12:30-12:45 Discussant: Christian Upper (Deutsche Bundesbank)


13:00-14:00 Lunch

14:00-16:00 Session 3.1: Regional and International Scope of Banks and CreditMarketsChair: Mark Carey (Board of Governors of the Federal Reserve System)

Claudia Buch (Kiel Institute of World Economics), Cross-border bankmergers: What lures the rare animal? (with Gayle DeLong, Baruch College)

Luigi Guiso (University of Sassari) Does local financial development matter?(with Paola Sapienza, Northwestern University, and Luigi Zingales,University of Chicago)

Steven Ongena (Tilburg University), To what extent will the banking industrybe globalized? A study of bank nationality and reach in 20 European nations(with Allen Berger, Fed Board, Qinglei Dai, Norwegian School ofManagement, and David Smith, Fed Board)

15:30-15:45 Discussant: Olivier de Bandt (Bank of France)


Session 3.2: International Financial LinkagesChair: Stijn Claessens (University of Amsterdam)

Lieven Baele (Ghent University), Volatility spillover effects in Europeanequity markets: Evidence from a regime switching model

Robin J. Brooks (International Monetary Fund), Firm-level evidence onglobal integration (with Marco Del Negro, Fed Atlanta)

Paul Ehling (University of Lausanne), The EMU and strategies of assetallocation (with Sofia B. Ramos, University of Lausanne)

15:30-15:45 Discussants: Stijn Claessens (University of Amsterdam) andDaniela Klingebiel (World Bank)



55ECB


ANNEX B

16:30-18:00 Plenary SessionChair: Vítor Gaspar (ECB)

Alberto Giovannini (C.E.O., Unifortune Asset Management), The work of theGiovannini Group and implications for research

Jesper Berg (ECB), The Eurosystem’s work on euro area financial structure

David Wright (European Commission), The status of the implementation ofthe Financial Services Action Plan

18:00-18:30 Keynote speech by Sirkka Hämäläinen (ECB)

Vítor Gaspar (ECB), Concluding remarks and future initiatives of the“Capital Markets and Financial Integration in Europe” Research Network


On 29-30 April 2002, the European CentralBank (ECB) and the Center for FinancialStudies (CFS) hosted a workshop at the ECB tolaunch their network initiative aiming atpromoting research on “Capital Markets andFinancial Integration in Europe”. The researchnetwork aims at co-ordinating and stimulatingtop-level policy-oriented research thatsignificantly contributes to the ECB’sunderstanding of developments in Europeanfinancial structure and the linkages betweenEuropean financial systems and those in theUnited States and Japan. The format is anetwork of people and its key feature will be astrong interaction between researchers in theECB, other Eurosystem central banks, otherofficial institutions and academia. Thisdocument summarises the launchingworkshop.

The workshop combined research “agendasetting” talks, research paper presentations,keynote addresses by ECB Executive BoardMembers and a plenary panel discussion.Prominent academics presented their views onthe state of the literature and provided adviceon future research agendas in three keyresearch areas: 1) the structure of financialsystems; 2) European financial integration; and3) global financial linkages. During the secondday, research papers selected through a call forpapers on European debt markets andinternational financial linkages were presentedand discussed. The second day ended with apanel discussion that included David Wright,the Director Internal Market of the EuropeanCommission, and Alberto Giovannini, whochairs an advisory group of financial marketparticipants. Three ECB Executive Boardmembers, Otmar Issing, Tommaso Padoa-Schioppa and Sirkka Hämäläinen, gavekeynote addresses at the workshop.

I N T RO D U C T I O N

Jan Krahnen (CFS) welcomed workshopparticipants and introduced Otmar Issing(member of the ECB Executive Board), whodelivered the opening keynote speech entitled“Monetary policy in an environment of globalfinancial markets”. In it, Issing stressed howfinancial markets are essential for thetransmission of monetary policy. Monetarypolicy sets only the current short-term interestrate, but for today’s investment decisionsfuture refinancing conditions are also relevant.In this context, the monetary policy strategy iscrucial to provide the markets with a referenceagainst which new information can beconsistently evaluated. The more predictablemonetary policy is, the smoother itsimplementation will be, since the market canadjust interest rates in anticipation of policyactions. Therefore, understanding of thedetermination of asset prices needs to bewidened and deepened. This is especially truefor financial markets in the euro area, whichhave seen the most remarkable pace of changeof all the developed financial markets over thelast few years. Issing encouraged new researchin the field of financial linkages, global trendsand other factors determining the evolution offinancial markets in Europe. In addition, newways to extract market information andexpectations as well as a better understandingof the propagation mechanism of monetarypolicy to economic activity through financialmarkets are highly relevant questions forpolicy-makers.

1) THE STRUCTURE OF FINANCIAL SYSTEMS

This session was chaired by Jan Krahnen(CFS) and started with Franklin Allen(University of Pennsylvania) presenting hisviews on “The Structure of Financial Systemsand Financial Stability”. Allen started bypointing out how the comparison between theAnglo-Saxon market-based system and theGerman-style bank-based system is the centraltheme of the research agenda on the structure of

BUILDING AN AGENDA – ISSUES ANDMETHODS

D AY 1

57ECB


ANNEX B

financial systems. The focus of theoretical andempirical research has been on market-basedsystems; more work is needed on theadvantages and disadvantages of the bank-based system. Issues that need to be addressedfall into four categories. First, the organisationof the financial system influences risk-sharingby households. This raises questions about therisk of the average household’s portfolios inbank-based and market-based systems.Second, the relationship between theavailability of information and the allocationof resources in bank-based systems deservesmore attention. Third, market-based systemsseem more successful than bank-based systemsin financing new firms. This raises the questionabout the best way to finance start-ups. Finally,the relationship between financial integrationand the euro is an important topic for research.The proposed research agenda on financialstability contains four categories as well. First,the relationship between the type of financialsystem (bank vs. market-based) and financialstability is of great relevance. For example, towhat degree are market-based financialsystems susceptible to financial crises (e.g.Long Term Capital Management)? Or is theregulation of banks a desirable way to ensurefinancial stability? Second, the large and rapideffect of financial crises on the real economy isan important topic. Third, the link betweenfinancial fragility and contagion needs to beexplored further. How important are contagionand financial fragility in practice and whatmechanisms trigger them? Finally, iscompetition in the financial services industrycompatible with financial stability?

In the second part of the session, Eli Remolona(Bank for International Settlements) presentedhis paper “Micro and macro structures in fixedincome markets: The issues at stake inEurope”. He discussed issues related to theeffects of the euro on market functioning (inproviding liquidity and forming prices) bydistinguishing between size and structuraleffects. A clearly interesting research topic inthis field is the measurement of liquidity inEuropean fixed income markets (e.g. the price

impact of trades, focusing on the role of privateinformation). The assessment of the structuraldifferences between European and US marketsis important as well. This includes differencesin the mix of major players, the marketmicrostructures, the benchmarks, theinformation structures and the way the centralbank operates in the markets. The second partof Remolona’s talk focused on the competitionbetween various market microstructures.Understanding the mechanisms of thiscompetition would be helpful to come to gripswith such issues as the advantages of electronictrading, the optimal level of transparency, therole of inter-dealer markets and the formationof benchmarks. Finally, the extent to whichprices in fixed income markets reflectfundamentals (factors that affect asset valuesbut are exogenous to the process of providingliquidity), including fundamentals about theunderlying economy and about credit risk, isimportant as well. The top three research issuesfor Remolona are: 1) the role of inter-dealermarkets (trading among dealers plays anessential role in the price discovery process);2) the benchmark tipping process (the tradingprocess leads to a critical level of activity inone of the instruments, so that it is used as apricing reference); and 3) the behaviour ofdefault risk premia.

The talk by Colin Mayer (Oxford University),“Corporate governance and legal structures”,was originally scheduled for the first day,together with Allen’s and Remolona’s. Due toMayer’s scheduling conflicts, this talk had tobe postponed until the following day. Forcoherence, we summarise it under this section,rather than with the submitted papers, asindicated in the programme. Mayer reviewedthe developments in corporate finance, startingwith complete markets models, movingthrough incomplete contract models up to thepolitical economy of finance. He focused onhow theoretical developments and empiricalevidence have shaped the research agenda inthis field. Starting from a point of view wherefinance decisions do not matter (Modiglianiand Miller theorem), corporate finance has now


expanded in many directions, includingfinancial institutions, corporate governance,law and finance. Great attention has been givento the role of law in the development of afinancial system. The significance and qualityof the legal system was emphasised by theintroduction of incomplete-contract models.One prediction of this branch of literature isthat whenever the legal system performs poorlynon-market processes play a more prominentrole. On the other hand, the political economyof finance stresses that law and regulation arean outcome of lobbying, and therefore cannotbe treated as exogenous. What emerges fromthe empirical literature is that there is noconvincing evidence that favours one theoryover another in explaining differences infinancial systems. Nevertheless, furtherunderstanding of these issues is of utmostimportance, not least for policy-making.According to Mayer, if one shares the view thatdifferent systems are best suited for differentactivities, the implication is that policy shouldbe enabling rather than restrictive orprescriptive. According to him, the ECB-CFSResearch Network can play an importantfunction in terms of identifying a researchagenda, facilitating research activities in thisarea through workshops, and particularlyencouraging exchanges of data as well as ideas.

Due to time constraints, there was nodiscussion after Allen’s and Remolona’spresentations. Antonio Sáinz de Vicuña(ECB) asked Mayer’s view about the conflictbetween harmonisation and competition, in thecontext of the European Union. In particular,he was worried that competition might hamperthe purposes of regulation. According toMayer, harmonisation is important, especiallyin the regulation of certain parts of the financialsystem. However, provided there is goodstandardised information disclosure, andprovided that the systemic risks associatedwith bank failures are not present, one canargue quite strongly for differences inregulatory structures and competition betweendifferent systems. Jan Krahnen (CFS) pointedout that most of the empirical work is carried

out on German and UK data. He asked whetherthere is a need for more analysis also focusingon other European countries. Mayer thinks thatthis is precisely the sort of requirement thatarises in this area of research. There has beenvery little serious international comparativework that tries to use standardised data banksand techniques across countries.

2) EUROPEAN FINANCIAL INTEGRATION

Jean-Pierre Danthine (University ofLausanne) chaired this session. The firstspeaker, Marco Pagano (University ofSalerno), talked about “Measuring FinancialIntegration”. His talk was based on a reportprepared by the Centre for Studies inEconomics and Finance for the EuropeanCommission: “Study to analyse, compare, andapply alternative indicators and monitoringmethodologies to measure the evolution ofcapital market integration in the EuropeanUnion” authored by himself, Klaus Adam,Tullio Japelli, Annamaria Menichini and MarioPadula. He evaluated the degree of Europeanfinancial integration by looking at indicatorsin four broad categories: 1) indicators of creditand bond market integration; 2) indicatorsof stock market integration; 3) indicators ofintegration based on economic decisionsof household and firms; and 4) indicatorsof institutional differences. Some of theseindicators are based on asset returns and prices,others on asset quantities (either flow or stockmeasures). When theoretical benchmarksare available, he uses the concepts ofβ-convergence and σ-convergence of returns.These indicators are taken from the economicgrowth literature. b-convergence regresses theaverage growth rate on the initial level of thevariable of interest and interprets a negativecorrelation as a sign of convergence. This isinterpreted as a measure of the speed ofadjustment of single countries towards thelong-run benchmark value. σ-convergencemeasures if countries tend to become moresimilar over time in terms of deviations fromthe benchmark, and it is computed as a cross-sectional standard deviation of a variable.

59ECB


ANNEX B

Pagano’s preliminary conclusions suggest that:a) Convergence in credit and bond marketsshould be achieved soon. It is important tomonitor the evolution of the share of foreignassets held by the national banking sectors inorder to get a measure of the home bias andtherefore of the degree of market segmentationof European credit markets; b) Monitoring ofstock market integration should be done byusing quantity-based indicators, such as theshare of equities managed by equity fundswith an international investment strategy;c) Indicators based on household decisions arerather volatile and hence unreliable. Oneshould focus instead on indicators based on thedecision of firms, such as those based onmerger and acquisition activities.

Xavier Vives (INSEAD) talked about“Industrial organisation of banking, bankcompetition and bank market integration”. Hestarted by pointing out how, over the last sixtyyears, the banking system has moved fromregulation, intervention and stability towardsliberalisation and greater instability. Thisresulted in an increase in competition,disintermediation, market integration, financialinnovation and financial fragility. Adding tothese factors, technological change ininformation technology and communicationspaved the way for a radical transformation ofthe banking sector. The banking businesswitnessed a major consolidation wave andmoved from taking deposits and granting loansto the provision of services to investors andfirms. Such developments pose key questions.Is too much competition in the banking sectorpotentially harmful, causing greaterinstability? Does the consolidation wave pose athreat to competition? Is the internet andglobalisation making banking contestable, thusreducing the scope for public intervention andcompetition policy? What is the impact ofmergers? Modern industrial organisation andfinancial intermediation analysis may provideuseful tools to answer these questions. Whatemerges is that there is a danger of bothexcessive competition and excessive market

power. The reason is that banking is a multi-product industry, with different levels ofcompetition in different product markets. Thusthe optimal level of competition depends on theinstitutional characteristics of regulation andon bank soundness. The implication is thatdifferent countries may have different optimallevels of competition intensity. Countries witha sound legal structure can benefit fromvigorous rivalry. By contrast, economies withweak institutional structures and high socialcosts of failure should moderate the intensity ofcompetition.

In the discussion, Philipp Hartmann (ECB)pointed out how the current literature is dividedabout the desirability of competition inbanking. One line of thought, based mainly onempirical contributions, supports it forefficiency reasons, while the other, based moreon theoretical arguments, is against it forstability reasons. He asked whether the speakerwould feel that the current situation in bankingsuggests more attention to competition orstability. Vives responded that the period inwhich banks were strictly regulated impliedhuge efficiency losses. The key question weneed to answer is: what is the optimal degree ofcompetition? His impression is that bankingcan take quite a bit of competition as long as itis regulated appropriately. Marco Pagano(University of Salerno) stressed the difficultyof defining integration in banking. Hesuggested that it would be a sign of integrationif a company could issue debt and take creditfrom banks in different countries on the sameterms. He then asked the following questions:How to make this definition operational at thedata level? Can we ever measure integration?Vives thinks that these are very importantissues, but that there are no clear-cut answers.For example, in retail, deposit and servicesmarkets there may be differences inpreferences and culture that explain variationsin conditions. The consequence is that thesedifferences need not be sub-optimal. The law ofone price might not make sense for some ofthese products.


3) GLOBAL FINANCIAL LINKAGES

The last session of the day was chaired by AxelWeber (University of Cologne). Philipp Lane(Trinity College Dublin), “Global FinancialLinkages: Some Theory and Empirics”,stressed the importance of gross asset trade andnet trade flows for the link between globalfinancial integration and macroeconomics.While according to the theory gross asset tradeis driven by international diversification andrisk sharing needs, reported net trade flowswould be the result of efficient capitalallocation and consumption-smoothingobjectives. Concerning gross asset trade,global factors like different risk aversions bydifferent types of investors determine assetpricing and could be used to explain systemicrisk events and contagion effects. Internationalfinancial exposure also gives rise tointernational wealth effects, which are thefocus of recent international macro research.Lane also mentioned the importance ofimproving the data situation in terms of cross-border stocks and flows. Due to the increasingimportance of valuation effects, net foreignasset positions are very difficult to measure.Using the accumulated current account as aproxy can be highly misleading. Caution is alsowarranted when deriving data on netinvestment income. The reverse sign of averagenet foreign asset positions and net investmentincome for many countries is the result of assetreturns exceeding (falling short of) liabilityreturns for debtor (creditor) countries. Detaileddata are thus essential to analyse the effects ofnet trade flows. As topics for future research,Lane suggested analysing internationalmonetary policy co-ordination from theperspective of incomplete risk sharing, lookingat non-fundamental explanations for capitalflows, reconsidering the issue of exchange ratemisalignments and further investigating globalfinancial panics and contagion events.

Giancarlo Corsetti (University of Rome III),“Global Financial Linkages and Open-MacroModels”, provided an overview of the new-open macroeconomics literature stressing the

importance of cross-border goods marketsegmentation. The recent literature followsearlier developments in international financeby showing that market imperfections providethe foundation for new cross-border spillovereffects, which here are represented by relativeprice adjustments in the goods market, i.e.movements in the terms of trade. Empiricalevidence suggests that there is a need formodels with both goods and asset marketsegmentation. This should lead to a rethinkingof the determination of exchange rates and theirrole in the international transmissionmechanism. This international transmissionchannel will affect optimal policy rules as well.There are many approaches to model goodsmarket segmentation. Corsetti mentionedsimple shipping costs, diverse cross-countrypreferences for home and foreign goods,nominal rigidities such as firms decidingwhether to price their goods in their homecurrency (producer currency pricing) or in theexport market’s currency (local currencypricing), and distribution costs paid in localcurrency. He advocated a new and promisingapproach for modelling goods marketsegmentation – vertical integration betweenfirms located in different markets (according tosome estimates 70% of international trade isintra-firm). This modelling approach is able toreproduce several attractive features, includinglocal price discrimination, an incompleteexchange rate pass-through, high exchange ratevolatility and the possibility of multipleequilibria. Corsetti suggested that future macroresearch should continue to increasingly relyon industrial organisation results and focus onthe vertical integration of firms, taking intoaccount their location decisions.

In the discussion, Harald Hau (INSEAD)pointed out that at first sight the twocontributions in this session appear quite“orthogonal” to each other, the first stressingthe importance of capital flows and the secondclaiming more attention for the micro-structureof goods market segmentation. This focus ofthe recent theoretical literature was consideredodd due to the fact that trade flows only explain

61ECB


ANNEX B

a minor share of today’s capital flows.Giancarlo Corsetti (University of Rome III)argued for patience, as researchers are trying toextend the finance part of the new-open macromodels, in the same way as the financeliterature had been improved like previouslythe finance literature had been improved byincorporating the real side of the economy.Axel Weber (University of Cologne) stressedthe similarities between international macrotheory and international finance as bothcontinue to further explore the importance offrictions in their respective fields.

After the agenda-setting presentations, JanKrahnen (CFS) gave a short presentationabout the status of the preparations of a“roadmap” to guide the future work of theECB-CFS Research Network. Krahnensuggested a first list of topics that the networkcould cover over the next two years. Thesetopics will be grouped by area of interest andwill form the basis for the future calls forpapers, announcing the next workshops andconferences. The detailed roadmap of thenetwork is discussed in a companion document:ECB-CFS research network on “CapitalMarkets and Financial Integration in Europe”:a roadmap.

Tommaso Padoa-Schioppa (member of theECB Executive Board) delivered the dinnerspeech, “Competition, co-operation, publicaction: three necessary drivers for Europeanfinancial integration”. He pointed out how theeuro can be considered not only as the crowningachievement of the European Single market,but also as the beginning of a deeper integrationprocess. By increasing price transparency,reducing transaction costs, and ultimatelystimulating competition, the single currency isacting as a powerful catalyst for financialintegration. As the title of his speech suggests,Padoa-Schioppa thinks that the completion ofthe financial integration process requiresinteraction between three (equally necessary)drivers: i) competition, ii) co-operation andiii) public action. Free competition acrossborders permits market forces to push towards

greater integration. Co-operation is needed toovercome situations where the optimaloutcome is not achieved by the simpleinteraction of independent individualdecisions. Public action should intervenewhenever a public good is at stake or when amarket failure occurs. One example whereinteraction between the three drivers is neededis the securities settlement systems. Since, inthe short run, some market participants maybenefit from its fragmentation, they lack theincentive to change. Competition by itselfcannot work and some form of co-ordination isrequired. In these circumstances, public actioncould act as a catalyst for change. He ended hisspeech by encouraging further research on howthe interaction between co-operation andcompetition can work towards the selection ofan efficient market structure.

DAY 2EUROPEAN DEBT MARKET STRUCTURES ANDINTERNATIONAL FINANCIAL LINKAGES – FIRSTRESULTS

LAW AND FINANCIAL MARKETS

Antonio Sáinz de Vicuña (ECB) chaired thesession. Harry Huizinga (Tilburg Universityand European Commission) presented thepaper “Deposit insurance and internationalbank deposits” (jointly with G. Nicodeme,European Commission), which examines howinternational depositors respond to nationaldeposit insurance policies. The authors findthat international non-bank depositors appearto favour banking systems covered by explicitdeposit insurance, and they are attracted tosystems with co-insurance, a privateadministration, and a low deposit insurancepremium. The sensitivity of non-bank depositsto deposit insurance policies opens up thepossibility of international regulatorycompetition in this area. They considered as anexample the EU deposit insurance directive of1994, which requires minimum standards fornational deposit insurance policies, but not on


the level of deposit insurance premium. Thisdirective thus may not preclude regulatorycompetition in the area of deposit insurance inEurope.

In the presentation “Law or finance: Evidencefrom Finland” by A. Hyytinen (ETLA), I.Kuosa (ETLA) and Tuomas Takalo (Bank ofFinland), the authors study the changes inFinnish corporate governance and financialsystems for the period 1980-2000. During thisperiod, Finland experienced simultaneousfinancial and currency crises, and a large-scalechange in industrial structure. Takalodescribed how the crises and the structuralchange occurred at the same time as a thoroughreform of Finnish corporate governance wasimplemented. In the paper, the authors explainthe development of the financial system in thelight of this corporate governance reform.Following the law and finance literature, theyequate corporate governance with the legalmechanisms by which outside investors areprotected. Their measures of shareholder andcreditor protection reveal that shareholderprotection has been strengthened, whereascreditor protection has been weakened. Theyfind that Finnish firms have substituted equityfor debt on a significant scale. Indeed, theweakening of creditor rights provides anexplanation for the gradual contraction incorporate lending during the 1990s. Finally,the authors find that the changes in investorprotection have to some extent preceded thereorganisation of the Finnish financial market,while changes in creditor rights occurred inparallel with market developments.

In the discussion of the first paper, KostasTsatsaronis (BIS) wondered whether depositinsurance limits are too low to be relevant forinternational depositors, such as bigcorporations and banks. Moreover, depositinsurance should have a second order effect,because other factors should dominate thebehaviour of depositors (location of trade,international capital flows, etc.). He thinks thepaper provides an interesting stylised fact (i.e.asymmetry in the behaviour of banks and non-

banks), but the interpretation of the results ispuzzling. Regarding the second paper, he foundthat it documented very well the legalframework in Finland. The paper lays a goodfoundation for future work, such as theresponse of Finland to the introduction of theeuro. Antonio Sáinz de Vicuña (ECB) askedhow it is possible to measure creditors’ orshareholders’ protection across borders in theeuro area. Takalo answered that the measurethey apply in the paper had been used in otherstudies, and that therefore he did not see anyproblem in extending their measurementcriterion to other EMU countries.

NOMINAL CONVERGENCE AND FINANCIALINTEGRATION

This session, chaired by Harald Hau(INSEAD), started with Giorgio Calcagnini(University of Urbino) presenting his paper“Financial convergence in the EuropeanMonetary Union?”, co-authored with FabioFarabullini (Banca d’Italia) and Donald Hester(University of Wisconsin). The paper reportsresults from three types of tests of financialconvergence. The first type, σ-test, is borrowedfrom the growth literature. This test was alsomentioned by Marco Pagano in his talk on thefirst day of the workshop. It examines whetherthe standard deviations of a set of cross-sectional measures of series that are believed tobe converging diminish over time. There isevidence of convergence in inflation rates,short-term nominal interest rates, but not forreal per capita GDP, relative to a group of non-EMU countries. The results from this testgenerally support the hypothesis that Europeanintegration has led to convergence in financialmarkets. A second type of test looks atconvergence in the banking industry structureby comparing levels and trends in interbankclaims and non-interest income at banks (suchas fees and commissions). One would expectthese flows to increase as barriers are removed.The results show that interbank claims werelarger at EMU banks than at banks in othercountries, but no clear trend emerged. The finaltest investigates the convergence of marginal

63ECB


ANNEX B

rates of return on assets and marginal costs ofliabilities. The main conclusion is that withinthe group of six biggest EMU countries, asingle market for banks started to emerge in thelast two years of the sample (1997 and 1998).

The second paper, by Stefanie Kleimeier(University of Maastricht) and Harald Sander(University of Cologne), was entitled“European Financial Market Integration”.The focus of the paper is on integration in retailbanking markets. The authors argue thatlooking simply at the law of one price could bemisleading, because of the high heterogeneityof the products exchanged in these markets.Therefore, they suggest a co-integrationanalysis to study the emergence of a uniformbanking market. They postulate the existenceof a long-run relationship between the nationalinterest rate of a country and the interest rate inthe rest of the euro area. While in the short rundeviations from the equilibrium relationshipare possible, in the long run these deviationsshould die out. They find very limited evidencefor co-integration before the introduction of theeuro. They also find that the relationship ofnational lending markets with the remainingeuro area lending markets exhibits strong signsof structural changes at the moment of theintroduction of the euro. According to theauthors, this is a clear indication that euro areacredit markets are changing dramatically,although more evidence is needed beforedrawing conclusions on the integration of thesemarkets. There are some tendencies towards amore uniform corporate lending market, whileconsumer lending markets are still morefragmented. Finally, they find that the pass-through process (i.e. the impact of monetarypolicy decisions on lending rates) is still farfrom perfect and exhibits strong asymmetriesacross countries, thus pointing again towardslack of integration.

The discussant Jos Jansen (De NederlandscheBank) asked how two papers on the samesubject could arrive at such different answersto the question of whether the banking sector isintegrated or not. Apart from sample periods

and other differences in the data, the answerlies in the different methodologies applied. Aproblem with Kleimeier’s approach is the shortsample available. Co-integration analysistypically requires a relatively long time series.In addition to this, extensive structural changeand, presumably, parameter instabilitycharacterised the times over which the analysisis performed. Hence care must be exercised inthe interpretation of the results. RegardingCalcagnini’s paper, Jansen found it puzzlingthat the sample ended in 1998, as we areparticularly interested in what happenedafterwards. Christian Upper (Bundesbank)made a similar point, saying that retail interestrates were prices of different assets before theintroduction of the euro and thus no conclusionabout integration could be drawn using pre-EMU samples. Francesco Papadia (ECB)asked whether similar results on convergencewould be obtained by applying the same testson average retail interest rates on bank depositsand loans. Calcagnini replied that, since nosuch analysis had been performed, he had noanswer to the question.

BANK COMPETITION, CREDIT MARKETINTEGRATION AND PUBLIC POLICY

The chairman for this session was XavierFreixas (Pompeu Fabra). The first contributionwas by Giovanni dell’Ariccia (IMF). Hepresented a joint paper with Robert Marquez(University of Maryland) entitled“Competition among regulators and creditmarket integration”. The objective of the paperis to analyse and compare the regulatoryequilibria that emerge in several countrieswhen the banking regulation decision iscentralised with the one that exists whenregulators are separate and independent acrosscountries. The underlying trade-off is betweenefficiency and flexibility. Independentregulators fail to take into account the positiveexternalities of their own regulations andtherefore tend to under-regulate compared withcentralised regulators. In other words,centralised regulators internalise positiveexternalities. However, a centralised regulator


who faces the constraint of imposing a singleset of rules on different countries necessarilyloses the flexibility that each national regulatorwas enjoying and might not fulfil country-specific needs. Therefore, the conclusion of thepaper depends heavily on the degree ofasymmetry among the countries concerned.Countries with similar financial systems willbenefit from a centralised regulator. However,the existence of multiple financial links matter,as they create a free rider problem: a countryenvisaging joining a single regulatory unionmight prefer free riding on the existingregulatory framework of its financial partners,while still benefiting from the flexibility ofhaving an independent regulator.

Reint Gropp (ECB) presented the paper“Contestability, technology and banking”written with Sandrine Corvoisier (ECB). Theobjective of the paper is to analyse whetherinternet banking increased contestability in thebanking market. The authors test a relationderived from a simple model of marketcontestability for banks. This relationexpresses the post-entry equilibrium interestmargin – a proxy for the degree of competition– as a function of market concentration,internet penetration and costs. The authors alsoconjecture that internet technology has reducedsunk costs for deposits more than for loans.Using semi-aggregated data for a group of 9 ofthe 12 euro area countries, they claim to findstrong support for an increase in contestabilityin time deposit markets. However, effects aremore moderate for loan markets. The resultsalso show that banking markets in Europe arenot perfectly contestable, which suggests thatphysical bank structures continue to matter.

Giancarlo Spagnolo (Mannheim University)presented joint theoretical work with ElenaCarletti (Mannheim University) and PhilippHartmann (ECB): “Bank Mergers,Competition and Liquidity”. The objective ofthe paper is to analyse the joint consequencesof consolidation on loan rates, reserve holdingsand interbank market liquidity. The modelfeatures stochastic withdrawal shocks on

deposits, which banks can finance either withreserves or by interbank market borrowing, andcompetition in differentiated loans. Whenliquidity shocks are uncorrelated a mergercreates an internal money market savinginterbank borrowing costs for the twoinstitutions. Surprisingly, for most parameterconfigurations this internalisation effectdominates the diversification of liquidity risk,so that merged banks increase reserve holdings.As a consequence of the internal moneymarket, they also enjoy lower liquidity risk andexpect lower liquidity needs than competitorbanks. As to the loan market, merged banksgain market power but also enjoy costadvantages through lower refinancing costsand potentially also through efficiency gains.Loan rates increase when the market powereffect dominates. Finally, aggregate banksystem liquidity improves through higherreserve holdings and deteriorates through anasymmetry in deposit bases induced by loancompetition. Hence with uncorrelated shocksthe aggregate liquidity effects of a merger arean empirical question, whereas with correlatedshocks they are unambiguously negative. Theresults have policy implications for centralbank liquidity management, antitrust analysisand financial stability.

Elena Carletti (University of Mannheim)discussed the first two papers, while CorneliaHolthausen (ECB) discussed the last one.Regarding the first paper, Carletti pointed outthat the results were strongly dependent onassumptions concerning preferences ofregulators that do not seem to be micro-founded. She argued that the reduced formequation used by the authors would need moremotivation. For example, it generates a trade-off in the regulator’s objective function thatdoes not seem to be consistent with somestandard banking theories, such as the chartervalue hypothesis. On Gropp’s paper, she askedwhich technology is the relevant one. Theweaker effect of internet banking on loanmarkets suggests that internet banking does nothave a significant impact on banks’ ability togather information about local markets.

65ECB


ANNEX B

However, other forms of technologicalprogress seem to have an impact and have madeit easier for banks to process and evaluateinformation as well as to obtain informationabout competitors. Vítor Gaspar (ECB) raiseda similar point. He remarked that, in the contextof the paper, it did not seem possible toestablish that the results found were comingfrom contestability as opposed to otherconsequences of technological progress. Forexample, internet banking led to lower searchcosts. In the last paper of the session, thediscussant Holthausen pointed out that theassumption concerning the constant rate on theinterbank money market was unsatisfactory.For instance, one can easily envisage thecentral bank changing this rate in response tomerger activity. Gaspar (ECB) suggestedextending the analysis to allow for a corridorfor interest rates in the interbank market,defined by the central bank’s standingfacilities.

THE EFFECT OF THE EURO ON BOND MARKETS

The session was chaired by Marina Emiris(National Bank of Belgium). Three papers onthe effect of the euro on tradable debt marketswere presented in this session. The firstanalysis conducted by Roberto Blanco (Bancode España), “The euro-area governmentsecurities markets: Recent developments andimplications for market functioning”, studiesthe recent key developments in the euro-areagovernment bond markets. He finds that thespreads over German bonds of previously high-yield debt have narrowed significantly. Bycontrast, the spreads of all other euro-areasovereign debt have widened since theintroduction of the single currency. The authorargues that this could reflect an increase indifferences in both liquidity and credit riskbetween German securities and other euro-areasovereign debt. The author also shows thatmarket microstructure factors, such as relativemarket liquidity, play a part in determiningrelative prices. Finally, the analysis suggeststhat the reduction in the relative supply ofgovernment bonds, as a consequence of the

improvement in public finances, has had alimited effect in the euro area, in contrast to theevidence in the US market.

The second paper, “Capital structure and theeuro”, was presented by Philip Vermeulen(ECB) on behalf of the authors, Yrjo Koskinen(Stockholm School of Economics), Arturo Brisand Matthias Nilsson (Yale School ofManagement), since the planned presenter hadfallen ill. The paper studies the changes incorporate leverage induced by the introductionof the euro. Three hypotheses are put forward.First, the diminished currency risks shouldimply a lower likelihood of financial distress,and hence should lead to increased corporateleverage. Second, the euro should lower thecost of capital, and as a result we shouldobserve decreased corporate leverage in thefirms of the countries that adopted the commoncurrency. Finally, in a fixed exchange system,firms have an incentive to maintain excessiveleverage, as the government can bail them outby devaluing the currency if they becomefinancially distressed. We then should observehigher debt ratios prior to the introduction ofthe euro for firms that benefited from currencydepreciation. The paper finds that the firms inthe euro area lowered their market-basedleverage after the introduction of the euro.Moreover, this leverage reduction is strongerfor those countries with a history of currencycrises. They also show that firms in the euroarea have increased their equity issuance sincethe advent of the euro. This supports the viewthat the euro has lowered firms’ cost of capitalby enhancing equity market integration and byeliminating currency risks within the countriesthat have adopted the common currency.

Kostas Tsatsaronis (Bank for InternationalSettlements) and João Santos (FederalReserve Bank of New York) made a jointpresentation of the paper “The cost of barriersto entry. Evidence from the market forcorporate euro bond underwriting”. The papershows that the arrival of the euro had animportant impact by reducing the underwritingfees of international corporate bonds issued in


the new currency. The euro brought down thecosts of this funding source to levels similar tothose in the US. This finding is particularlystriking, given the fact that in 1994 the averagefee for bonds denominated in Europeancurrencies was twice the corresponding figurein the US. The paper also shows that thereduction in the cost of corporate bondunderwriting was largely due to greatercontestability of the investment bankingbusiness achieved with the introduction of theeuro. Therefore, one can conclude that the costsof economic barriers to entry in the pre-EMUmarket were significant, and their eliminationgave borrowers in the new currency a widerrange of options. A second finding of the paperis that the elimination of market segmentationincreased the business of the largerinternational investment banks, rather thanintensifying the links between euro areaborrowers and bankers from the same country.This is interpreted as evidence that borrowersput a greater weight on reputation than onbusiness relationships in the choice of anunderwriter.

Christian Upper (Deutsche Bundesbank)acted as discussant. Regarding Blanco’s study,he questioned the use of the premium of the on-the-run bond as a proxy for liquidity. The on-the-run bond, for the German market at least,does not command a narrow spread, and hencecannot be considered more liquid than otherbonds. This is likely to introduce somemeasurement error in the analysis. RegardingSantos’ and Tsatsaronis’ paper, Uppermentioned that gross fees constitute only a partof the costs of underwriting. This might havesome consequences in the interpretation of theresults, as a reduction in the costs ofunderwriting should also include the discountone actually gets on the market. Issam Hallak(CFS) questioned whether the increaseddemand for euro assets from US investors couldexplain the reduction in underwriting costs inSantos’ and Tsatsaronis’ paper. Marco Pagano(University of Salerno) asked how manycompanies gained access to the bond market forthe first time. In his opinion it would be

interesting to know the difference betweentheir yield and the normal yield in the samesector. Santos answered that these are veryinteresting questions, but their data set does nothave the detailed information to answer them.Regarding the effective costs of theunderwriting, Jan Krahnen (CFS) askedwhether the authors had discussed the issuewith market practitioners, in order to be surethat competition really happens on the nominalfees and not on something else outside the data(similar to what is observed in credit markets).Tsatsaronis spoke with a representative fromthe Association of Bank Underwriters, whoagreed with the basic thrust of the paper, in thesense that the corporate bond market hasbecome more competitive in the euro area.

REGIONAL AND INTERNATIONAL SCOPE OFBANKS AND CREDIT MARKETS

Mark Carey (Fed Board) chaired the session.Claudia Buch (Kiel Institute of WorldEconomics, with Gayle de Long) in “Cross-border bank mergers: What lures the rareanimal?” provides an empirical analysis onthe causes of cross-border bank mergers. Theauthors use a novel data set of over 2,300mergers that took place between 1978 and2001. They study the impact of countrycharacteristics such as language, regulation, ordistance, on the likelihood of a merger betweena country pair. They also analyse changes inmerger behaviour over time, and study theimpact of selected regulatory changes onmerger activity. It is found that regulation is adriving factor behind international mergers. Inparticular, their results suggest that banksoperating in more regulated environments areless likely to be the targets of internationalbank mergers. Therefore, the authors concludethat the lifting of regulatory barriers can spurgrowth in cross-border bank mergers. Still,informational barriers such as distance andcommon cultural factors will continue to holdback merger activity. The paper also providesevidence that banks from developed countriestend to take over those from developingcountries. Finally, the results in the paper

67ECB


ANNEX B

suggest that during the 1980s, bank mergerstended to occur between banks from similarcountries, even if those banks were located indifferent continents. In the 1990s, as easternEuropean and Latin American countriesopened up, banks from western Europe and theUnited States began to engage in cross-border,intra-continental mergers.

Luigi Guiso (Ente Einaudi, with PaolaSapienza and Luigi Zingales) in “Does localfinancial development matter?” studies theeffects of differences in the development ofnational financial institutions, when domesticagents have access to foreign financialmarkets. Rather than studying data fromseveral countries, the authors use Italian data atprovincial level, and argue that the results canbe interpreted as an upper limit to whatintegration can achieve. An indicator of localfinancial development is constructed, whichreflects the notion that developed financialmarkets grant easier access to external funds todomestic households and firms. By using thisindicator, the authors find that (local) financialdevelopment enhances the probability that anindividual starts his own business, favoursentry, increases competition, and promotesgrowth of firms. These results are interpretedas evidence that even if financial marketsbecome increasingly integrated, domesticfinancial institutions do not become redundant.Furthermore, as predicted by theory, the resultsindicate that large firms are more likely to seekfinance on the international market thansmaller firms. Overall, the results suggest thatlocal financial development is an importantdeterminant of the economic success of an area,even in an environment where there is nofriction between capital movements.

Steven Ongena (Tilburg University, withAllen Berger, Qinlei Dai, and David Smith),“To what extent will the banking industry beglobalized? A study of bank nationality andreach in 20 European nations” argues that thebanking industry may never become fullyglobalised, even after adjusting to the fulleffects of deregulation, technological progress

and increased cross-border non-financialactivity. The proper question is, rather, to whatextent will the banking industry be globalised?The paper studies the importance of differentfinancial service providers (ranging from asmall bank located in the host nation, to a largeglobal bank in a distant financial centre) for theprovision of cash management to largemultinational corporations. The term cashmanagement refers to all short-term bankingneeds (including liquidity management, short-term lending, foreign exchange transactions).The authors, using data on more than 2,000foreign affiliates of these corporations in eachof 20 European nations, identify two distinctdimensions of globalisation: bank nationalityand bank reach. The first refers to the locationof a bank’s headquarters. Bank reach refers tothe geographic scope and size of the bank. It isfound that for their local cash management,most multinational affiliates choose a small,local institution located in the country in whichthe services are demanded, while only a rathersmall percentage uses the services of a globalfinancial institution. Moreover, it is also foundthat bank reach is strongly associated with banknationality, i.e. firms that use host nation banksfor cash management services are less likely touse a global bank and more likely to use a localor regional bank. The findings suggest that theextent of future globalisation of the bankingindustry is limited.

In his discussion, Olivier de Bandt (Banque deFrance) emphasised the common theme of allthe papers in the session, namely testingfinancial integration at national and/orinternational level. The focus is particularly oncredit market integration, and the behaviour ofhouseholds and firms. All papers providedsome evidence of potential limits toglobalisation in the banking industry: localfinancial institutions will continue to play arole for domestic markets. Another distinctivefeature of the three papers is the use of noveldata sets. Concerning the paper by Buch, heremarked that mergers are only one way topenetrate a foreign market. To obtain a measureof financial integration, other strategies such as


the number of foreign branches or subsidiariesshould be taken into account. One criticismraised was the pooling of acquiring andtargeted banks in the sample. On the paper byGuiso, de Bandt encouraged the authors toprovide a clearer interpretation of the newlydeveloped financial indicator. He alsocriticised the lack of measures for the goodnessof fit in the estimations. Finally, he raised somedoubts about the stability of the resultsobtained by Ongena, noting that more detailsshould be given to explain why variables linkedto eastern Europe were the only ones to besignificant in the empirical results. All in all,he concluded that the analyses pointed to theexistence of two polar cases, which may berelated to the underlying strategy of clients:regional banks or global banks.

INTERNATIONAL FINANCIAL LINKAGES

Stijn Claessens (University of Amsterdam)chaired the session. Lieven Baele (GhentUniversity) presented his paper “Volatilityspillover effects in European equity markets:Evidence from a regime switching model”. Heestimates the magnitude and the time-varyingnature of volatility spillovers from aggregateEuropean and US equity market indices to 13local European equity markets. The ultimategoal is to analyse the fundamental forcesdriving volatility in European stock markets.In the analysis, he allows volatility in thedifferent European markets to depend on apurely country-specific shock, a regionalEuropean shock and a global shock from theUS. More specifically, he investigates whetherthe process of economic, monetary andfinancial integration have fundamentallyaltered the sources and intensity of shockspillovers to individual European stockmarkets. He uses a regime switching model todistinguish between periods of high and lowspillover intensity, and high and low volatility.He finds these regime switches to be veryimportant, which implies that shock spilloverintensity varies significantly through time. Theimportance of EU shocks increased for mostmarkets during the 1990s, and has become

important for euro area countries and Denmark.For countries like Norway, Sweden,Switzerland, and the United Kingdom, theimportance of US shocks is still higher thanthose coming from the EU.

Robin Brooks (International MonetaryFund) presented his joint paper with MarcoDel Negro: “Firm-Level Evidence onGlobalization”. He argued that the degree ofco-movement across national stock marketshas increased dramatically in recent years. Thepaper aims at exploring some of the causesbehind this phenomenon. It does so byconstructing a new firm-level data set thatcovers monthly stock market data and annualbalance sheet and income statementinformation from 1985 to 2002 for about10,000 companies in 42 developed andemerging stock markets. The authors find thatthe ability of country-specific effects toexplain international variation in asset andsales growth and returns on assets fellsignificantly during the late 1990s, while theexplanatory power of global industry effectsgrew and in some cases surpassed that ofcountry effects. Using this data set, they alsoestimate a factor model, which decomposesfirm-level equity returns into a global, acountry-specific, an industry-specific and afirm-specific component. Compared with theprevious literature, the factor model drops theassumption that firms have the same exposureto a given country or industry factor. Theirmain finding is that country-specific shocksplay a smaller role in explaining the stockreturn variation of firms that are diversifiedinternationally, while the country-specificcomponent plays a greater role for companiesthat are more domestically oriented.

Paul Ehling (University of Lausanne)presented his joint paper with Sofia Ramos(University of Lausanne) entitled “The EMUand strategies of asset allocation”. Theauthors test two strategies for portfolioallocation: country versus industrydiversification. They do this by testing whethera set of industry portfolios can improve the

69ECB


ANNEX B

efficient frontier of country portfolios and, viceversa, whether a set of country portfolios canimprove the efficient frontier constructed fromindustry portfolios. They isolate the effects ofEMU by considering different kinds ofsamples: euro area countries, European Unionbut not EMU countries, and other Europeancountries. Their results show that since theintroduction of the euro, country portfolioshave provided equally good diversificationopportunities as industry portfolios. However,mixing both strategies still offersdiversification gains. They do not findsignificant differences between euro areacountries and different groups of non-euro areacountries. Therefore, the authors conclude thateither EMU is not responsible for the apparentshift in the efficient portfolio frontier, or that ithas affected all the countries in Europe,regardless whether they have joined EMU ornot.

Daniela Klingebiel (World Bank) discussedRobin Brooks’ presentation. She raised somedoubts about the dependence of the results onthe particular data set used. The data set seemsto be heavily biased towards the U.S. Thereforeit is not clear to what extent the results reallyhold across countries and globally. She isparticularly concerned about the situation foremerging markets. She presented some resultsof her own work, showing that during the 1990sinternationalisation in capital raising andtrading in foreign shares grew rapidly,especially in middle income countries. In thesecond part of the discussion, Stijn Claessens(University of Amsterdam) pointed to the needin the three papers for a well-defined asset-pricing model. The papers use a single factorasset-pricing model, running into the problemthat it becomes impossible to distinguishbetween tests for lack of diversification andtests for lack of integration. Only with a multi-factor model can one explicitly test for thesources of financial integration: is a higherdegree of integration the result of financialconvergence, real convergence or similarconcepts of economic policy? Further, hestressed how these works suffer from being

partial equilibrium models. He pointed towardsa growing body of literature that shows how theintroduction of small transaction costs in ageneral equilibrium model of internationalasset pricing can generate a large deviationfrom the optimal prices. This suggests theinclusion of stock positions or flow side inempirical analyses of this kind.

PLENARY PANEL SESSION

Vítor Gaspar (ECB) chaired the last session ofthe workshop. In this panel, technical andpolitical experts were invited to give theirviews about the status of the Europeanfinancial system and the way forward tocomplete integration.

Alberto Giovannini (C.E.O., UnifortuneAsset Management) began his presentation,“The work of the Giovannini Group andimplications for research”, by recalling themandate of the Giovannini Group and the workit has performed so far. The group was set upin 1996 and consists of financial marketparticipants. Its purpose is to advise theEuropean Commission on financial marketissues. Specifically, the group puts togetherfinancial, economic and legal viewpoints, withthe aim of identifying inefficiencies andproblems in EU financial markets andproposing possible solutions to increasemarket integration. A key area where seriousbarriers to integration exist is the clearing andsettlement area. Here, technical impediments,as well as fiscal and legal cross-border barriersare still in place. The group’s analysis foundthat, as a result of this type of impediment,cross-border transactions are subject to post-trading costs 8–10 times higher than those fordomestic transactions. Giovannini argued thatthe combination of widespread barriers andhigh transaction costs paints a bleak picture ofEU financial markets in the near future. He alsodiscussed the potential role of policy-makersin promoting EU financial integration. Hementioned several problems policy makerswould have in tackling this process, includingpowerful special interests, the information


disadvantages of policy-makers vis-à-vismarket participants, and the fact that it is hardto describe and quantify the benefits ofintegration. He concluded by raising a numberof questions. These included the impact ofeliminating barriers to integration onindividual national markets, the possibility thatincreased integration may be destabilising, andthe role of regulators in an integrated EUmarket.

Jesper Berg (Head Capital Markets andFinancial Structure Division, ECB) outlinedthe ongoing non-research work of the ECB inenhancing the Eurosystem’s analysis offinancial structures in the euro area in “TheEurosystem’s work on euro area financialstructure”. The goal is to produce, within thenext few months, a report that will serve as themain reference source in this area. A number ofreasons were given for the importance of thiswork, including the need to improve the overallstructural analysis, to improve understandingof the monetary policy transmission process,and to better understand key determinants oftrends in financial structures. Berg mentionedthat while the Eurosystem has alreadyproduced extensive work on financial structurein the euro area, the current project focuses onmore general aspects of the euro area financialstructure. The envisaged output of the project isa publication with 12 country chapters, as wellas a euro area chapter. Each of these chapterswill deal with the specifics of the sources offinancing, the markets/intermediaries, and theuses of financing in each country and the euroarea. Berg advanced a few preliminaryconclusions that could be drawn from thework done so far. For example, householdsappear to be the main providers of funds(through intermediaries), and non-financialcorporations are the main recipients of funds inthe euro area. A relatively clear trend in thefinancial structure of the euro area is the rapidinternationalisation of the financial sector,although more so in markets than amongintermediaries. In contrast to these findings,which seem common across a majorityof countries, there appear to be substantial

differences between countries regarding theratio of external to internal financing for non-financial corporations. Finally, with respect tothe household sector, housing loans seem to bethe major source of financing, although someimportant differences between countries exist.

David Wright (Director Internal Market,European Commission) discussed “The statusof the implementation of the Financial ServicesAction Plan”. He described the progress madeto date on the Financial Services Action Plan(FSAP) and reflected on future issues. TheFSAP, which was adopted in 1999 by theEuropean Commission, outlines the overallpolicy of the EU for achieving integratedfinancial markets. Wright reported thatimportant progress had been made and the workis currently halfway through. There isincreasing support within the EU for deeperfinancial integration, and therefore a growingneed for more resources in order to monitorprogress in the future. Despite the recentprogress in implementing the FSAP, Wrightnoted that a large number of important issuesare still outstanding. For example, further workis required on the Investment ServicesDirective, ultimately depending on the choiceof trading structures in the EU. Cross-bordermergers are another important area that needsfurther development. In particular, hehighlighted the need for a new directive ontakeovers and for common disclosure rules. Healso mentioned a number of legal issues,including the need to harmonise and modernisecompany law. More work is needed in the areaof settlement and clearing. The main concern atthe moment, according to Wright, relates topension funds, where no progress had beenmade to date. Specifically, he raised thequestion whether Europe’s markets could copewith future expected pension flows. In the lightof these and other outstanding issues, Wrightdiscussed some areas where he believes futureresearch would be particularly valuable. Hementioned the need for long-term monitoringsystems for the progress of integration in EUfinancial markets. The role of EU enlargementand its possible consequences and benefits was

71ECB


ANNEX B

also pointed out. He suggested that theconsequence of tax distortions in capitalmarkets should be studied more, in order to findout whether these are important and whetherthe market can be expected to circumvent suchdistortions by itself. He also highlightedinformation issues. Specifically, the potentialrole of disclosure externalities and their impacton market volatility and liquidity should beinvestigated further, as well as the trade-offbetween efficiency and supervision.

In the general discussion, the question wasraised about the ranking of the importance ofvarious barriers and impediments for financialintegration in the EU. Giovannini respondedthat it is difficult to make a specific ranking ofsuch barriers. Moreover, he prefers to focus onways of disposing of integration impediments,rather then ranking them. Relating to thesubstantially higher cost of cross-bordertransactions, as compared with domestictransactions, Giovannini was asked what hethought the real implications of this higher costwere. Giovannini responded that he saw thehigh cross-border costs as a prohibitive tariff,resulting in fewer cross-border transactionsthan would otherwise be the case. He expressedthe view that a number of additional marketfunctions could be carried out and a betterallocation of resources could be achieved if thisbarrier were eliminated. Giovannini was askedhow his conclusions will be conveyed. Heresponded that the final report will makeproposals on how impediments for marketintegration can be eliminated. It will also covervarious legal aspects relating to this.Furthermore, it will have an analysis ofdifferent financial structures and theirproperties with respect to efficiency. Wrightnoted that while a large number of problemshave been identified in the area of cross-borderclearing and settlement, domestic clearing andsettlement within each EU country cannot bedescribed as inefficient. He expressed the viewthat the market alone cannot be expected tosolve the current problems associated withcross-border clearing and settlement, since

these involve legal, competition, and accessissues, among others.

Sirkka Hämäläinen (member of the ECBExecutive Board) delivered the concludingkeynote speech, “Consolidation in theEuropean securities infrastructure. What isneeded?”. She pointed out how the securitiesinfrastructure in the euro area remains highlyfragmented. A large number of stock andderivative exchanges, and several nationalclearing and settlement institutions stillsurvive, despite the consolidation processtriggered by the single currency. In order tofully reap the benefits of the single currency infinancial markets, this problem must be solved.The key question she posed is whether theconsolidation process of the securitiesinfrastructure will proceed without some formof public involvement. It is true that in acompetitive environment market forces pushfor the most efficient solution. But howcompetitive is the European securitiesinfrastructure? And, in any case, will marketpressures be sufficient to overcome existingnational interests? Available analysis of thebusiness environment for securities trading,clearing and settlement indicates the existenceof several sources of insufficient competition.According to Hämäläinen, there are two mainfields in which public action can play aprominent role: removing obstacles toconsolidation and ensuring an integratedregulatory and oversight framework. Sheconcluded her speech by suggesting a fewtopics for research: 1) analyse the economiccircumstances under which there is anargument for public involvement in theconsolidation process itself; 2) what is theoptimal degree of concentration in thesecurities industry?; and 3) what is the impactof the “insourcing” of securities services on theconsolidation of the industry and on theoversight, supervision and financial stabilityfunctions?

Vítor Gaspar (ECB) closed the workshop,stressing that the research network will cover


three broad topics: 1) European financialintegration; 2) European financial systemstructures; and 3) financial linkages betweenEurope, the US and Japan. The main purposesof the network will be to provide acomprehensive research agenda, provide a“roadmap” identifying the topics to be coveredin future workshops and conferences,encourage and facilitate exchanges of data, andreserve a special role for young researchers.Young researchers will be invited to submitresearch proposals through a call for projects.The most promising proposals will be selectedand funding will be available. All researchersinvolved in the network are expected toproduce original research on relevant topics,present it at the forthcoming workshops andconferences, interact with other researchers inthe network, and eventually publish theirresearch in top academic refereed journals. Heannounced that the summary proceedings of theworkshop and the research roadmap of thenetwork will be made available shortly andposted on the network website. Dates andvenues of the next workshops will also beannounced on the network website.

73ECB


ANNEX C

A f f i l i a t i on

(% share of total participants)

ANNEX C A F F I L I AT I ON O F P A RT I C I PAN T STO N E TWORK E V EN T S

Launching Workshop Second Workshop

Third Workshop Symposium

Academia34%

EuropeanCentral Bank

33%

Eurosystem National

Central Banks16%

other Central Banks

5%

private sector and other institutions

12%

Academia43%

European Central Bank

17%

EurosystemNational

Central Banks27%

other Central Banks

3%


10%

Academia47%


21%

EurosystemNational

Central Banks16%

other Central Banks

7%


9%

Academia38%


34%

EurosystemNational

Central Banks13%

other Central Banks

5


10


Day 1 (Tuesday March 11)

8:45-9:15 Welcoming coffee

9:15-9:30 Sinikka Salo (Executive Board Member, Bank of Finland), Opening remarks

9:30-10:15 Steven Ongena (Tilburg University),Key lecture on the impact of technology and regulation on the geographicalscope of banking activities: theory and evidence


10:30-11:00 Coffee break

11:00-13:00 Parallel sessionsSession 1.1: Bank consolidation and its implicationsChair: Rune Stenbacka (Swedish School of Economics, Helsinki)

Emilia Bonaccorsi di Patti (Bank of Italy), Winners or losers: the effects ofbanking consolidation on corporate borrowers (with Giorgio Gobbi, Bank ofItaly)

Maria Fabiana Penas (Free University of Amsterdam), Gains in bank mergers:evidence from the bond markets (with Haluk Unal, Robert H. Smith School ofBusiness, University of Maryland)

Dag Morten Dalen (Norwegian School of Management), Regulatorycompetition and multi-national banking (with Trond E. Olson, NorwegianSchool of Economics and Business Administration)

12:30-12:45 Discussant: Jukka Vesala (ECB)


Session 1.2: European equity markets and corporate governanceChair: David Mayes (Bank of Finland)

Yrjo Koskinen (Stockholm School of Economics), The euro is good after all:corporate evidence (with Arturo Bris, Yale School of Management, andMatthias Nilsson, Stockholm Institute for Financial Research)

Thomas Gehrig (University of Freiburg), Cross-listings and the geography offirm’s ownership (with Thierry Foucault, HEC)

Mariassunta Giannetti (Stockholm School of Economics), Which investorsfear expropriation? Evidence from investor stock picking (with AndreiSimonov, Stockholm School of Economics)

ANNEX D PROGRAM AND SUMMARY OF THESECOND WORKSHOPHOSTED BY THE BANK OF FINLAND IN HELSINKI,MARCH 11-12, 2003

75ECB


ANNEX D

12:30-12:45 Discussant: Leo de Haan (Dutch Central Bank)


13:00-14:00 Lunch break

14:00-14:45 Tullio Jappelli (University of Salerno)Key lecture on financial market integration, corporate financing and economicgrowth



15:30-17:30 Parallel sessionsSession 2.1: Financing structure of firms; theoryChair: Jan Krahnen (Center for Financial Studies)

Xiaoyun Yu (Indiana University), Informational efficiency and liquiditypremium as the determinants of capital structure (with Chung Chan, Universityof Minnesota)

Tuomas Takalo (Bank of Finland), Equilibrium in financial markets withadverse selection (with Otto Toivanen, Helsinki School of Economics)

Thorsten Koeppl (ECB), Limited enforcement and efficient interbankarrangements (with Jim McGee, University of Western Ontario)

17:00-17:15 Discussant: Leo Kaas (University of Vienna)


Session 2.2: Integration of equity marketsChair: Harry Huizinga (Tilburg University and Economic Advisor to theEuropean Commission)

Michael Haliassos (University of Cyprus), Household stockholding in Europe,Where do we stand and where do we go (with Luigi Guiso, University of Sassariand Tullio Jappelli, University of Salerno)

Eric Theissen (University of Bonn), Competition between exchanges: Euronextvs. Xetra (with Maria Kasch-Haroutounian, University of Bonn)

Miguel Angelo Ferreira (ISCTE School of Business), The importance ofindustry and country effects in the EMU equity markets (with Miguel AlmeidaFerreira, ISCTE School of Business)


17:00-17:15 Discussant: Simone Manganelli (ECB)


19:30 DinnerBenn Steil (Council on Foreign Relations, New York),Dinner speech on building a transatlantic securities market

Day 2 (Wednesday March 12)

8:00-8:30 Coffee

8:30-9:15 Geert Rouwenhorst (Yale School of Management)Key lecture on international financial linkages



10:00-12:00 Session 3: International portfolio choice and asset market linkagesChair: Philipp Hartmann (ECB)

Helene Rey (Princeton University), Exchange rates, equity prices and capitalflows (with Harald Hau, INSEAD)

Michael Ehrmann (ECB), Interdependence between the euro area and the US:what role for EMU? (with Marcel Fratzscher, ECB)

Charlotte Ostergaard (Norwegian School of Management), Internationaldiversification in bank asset portfolios (with Claudia Buch, Kiel Institute forWorld Economics, and John C. Driscoll, Board of Governors of the FederalReserve System)

11:30-11:45 Discussant: Yigal Newman (Stanford School of Business)


77ECB


ANNEX D

12:00-13:30 Policy panelThe future of exchanges: competition or consolidation?Moderator: Juha Tarkka (Bank of Finland)

Niall Bohan (European Commission, Brussels)

Peter Gomber (Deutsche Boerse AG, Frankfurt)

Hannu Halttunen (Nordea, Finland)

Andre Went (Euronext, Paris)

13:30-13:45 Vítor Gaspar (ECB), Concluding remarks

13:45-14:45 Lunch


On 29-30 April 2002, the European CentralBank (ECB) and the Center for FinancialStudies (CFS) hosted a workshop at the ECB tolaunch their network initiative aiming atpromoting research on “Capital Markets andFinancial Integration in Europe”. The researchnetwork aims at co-ordinating and stimulatingtop-level policy-oriented research thatsignificantly contributes to the ECB’sunderstanding of developments in Europeanfinancial structure and the linkages betweenEuropean financial systems and those in theUnited States and Japan. The format is anetwork of people and its key feature is a stronginteraction between researchers in academia,the ECB, other Eurosystem central banks andother official institutions. On the basis of thediscussions held during the Launchingworkshop regarding the areas where research isneeded, five top priorities areas have beenselected: (1) bank competition and thegeographical scope of banking activities;(2) international portfolio choices and assetmarket linkages between Europe, the UnitedStates and Japan; (3) European bond markets;(4) European securities settlement systems;and (5) the emergence and evolution of newmarkets in Europe (in particular start-upfinancing markets). Subsequent workshopswere designed to cover these different areas.

The second workshop of the network washosted by the Bank of Finland in Helsinki on11-12 March 2003. The main priority areasanalysed in the course of the workshop were (1)bank competition and the geographical scopeof banking activities; (2) Internationalportfolio choices and asset market linkagesbetween Europe, the US, and Japan and (5) theEuropean equity markets. The workshopcombined research key lectures, research paperpresentations, and a plenary panel discussionon “The future of exchanges: consolidation orcompetition”, that included Niall Bohan(European Commission, Brussels), PeterGomber (Deutsche Boerse AG, Frankfurt),Hannu Halttunen (Nordea, Finland), and Andre

I N T RODUC T I ONWent (Euronext, Paris). This documentsummarises the second workshop of theNetwork.

79ECB


ANNEX D

Juha Tarkka (Bank of Finland) welcomedworkshop participants and introduced SinikkaSalo (member of the Bank of Finland executiveboard), who delivered the opening remarks. Sherecalled that many of the processes that areshaping Europe today are closely linked to thedevelopment and working of financial markets.On the one hand, changes in Europe politicalstructures are reflected in its financial system.For example, the most crucial change on ourcontinent at the present time is the enlargementof the European Union. The challenge forfinancial markets is the collection and transfer ofhuge amounts of capital that is needed in the 10new member states to enable them to catch upwith the living standards of the older memberstates. It is most likely that, 10 years into thefuture, the European financial markets will inmany respects be quite different from theircurrent situation. On the other hand,developments in financial markets have a crucialimpact on how the EU can be renewed and how itwill develop in the years to come. Despite thesuccessful introduction of the euro and the singlemonetary policy, EU financial markets are inmany respects still fragmented and national.Finally, she recalled the result of several studiesshowing that further financial integration can beexpected to have significant positive effects ongrowth and increase market efficiency.

KEYNOTE LECTURE ON THE IMPACT OFTECHNOLOGY AND REGULATION ON THEGEOGRAPHICAL SCOPE OF BANKINGACTIVITIES

Juha Tarkka (Bank of Finland) thenintroduced Steven Ongena (TilburgUniversity) who addressed two themes in hiskey lecture “The impact of technology andregulation on the geographical scope ofbanking”, (i) the implications of distance andborders in banking and (ii) the effect oftechnology and regulation on thesegeographical barriers.

Ongena first clarified the concept of borders.Economic borders arise endogenously frominformation problems due to strategic

behaviour of market participants, e.g.relationship lending can be seen as a barriers toentry. Economic borders also have anexogenous component, such as differentstandards of legal origin, corporate governancepractices, supervisory practices, politicalsystems or languages.

He then reviewed the theory and evidencerelated to distances and borders in four areas:1) spatial pricing and rationing; 2) branchingand servicing; 3) segmentation and 4) entry andmergers and acquisitions. In the first area,theories on the effects of transportation costs,monitoring costs or asymmetric informationpredict that loan rates should be positivelyrelated to distance to the closest competitor,while it should be negatively related to distanceto the lender. These relationships are stronglysupported by the empirical evidence.Regarding the second area “branching andservicing”, Ongena reported that there is littlesupport that the strategy of banks to openbranches is affected by distance. Relatedsegmentation, there is convincing evidencethat banks over-invest domestically. Theexistence of country-specific credit risk or thefact that foreign banks may not be in a betterposition than local lenders to extend credit toborrowers than local lenders might explain thisfact. Regarding the fourth area, entry andM&As, Ongena reported that cross-borderbank M&As are less frequent than cross-borderM&As in other industries or than M&Asbetween domestic banks. So borders are asignificant impediment to bank M&As.Finally, on strategies of entry, the evidence ismixed. Contrary to banks in the US, banks inGermany e.g. appear to adopt a “follow-your-customer” strategy. However, it seemsaccepted in the literature that lower efficiencyin banking activities is related to borders andnot to distance.

Ongena then elaborated on the likely impact oftechnology and regulation on distance andborders, still focusing on the four areas listedabove. Regarding pricing, new technology likeon-line banking should spur competition thus

DAY 1


reducing loan rates. However there is only littleevidence that this is the case. Indeed, there aretwo forces at work. A better access toinformation has a negative effect on loan rates,but an improved ability to process informationincreases loan rates and bank profits.Regarding rationing, banks should faceincreased competition from capital markets,which might induce banks to lend across largerdistance. Ongena presented empirical supportfor this claim, showing that the estimateddistance between banks and lender increased inBelgium and in the US. Regulation has a largerimpacts on areas 3) and 4). In particular,Ongena reported that while the integration ofthe EU wholesale capital market was achieved,there is still a long way to go in order tointegrate the retail loan market. In particular,contrary to the US, the distance at which bankslend internationally has not increased inEurope. As a bottom line, while regulatory“borders” in the EU have been removed, theexogenous economic borders remain verystrong. Against this background, Ongena drewtwo main policy implications. First, it appearswarranted to facilitate cross-borders M&Asusing for instance pro-active competitionpolicies. Second, it appears necessary to reflecton the current status of the regulatoryframework, which might not be able to dealwith complex corporate structure arising fromcross-borders M&As.

In the discussion, Thomas Gehrig (Universityof Freiburg) asked whether Hotelling’sframework – a theory to analyse the choice oflocation as part of a competitive strategy –could be used to understand the pattern of bankbranching in Europe. In his view, an increase indistance has a negative effect on the number ofbranches opened and he invited the authors toexplore the interaction between informationand market structure effects. Ongenaresponded that he and his co-author HansDegryse (Tilburg University) did not exploreyet the essence of local competition in Europebut suggested that, to do so, it would benecessary to explore the identity of banks thatare located nearby the lending institution.

Philipp Hartmann (ECB) then asked theauthors to elaborate further on the future ofbank supervision in Europe. In Ongena’s view,supervision should become more integrated atthe European level so as to build up expertise asEuropean M&As proceed. At the same timenational supervisors should be maintained.Regarding the impact of technology on bankingactivities, Jukka Vesala (ECB) stressed thatthe amount of loans granted by online banks islow, possibly because of the informationaldisadvantage these banks face. However,Vesala wondered whether these banks willbecome more serious competitors when theywill be able to receive deposits. Ongena repliedthat ultimately technology will have an effecton the competitiveness of online banks, but healso emphasised that it will become moredifficult for them to compete when they willengage in cross-border activities. Juha Tarkka(Bank of Finland) invited the author toelaborate further on the interaction between theexploitation of economies of scale andtechnological progress in banking. In his view,because of the possibility to outsource manyoperations, size does not really matter anymore. On this matter, Ongena referred to theUS experience where small banks outsourcemany activities to information technologyproviders but retain some aspects of theirbusiness such as lending activities. The sameprocess is taking place in Europe, although at amuch slower pace. Then the authors were askedto be more precise why they regardtransportation costs as important, given newtechnologies. Degryse replied that in his view,transportation costs are important for thedecision to take a loan as well as for the easewith which a loan can be monitored. He alsoemphasised that transportation costs arerelated to “informational distance”. Finally,the authors were asked their views on thelikelihood of a pan-European retail market.Degryse answered that so far there is noempirical evidence that retail markets becomepan-European. Hartmann further commentedthat progresses on the retail side are limited ona European-wide level because of theexogenous borders underlined by Ongena.

81ECB


ANNEX D

There are some inherent limits to theintegration of retail business, some of whichmay be impossible to overcome.

SESSION 1.1 BANK CONSOLIDATION AND ITSIMPLICATION

The session was chaired by Rune Stenbacka(Helsinki Swedish School of Economics). Thefirst paper, “Winners or losers? The effect ofbanking consolidation on corporateborrowers”, was presented by EmiliaBonaccorsi di Patti (Bank of Italy) (co-authorGiorgio Gobbi, Bank of Italy). The paper uses arich dataset on Italian firms and banks toinvestigate borrowing and lending behaviour.The paper investigates the impact of bankM&As from borrowers’ perspective, contraryto the prevalent approach in the literature thatconcentrates on the perspective of mergingbanks. The main question is whether bankM&As change borrowers’ credit availabilityand investment capability because of a loss ofinformation in relationship-based lending. Theauthors suggest that small firms with few bankrelations or high risk firms could be mostaffected because of the dissipation ofinformation in bank M&As and the followingrestructuring of the organisation. However, themain conclusion of the paper is negative: Noneof the different classes of borrowers areadversely affected – not even the smallest orweakest firms, or firms most dependent on fewbanks.

The second paper, “Gains in bank mergers:Evidence from the bond markets”, waspresented by Maria Penas (Vrije UniversiteitAmsterdam) (co-author Haluk Unal,University of Maryland). The paper documentssignificant effects of US bank M&As on theconcerned banks’ bond returns and spreads.Thus the authors conclude that banks aredifferent from non-financial firms as firms’spreads are typically not affected by M&As.The paper then analyses the determinants of thepositive bond returns and lower spreadsassociated with bank M&A-announcements.The main findings are that increasing risk

diversification increases returns and lowersspreads as default risk is reduced. The moststriking finding is that the incremental size-increase of the new banking organisations is asignificant factor. Notably, return increasesand spread reductions mainly occur formedium-sized banks that become sufficientlylarge through consolidation. If medium-sizedbanks grow large enough (and become too-big-to-fail), banks’ bond-holders will expect abailout by public authorities in case of troubleto be more likely and thus they will demand alower risk premium.

The last paper of the session was presented byPeik Granlund (Finnish FinancialSupervisory Authority), “Economic evaluationof bank exit regimes in the US, EU andJapanese financial centres”. The paperdocuments the legal frameworks for bankprotection and creditor rights in the differentfinancial centres and studies then the impact ofthese legal features on banks’ bond spreads.Banks’ cost of funds should be the lower thegreater creditor protection and the higher bankbailout probability, as the risk premium wouldbe lower. The paper also investigates theimpact of shareholder protection on bank assetgrowth and predicts a positive relationship.The main conclusion is that banks seem toenjoy lower spreads in more “protected”regimes. The evidence as regards asset growthis ambiguous.

Jukka Vesala (ECB) discussed the threepapers. He noted on the first paper that it wouldhelp to articulate the underlying welfareanalysis more clearly. In particular, the paperseems to focus on how producers’ surplus isaffected by bank M&A and what factorsdetermine how this is shared betweencustomers and the bank. But it is notimmediately clear how the focus on borrowersfits into this. In addition, it would beworthwhile analysing further whyconsolidation should be an important influenceon bank-customer relations. Finally, somesuggestions were made about using two-stageestimation techniques, inclusion of non-


performing loans, and robustness checksconcerning the measurement of “firm quality”.Regarding the second paper, Vesala warnedagainst interpreting too sharply theimplications for the too-big-to-fail problem.Under a full bailout expectations banks’spreads would be zero (or amount only to theliquidity premium), but in the sample, banks’have significantly positive spreads (on average90 basis points in the US) and the impact ofM&As is found at most at 16 basis points.Another major suggestion was that sincefinance theory predicts bond spreads to be non-linearly related to default probability, spreadsshould remain stable and close to zero for largeintervals and only react significantly relativelyclose to the default point. Against thisbackground, the relatively large spreadreductions found for targets could suggest thatthese banks were in a relatively weak condition(and therefore targets) and this could beusefully studied further. Comments on the lastpaper largely focused on the empiricalanalysis. Especially, the measurement offinancial assistance probability is quitejudgmental. Rating agencies’ “support ratings”(measuring exactly the bailout probability bypublic authorities or parent financialorganisations) could be used to increaseobjectivity. Moreover, the credit spread (andasset growth) analysis should preferably bebased on a larger sample of banks and oninstruments that are sufficiently liquid. Itshould also be extended to multivariateregressions (including bank-specific risk andother determinants, market-specific factors).

SESSION 1.2 EUROPEAN EQUITY MARKETSAND CORPORATE GOVERNANCE

The session was chaired by David Mayes(Bank of Finland). The first speaker YrjoKoskinen (Stockholm School of Economics),presented the paper “The euro is good after all:corporate evidence” (co-authored with ArturoBris, Yale School of Management and MatthiasNilsson, Stockholm Institute for FinancialResearch). The objective of the paper is todetermine the effects of the euro on corporate

behaviour since its introduction in 1997. Theanalysis relies on data at the firm level from1995 to 2000, from 10 EU countries thatadopted the Euro, 3 EU countries that did notjoin EMU, as well as Norway and Switzerland.The authors analyse firms’ valuation andinvestments in the sample. Large firms in EMUcountries increased in value, as measured byTobin’s Q, especially those in countries thatexperienced currency crises. In itself this is notenough to show the positive effect of the euroas investment opportunities could also haveimproved. However, the authors’ view is thatbetter investment opportunities were createdby the single market, much before the advent ofthe euro. Given the assumption that all firmsface similar investment opportunities, oneshould observe that firms with reduced cost ofcapital would invest relatively more. Indeed,the authors find that the introduction of theeuro had a positive effect on investments offirms in the euro area. This justifies the claimthat the cost of financing decreased due to theintroduction of the single currency. Howeverthe authors emphasised that the introduction ofthe single currency implied the adoption of asingle monetary policy. Some firms could beadversely affected by asymmetric inflationshocks among EMU countries, which areimpossible to overcome when a singlemonetary policy is used. This effect could havedriven the cost of capital up. In this regard,increased capital markets integration is key inorder to improve risk sharing opportunities inEurope. Hence, the authors reached theconclusion that the introduction of the euro haslowered firms’ cost of capital by furtherincreasing capital markets integration inEurope and by eliminating currency risksamong the countries that joined EMU.

Thomas Gehrig (University of Freiburg)presented the paper “Cross-listings and thegeography of firm’s ownership”, co-authoredwith Thierry Foucault (HEC). The objective ofthis theoretical paper is to propose anotherexplanation for the increased cross-listings ofEuropean stocks in the US. The authors arguethat cross-listing is a mechanism through

83ECB


ANNEX D

which market information is generated. Theyconsider a framework with two types ofshareholders: sophisticated shareholders whocan engage in multi-market trading andunsophisticated shareholders who exclusivelytrade in their domestic markets. Shareholdersmight have private information on the value ofa firm. A firm facing the option to go publicmight choose to cross-list or not, may choose toallocate equities between different class ofshareholders. Cross-listing creates marketsegmentation. This segmentation reducesmarket liquidity for a firm’s shares. But theauthors show that it also forces informedinvestors to trade larger quantities overall. As aconsequence, prices reveal more of theinformation possessed by informed investors.Therefore a firm faces a trade-off betweenliquidity and information. The benefits ofcross-listing are maximal if the allocation ofequities is judicious. In particular, firms mustallocate shares to unsophisticated investors inboth markets so as to create the informationeffect. An interesting consequence of theanalysis is that firms and exchanges would notnecessarily benefit from market integration. Asthe trade-off vanishes with marketsintegration, firms will loose the informationalbenefits of cross-listing. As a consequence,revenues of exchanges from listings willdiminish. Hence, firms and exchanges mightoppose further markets integration.

Mariassunta Giannetti (Stockholm School ofEconomics) presented the paper “Whichinvestors fear expropriation? Evidence fromstock picking”, co-authored with AndreiSimonov (Stockholm School of Economics).This paper investigates whether investors takefirms’ corporate governance into account intheir investment decision. To this end, theauthors use a data set that providescomprehensive information on almost allstockholders of companies listed on theSwedish stock market. They find that fears ofexpropriation in companies where theextraction of private benefits is expected to begreater, discourage investors who enjoy onlysecurity benefits from buying shares. Their

analysis relies on the ratio of control to cashflow rights of the principal shareholder. Thisratio is expected to be positively correlatedwith the extraction of private benefits by a firmand is therefore used as a proxy for badgovernance. The authors find that this ratiomatters, in the sense that the overall impact of amarginal change in the ratio for a specific firmon the probability that investors buy the stockof this firm is negative and significant. Theeffect of improved corporate governance isfound to be stronger for sophisticatedinvestors, like financial institution and foreigninvestors, while large domestic investors andindividuals who are board members, do notbase their investment decisions on corporategovernance grounds. Hence, the authors findthat corporate governance is an importantfactor in order to understand portfolio choicesacross countries.

Leo de Haan (Dutch Central Bank) acted asdiscussant. Regarding the Bris, Koskinen andNilsson paper, he questioned the result on theground that Tobin’s Q for non-EMU countriesdecreased substantially in the early 1990s, asthey had been hit by a severe currency crisis.He argued that in the aftermath of such a crisis,the authors crude control for this event mightnot take into account all of its consequences.He also called for further explanation on whythe difference between EMU and non-EMUcountries’ Q is only significant until 1997.Finally he questioned the finding that the euroeffect on Q appears in 1998 and is non-recurrent and wondered why this is the case.Regarding Foucault and Gehrig’s paper, deHaan asked whether listing fees and tradingrevenues could be set in such a way as to restoreinvestors’ incentives to promote financialintegration, instead of imposing regulation. Healso wondered how firms would do in practiceto allocate the proper amount of shares tounsophisticated investors. Finally, given therecent progress made in facilitating access toexchanges, he enquired about the future ofcross-listings should unsophisticatedshareholders disappear. Regarding Gianettiand Simonov’s paper, de Haan found it


surprising that the median investor holdsstocks from only a single firm. He questionedthe methodology on the ground that privateinvestors hold stocks mainly indirectly throughmutual funds, which are assimilated in thepaper to financial institutions. From the floor,it was asked whether the results of Bris,Koskinen and Nilsson’s paper are robust todifferent specifications of Tobin’s Q. On thisquestion, Koskinen replied that they did notperform any robustness checks yet. Also heclarified that the euro effect on Q is nottemporary. Rather this effect is permanent anddoes not revert after 1998. In his reply to deHaan, Gehrig expressed the opinion thatregulation is unlikely to provide the rightincentives as long as benefits from cross-listing are present. Although he reckoned thatthe absence of unsophisticated investors wouldchange the results, as this would eliminateany incentives of firms to cross-list. However,he pointed out that the assumption of havingsome investors tied to a specific market isreasonable. Finally, Gianetti pointed out thatmany empirical studies show that smallinvestors with brokerage account in the UShold one stock or a few stocks only.

KEYNOTE LECTURE ON FINANCIAL MARKETINTEGRATION, CORPORATE FINANCING ANDECONOMIC GROWTH

Jan Krahnen (CFS) introduced TullioJappelli (University of Salerno) who lecturedabout “Financial market integration,corporate financing and economic growth”.His talk was based on a report prepared by theCenter for Economic Policy Research for theEuropean Commission by himself, LuigiGuiso, Mario Padula and Marco Pagano. Theaim of the report is to determine the gains fromeconomic integration, the potential losers andwinners at the country and sector level. FirstJappelli reviewed the evidence on the linkbetween financial development and growth. Heconcluded that financial development of acountry benefits more industries with highdependence on external financing relative tothose using internal cash flow. Then he

exposed reasons why financial integrationmight promote financial development. Hehighlighted three main channels. Bank mergersand increased competition facilitate access tofunds. Regulation might be improved andfinancial integration might give laggardcountries access to financial markets.However, he pointed out that the full effect offinancial integration on financial developmentmight be limited by the lack of convergence ofinstitutions or of their operation. In particular,he emphasised that setting rules is of little helpunless they are well enforced. In addition, hestressed the difficulty for foreign banks topenetrate the local credit markets, which is theones that matter for most firms. In order toempirically determine the effect of financialintegration on growth, the authors use aninternational industry-level panel of 49countries covering the period 1981 to 1995.They regress the growth in value-added in agiven country and sector on three explanatoryvariables. 1) Financial development, definedalternatively as the combination of financialactivity (private credit over GDP) or the size ofthe financial sector (private credit plus marketcapitalisation), multiplied with an indicator offinancial dependence (the degree of externalfinancing for each sector). If all countries werefully integrated the effect of this variablewould be nil, because national (or local)financial development should not matter for thegrowth of national firms, whatever theirdependence on external finance. 2) A proxy forfinancial efficiency, measured by an index ofaccounting standards, is used as instrumentalvariable for financial development. 3) Somecountry-specific and sector-specific controlvariables and fixed effects. The authors findthat the coefficient of interaction betweenfinancial dependence and financialdevelopment is positive and stable over time.Furthermore, accounting standards matter forgrowth. Results are similar for the period after1991. Hence, Jappelli concluded that furtherpromotion of financial integration is warrantedas integration was only partial in the 1990s.Based on these estimates, Jappelli thenpresented the results of simulations of the

85ECB


ANNEX D

impact of financial integration in Europe.There, two alternatives definition ofintegration have been used. Financialintegration is either defined as a convergenceof European financial development to the levelof the US, or as convergence toward the level ofthe best country in Europe. In the first case, theimpact of financial integration is a 1 percentagepoint gain in EU GDP growth, while it is 0.75percentage points in the second case. Jappellistressed that these numbers are lower bounds,as the effect of financial integration on servicesectors was not considered. Furthermore hepointed out that countries that are alreadyfinancially developed will benefit less fromfurther integration, while in countries that areless financially developed, the financial sectorwill in all likelihood lose market shares andprofits.

In the discussion, Jappelli was asked to clarifywhether state subsidies were taken into accountin the measures of financial development orfinancial integration. Jappelli answered thatthe use of country and sector dummies shouldtake care of this issue. Juha Tarkka (Bank ofFinland) asked whether transportation costsand informational barriers were a limit tofinancial integration and wondered whetherthese barriers could be removed by regulation.Jappelli replied that some determinants offinancial development can be modified bypolicy makers. He reiterated that the maindifficulty resides in the implementation ofrules rather than in their adoption. Jan Krahnen(CFS) wondered if the fact that the US has afinancial system which relies more on financialmarkets relative to Europe can influence theresults. Jappelli answered that it is not clearwhether the composition between market-based and bank-based finance matters forgrowth. Rather, the overall size of financialmarkets seems to be more important. Finally,an alternative approach was suggested from thefloor, that would use fewer countries but longertime-series so as to verify that the results arenot driven by time-series effects. Jappellireckoned that the analysis would be usefulalthough difficult to carry out. For instance, he

mentioned the fact that a time-series analysiswould require time-dependent indicators forfactors related to financial development and tofinancial integration. Furthermore, analysingfewer countries would necessarily add noise tothe data.

SESSION 2.1 FINANCING STRUCTURE OFFIRMS; THEORY

Jan Krahnen (Centre for Financial Studies)chaired the session. The first paper of thesession, “Informational efficiency andliquidity premium as the determinants ofcapital structure” by Chung Chang (Universityof Minnesota) and Xiaoyun Yu (University ofIndiana), analysed the optimal capital structureof firms when the firm’s debt and equity istraded on secondary markets. A firm initiallydecides to finance a project by issuing debt andequity. Both instruments are traded lateramong two types of investors that choosewhether to trade in the market for equities orbonds: traders that use private – and unknownto the firm - information concerning the valueof the project and traders that are uninformedbut need to obtain liquidity. At this stage, thefirm uses information that is revealed throughprices to decide whether to liquidate or finalisethe project. Yu emphasised that the leveragedecision of a firm determines not only how wellprices will reflect private informationconcerning the project value, but also howmany liquidity traders are participating in theequity market. On the one hand, increasingleverage makes equity more sensitive toinformation and, hence, increases the profit ofinformed traders. On the other hand it reducesthe amount of liquidity trading in equitythereby lowering profits for informed traders.Based on this trade-off the authors derive thefollowing results. First, the optimal leverage issuch that debt is not any longer default-free, butnot so risky as to induce informed trading in thebond market. Hence, there is only a defaultpremium on debt when issued but no liquiditypremium that compensates bondholders forlosses due to informed trading. Second, sinceall informed trading takes place in the stock


market, at the optimal level of debt stock pricesreveal the maximum amount of information,while the liquidity premium, required toprotect against informed trading, is minimised.Finally, over-leveraged firms will not be ableto obtain information from securities pricesefficiently, and, hence, excessive debt caninvolve costs beyond the costs associated withfinancial distress.

The second speaker, Tuomas Takalo (Bank ofFinland) also emphasised the importance ofprivate information for optimal financialdecisions by firms in his paper “Equilibrium infinancial markets with adverse selection”(joint with Otto Toivanen, Helsinki School ofEconomics). The authors look at a standardfinancing problem with adverse selection.There are two types of entrepreneurs eitherhaving a project with positive or negative netreturn. When entrepreneurs have access tofunds from outside the economy, this unlimitedsupply of funds leads to low interest rates andinefficient pooling equilibria with over-investment in projects. The authors depart fromthis benchmark by allowing entrepreneurs toinvest their own funds. Entrepreneurs are thenalso investors in the sense that they face aninitial choice whether to use their wealth tofinance their own or somebody else’s project.The scarcity of funds leads to an increase in theinterest rate at which projects get financed. Forcertain parameter values, this increase in theinterest rate allows to screen for entrepreneurswith good projects and, hence, allows to re-establish efficient outcomes in form ofseparating equilibria. Within these equilibria,the mode of finance is equity if the averagequality of projects is high and debt otherwise.Finally, Takalo stressed that the model hassome counter-intuitive implications as far asincreases in wealth are concerned. First, whenwealth increases inefficient equilibria becomemore likely since the supply of funds isabundant. Second, while equity is only used tofinance good projects in wealth constrainedeconomies, increases in wealth make equity thechoice of finance for bad projects.

The final paper of the section, “Limitedenforcement and efficient inter-bankarrangements” by Thorsten Koeppl (ECB)and James MacGee (University of WesternOntario), puts forward limited enforcement ofcontracts as an explanation of financialarrangements rather than private information.Starting from some stylised empirical evidenceon co-operative arrangements between banksin contemporary Germany and the historic US,the authors try to assess whether banks canmutually insure against their asset risk even ifthere are no formal institutions to enforceinsurance transfers. In the model, banks facetwo types of risk. The liability side of a bank’sbalance sheet is affected by random withdrawaldemands from depositors. Furthermore, banksface default risk on their loan portfolios withthe aggregate size of default being stochastic.Koeppl first pointed out that it is the interactionbetween both types of risk that enables banks toovercome limited enforcement. In the absenceof liquidity risk, insurance against asset risk isnot feasible since – in the absence ofenforcement – a bank that is called upon tomake a transfer has no incentive to do so afterthe risk has materialised. When banks faceliquidity risk, however, the threat of beingexcluded from liquidity provision when failingto make a transfer is enough to enforce partialrisk sharing. The amount of risk sharing heredepends on the correlation between the two riskfactors with a negative correlation fosteringrisk sharing. In the second part of hispresentation, Koeppl demonstrated that acompetitive market for liquidity can notprovide this insurance. While the pricemechanism is able to channel liquidityefficiently between banks, it can never exploitthe correlation between the shocks to allow forself-enforcing insurance arrangementsbetween banks. Finally, Koeppl stressed theco-existence of liquidity provision and assetinsurance in the empirical examples. Whileinterpreting this fact as supporting evidence forthe paper, he emphasised, however, that thefindings of the paper go beyond the narrowfocus on inter-bank arrangements and applymore generally.

87ECB


ANNEX D

Leo Kaas (University of Vienna) discussed thepapers. He saw the strength of the paper by Yuin combining an optimal choice of financialinstruments with a friction arising from latertrade in secondary markets. He wondered,however, how robust the findings were withrespect to the informational assumptions madein the model. Traders can observe prices only inthe market where they trade, while the firm isable to observe both prices. Furthermore, whendebt or equity are issued traders are not yetinformed while they are potentially informedlater. Regarding the paper by Takalo, Kaasexpressed concerns about the interpretation ofthe model in light of financial development.According to him, financial market integration– via enhanced access to markets – lowersfinancing costs and should per se have positivewelfare effects in general. Kaas attributed thedetrimental effect of a wealth increase in spiteof lower financing costs to the particular modelof adverse selection used. For the paperpresented by Koeppl, Kaas appreciated thenovel approach used to model and analyseinteractions between the asset and liability sideof banks’ balance sheets. He pointed out,however, that – due to risk-neutrality and thetiming of the model – one should not use theterm “insurance”, but rather “ex-post trade”.Furthermore, he asked Koeppl whether theinefficiency of markets arouse only from thefact that banks are required to be solvent whentrading on the market for liquidity.

In her response Yu pointed out that modelinginformation acquisition – possibly in adynamic setting – could help to address thecomments, but was likely to increases thecomplexity of the model tremendously. WhenElena Carletti (University of Mannheim)remarked that the liquidation decision shouldnot be carried out by the firm, but rather byinvestors, Yu stressed that it would be difficultto introduce an agency problem within theexisting framework and that it was not clearhow results would change. Finally, Koepplexplained that the solvency constraint forbanks in the market for liquidity is preciselymodelled to mimic the enforcement problem of

co-operative arrangements between banks.This fact allowed for a sensible comparisonbetween different forms of liquidity provision.Thomas Gehrig (University of Freiburg)mentioned that the inefficiency of markets isseemingly driven by the fact that prices arerestricted to be linear and that more generalpricing structures are potentially able toimplement efficiency via competitive markets.

SESSION 2.2 INTEGRATION OF EQUITYMARKETS

The session was chaired by Harry Huizinga(Tilburg University and EuropeanCommission). The first speaker, MichaelHaliassos (University of Cyprus), presentedthe paper “Household stockholdings in Europe.Where do we stand and where do we go?” (co-authored with Luigi Guiso, University ofSassari and Tullio Jappelli, University ofSalerno). Haliassos started by documentinghow the 1990’s witnessed a significant increasein households’ stock market participation in allmajor European countries and in the UnitedStates. Despite this common trend, persistentdifferences across countries remain, with theU.S., the British and Swedish householdshaving considerable more participation thanFrench, German and Italian ones. The keycontribution of the paper is to explain howdifferences in stock market participation andstockholdings across countries can beattributed to households’ characteristics, suchas age, income, wealth and education. The mainfinding is that there is a positive correlationbetween the participation decision on the onehand and wealth and education on the otherhand. However, wealth and education haveonly a small effect on the portfolio shares ofstocks, among those households who doparticipate in the stock market. The authorsargue that different participation rates areconsistent with the international pattern ofentry costs. More specifically, countries withlower transaction and information costs arethose with higher participation rates. Inaddition, since the lowering of such costsbrings into the markets less sophisticated


investors, the authors argue that this couldinduce greater volatility in stock markets. Forexample by reacting excessively to marketsignals because of limited sophistication orbecause of limited ability to withstandfinancial pressure. This should prompt adiscussion about policies that could mitigatethese concerns, such as improving the access toaccurate financial information and ensuringsufficient financial education to newcomers.

Eric Theissen (University of Bonn) presentedthe paper “Competition between exchanges:Euronext versus Xetra”, co-authored withMaria Kasch-Haroutounian. The motivation forthe paper is that many exchanges in Europe maysoon face the decision to join one of the twodominating continental European tradingsystems, Euronext and Xetra. Euronext is acommon trading platform that originated fromthe merger of exchanges in Amsterdam,Brussels, Paris and (in 2002) Lisbon. Xetra,instead, is the electronic trading system of theDeutsche Börse AG, recently adopted byAustria and Ireland. It is plausible to expect thata factor leading to further consolidation ofexchanges in Europe should be the quality ofthese markets, the more efficient ones having anadvantage over the others. The paper analysesone of these factors, namely the execution costsin Xetra and Euronext. The comparison is madewith a sample of 40 pairs of stocks, matched bymarket capitalisation, trading volume andvolatility. Each pair consists of one Frenchstock traded on Euronext Paris and one Germanstock traded in Xetra. The main finding of thepaper is that, although there are no significantdifferences in quoted spreads, effective spreadsand their components as well as the realisedspread and the adverse selection componentare lower in Germany. Neither differences inthe number of liquidity provision agreements(i.e. the existence of sponsors or liquidityproviders for stocks), nor differences in theminimum tick size can explain the higherexecution costs in Euronext. The authorsconclude that investors in Euronext are lessprotected against informed traders, and thatEuronext offers lower operational efficiency.

Miguel Almeida Ferreira (ISCTE School ofBusiness) presented the paper “The importanceof industry and country effects in the EMUequity markets”, co-authored with MiguelAngelo Ferreira (ISCTE School of Business).This paper uses the methodology developed byHeston and Rouwenhorst to test the relativeimportance of country effects versus industryeffects in the euro area. This is relevant forfinancial integration, because when previouslysegmented markets start to integrate, gainsfrom diversification through a country-basedapproach should decrease with respect to thoseone could obtain through a sector-basedapproach. Using a sample of 10 industryindices and 11 EMU countries, the authors findthat country effects still dominate overindustry effects. However, although over thewhole sample country effects have beenrelatively more important in determiningequity returns, the evolution of the national andglobal effects through time reveals anincreasing relative importance of industryeffects. In particular, in the last sub-sampleperiod 1999-2001, industry effects arebecoming similar in size to country effects. Acomparison with 5 non-EMU Europeancountries shows that the increasing importanceof industrial influences is not limited to EMUcountries.

The discussant Simone Manganelli (ECB)stressed how the three contributions of thissession could be viewed as different pieces ofevidence of ongoing financial integration in theeuro area. The paper by Haliassos – by lookingat households’ portfolio composition – showsthat country dummies still explain households’decisions, after controlling for their age,income, wealth and education. Manganellisuggested to extend the time series of the dataset, and to look at how these results evolve overtime. Regarding the paper by Theissen, hecriticised the use of number of traded shares asa proxy for traded volume, which issubsequently used as a matching criterion forFrench and German stocks. According to thisproxy, the same company would double itstrading volume after a 1:2 stock split. This is

89ECB


ANNEX D

obviously inadequate and casts some doubts onthe reliability of the results. The results of thepaper by Ferreira are difficult to interpret in thelight of financial integration, becauserelationships in equity markets are highly time-varying and therefore it is difficult todisentangle cyclical movements fromstructural changes. Philipp Hartmann (ECB)asked whether the spreads in Theissen’s paperwere computed using daily or intra-day data, asdifferent definitions might produce verydifferent results. Theissen clarified that thedaily spreads are computed as averages ofintra-day spreads. He admitted the limitation ofthe number of traded shares as a proxy fortrading volume. He speculated however thatthis is unlikely to affect the results because thedescriptive statistics show that there are nosignificant differences among the differentgroups of stocks. Haliassos replied that the dataavailable for the United States as of 2001 showa slight increase in stock market participation,thus confirming the trend outlined in the paper.


KEYNOTE LECTURE ON INTERNATIONALFINANCIAL LINKAGES

Philipp Hartmann (ECB) introduced GeertRouwenhorst (Yale School of Management)who gave a key lecture on “Internationalfinancial linkages”. The main objective of histalk was to illustrate the effects of globalisationand integration on financial linkages. He firstrecalled the common wisdom that globalisationand integration of markets accentuate financiallinkages. As business cycles are more likely tobe synchronised, policies and institutions willbe more co-ordinated so that asset prices willmove together. Furthermore, integrationshould go together with a decrease in the “homebias” of investors and the globalisation offirms. All these effects will accentuatefinancial linkages. However, other factors thatare due to further integration are likely totemper financial linkages. Among others,Rouwenhorst mentioned the greaterspecialisation of countries which might favoursector-related business cycles. In addition,globalisation expands the set of investmentopportunities, which might allow greaterdiversification and insurance. Also,competition among exchanges might fostergreater efficiency. Overall, he concluded thatthe net effect of greater globalisation andintegration on financial linkages is unknown.

Next Rouwenhorst presented empiricalevidence on the evolution of financial linkages.He first presented a time-series plot of theaverage correlation of four markets (France,Germany, UK and US – the “core countries”)from 1870 to 2000, which shows that there is ageneral increase in correlation, but that thecorrelation is generally very low (below 0.6).He explained that this could not be the result ofperfectly diversified markets, as thecorrelation between two diversified portfoliosis around 0.98. Furthermore using pair-wisecorrelation over time, he showed that theaverage correlation is generally higher inperiods where markets are integrated thanwhen they are segmented.

DAY 2Rouwenhorst then addressed the role ofemerging markets in financial linkages. Heshowed that emerging markets have twoeffects. First, they lower the overall correlationbetween markets, as they are typically lesscorrelated to existing non-emerging markets.Second, they increase the investmentopportunity set for investors. The overall effectof emerging markets is an increase ofdiversification opportunities, as measured bythe ratio of the market portfolio volatility overthe average market volatility. How docorrelations change as a consequence ofglobalisation? Rouwenhorst presentedevidence that the country effect is still verylarge. In a globally integrated market, firms’location should not matter, i.e. from prices andreturns it should be impossible to tell a firms’location. However, where firms are locatedseems to still matter more than what theyactually produce, although there is evidencethat, in Europe, industry effects are gainingsome importance. Rouwenhorst concluded histalk by saying that, although the world is notyet fully globalised, there is some ground tobelieve that international linkages arebecoming stronger. He also pointed out that theexpansion of the opportunity set should givesome compensation for investor who seekdiversification.

In the general discussion, Rouwenhorst wasasked whether the “home bias” of investorswill decrease in the long run. Rouwenhorstanswered that the home bias was extremebecause of restrictions imposed on financialinstitutions. He added that even if thesebarriers were removed, he expected the homebias to persist, as he suspects behaviouralbiases can never be overcome. Helene Rey(Princeton University) pointed out that onedifficulty of sharing risks further is that it isdifficult to create new markets, such as marketsfor bonds indexed on GDP. She then asked whymarkets were so difficult to create.Rouwenhorst replied that a successful marketmust have demand and supply. Hence, thewillingness of an institution to create a market

91ECB


ANNEX D

is not sufficient, as it must generate enoughtrades for the market to survive. Jan Krahnen(CFS) questioned the assumption that the “corecountries” in 1870 are identical to the ones in2000 and asked whether changing thedefinition of “core countries” would affect theresult. Rouwenhorst acknowledged that thiswould be a concern if average returns wereanalysed. However, since average correlationsare looked at, the selection bias is not a concernhere. Hartmann remarked that Europeangovernment bond holdings by non residentshad increased by 10-20 percentage points in thepast, thus suggesting a decrease in the “homebias” of investors.

SESSION 3 INTERNATIONAL PORTFOLIOCHOICE AND ASSET MARKETLINKAGES

The session was chaired by Philipp Hartmann(ECB). The first paper, “Exchange rates,equity prices and capital flows” by Harald Hau(INSEAD) and Helene Rey (PrincetonUniversity), focused on explaining exchangerate movements with macro fundamentals. Themain objective of the paper is to develop amodel that can bridge the gap between foreignexchange market microstructure andmacroeconomic fundamentals. By developingan equilibrium model in which exchange rates,stock prices and capital flows are jointlydetermined, the authors are able to derive anumber of testable implications for the jointdynamics of these variables, which explicitlydepend on assumptions about traders’ degree ofrisk sharing. In the most interesting case ofincomplete risk sharing, the model implies ahigh level of exchange rate return volatility(albeit lower than equity return volatility), anegative correlation between foreign equityreturns in local currency and the value of theforeign currency, and a positive correlationbetween the value of a foreign currency and thenet equity flows into that market. Interestingly,using data for OECD countries relative to theUS, the authors find these implications of themodel to be supported by the data.

Michael Ehrmann (ECB), presented thesecond paper, “Interdependence between theeuro area and the US: what role for EMU?”,which is co-authored with Marcel Fratzscher(ECB). In this paper, the authors study thedegree of interdependence between the US andthe euro area from the viewpoint of financialmarket participants, by examining the effectsof monetary policy announcements andmacroeconomic news on interest rates in themoney markets. Using pre-EMU German dataas well as euro interest rate data following theintroduction of the single currency, the authorsfind that the interdependence of money marketshas steadily increased over time. In addition,there is stronger evidence of spill-over effectsfrom the US to the euro area than in theopposite direction. However, the authors alsofind that euro area news have becomeincreasingly important for the euro moneymarket, suggesting that markets haveundergone a learning process about themonetary policy of the ECB.

Charlotte Ostergaard (Norwegian School ofManagement) presented the third paper of thesession, “International diversification in bankasset portfolios” (joint with Claudia Buch, KielInstitute for World Economics, and JohnDriscoll, Board of Governors of the FederalReserve System). This paper analysed theinternationalisation of the banking industry,focusing in particular on direct cross-borderlending, i.e. “international banking”. Using astandard mean-variance portfolio model, theauthors estimate benchmark portfolio weights,which they subsequently compare to the actual“portfolios” of banks’ foreign and domesticassets. Based on data from four majorindustrialised economies, the authors find thatbanks in general tend to over-investdomestically, relative to the benchmark,showing that there is a similar home bias forbank credit, as there is one for stockinvestments. In investigating the possiblereasons for this type of domestic over-investment, they find that proxies forinformation costs associated with foreign


investments did not seem to be important.Instead, credit risk considerations appeared toplay a more important role, in that banks tend toover-invest more in markets characterised byhigh credit quality and also to invest even moreif credit conditions further improve.

The discussant of this session was YigalNewman (Stanford Graduate School ofBusiness). Newman first highlighted a verystrong feature of Hau and Rey’s paper: the factthat it produced a number of clearly testableimplications, which turned out to be confirmedby the data. He also noted that the implicationsof the model and the empirical results related tothe correlation between equity returns andexchange rate movements run counter to“conventional wisdom”. He therefore urged theauthors to be more clear in the paper inexplaining the mechanisms in the model thatproduce this result, and to also provide bettereconomic intuition for these findings.Commenting on the paper by Ehrmannand Fratzscher, Newman suggested that intheir investigation of money marketsinterdependence prior to EMU, more Europeanmarkets could be examined than only theGerman one. Furthermore, he also asked whatrole the relative size of markets may play inexplaining interdependence between markets.Finally, on the paper presented by Buch,Newman noted that a nice feature of the paperwas the unique data set on cross-border claimsof banks in four major industrialised countries.On the other hand, he questioned whether thisdata set could be seen as fully representative,since it excludes a number of markets (notablyemerging markets) that may representimportant investment opportunities for banks.The discussant also suggested that the authorsshould discuss the way banks choose to hedgeor not to hedge their international exposures, aswell as the reasons underlying these decisions.

In the general discussion, Philipp Hartmann(ECB) argued that developments happeningaround the time of the conference seem toprovide further evidence in favour of thearguments in Hau and Rey paper. In particular,

one could observe rising US stock markets witha depreciating dollar and declining Europeanstock markets with an appreciating euro.


Juha Tarkka (Bank of Finland) chaired theconcluding panel discussion of the workshop.Participants were invited to give their viewsabout the future of exchanges: competition orconsolidation?

Niall Bohan (European Commission)presented channels through which competitionis changing the economics of trade execution.He first concentrated on competition betweenexchanges and multilateral trading systems. Herecalled that these arrangements are the basisfor the price formation mechanism. Thedimensions along which exchanges cancompete with each other are many, rangingfrom the initial listings fees, the increase ofliquidity on platforms thus reducing the bid-ask spread, the reduction of trade executioncosts, to the vertical integration of exchangeswith clearing and/or settlement systems. In hisview, it is impossible to predict howcompetition will shape the future of theEuropean trading landscape, although hereckoned that consolidation is inevitable.However, he stressed that from the standpointof regulatory authorities, one should avoidfavouring a preferred market structure whilestill being concerned that the consolidationprocess and outcome are not distorted by anti-competitive strategies or behaviours thatwould undermine the efficiency of the price-formation process.

Then Bohan discussed the issue of tradeinternalisation by banks. He recalled that 22-25% of client orders of large houses are nowinternalised. This increased internalisation wasthe result of the steady concentration of clientbrokerage accounts over the course of the1990s combined with increased netting orsettlement capacity within large banks.Internalised trading differs from exchange-based trading as the system operator

93ECB


ANNEX D

determines the conditions of trades andwhether they are executed internally. Thisfeature implies that internalised trades do notface the same regulation as “order book”trades, while still competing with exchanges.Bohan recalled that concerns were expressedon the possibility that internalised tradingwould siphon so much liquidity away fromexchanges that the price mechanism wouldsuffer. He admitted that there is little evidence,as aggregate internalised order flows or thetypes of orders that are internalised areunknown. Therefore he concluded that theregulatory authority should concentrate on therisks pertaining to the combinations ofdifferent systems in the executions of trades,and ensure that trades are internalised onlywhere there is benefits for the clients. In thisway, the integrity and efficient price formationof the overall market system can be preserved.

Andre Went (Euronext) then described thebusiness model chosen by Euronext, horizontalintegration and vertical efficiency. Euronextoffers local access to members and listedcompanies, where markets are regulatedaccording to local jurisdiction. However, inconjunction to this local presence, there is asingle central order book with single marketrules. Among the benefits offered by thissystem are the international exposure andliquidity for listed companies and costreductions for members. Members appear torationalise progressively their access bymaking use of their direct access, whileinvestors pay fees on cross-border transactionsthat are close to the ones charged for domestictransactions. The system also exploitseconomies of scale and synergies by removingthe duplication of systems such as the numberof clearing platforms. Went asserted that theconsolidation process will continue, as crossborder activities are bound to increase andcompetition puts pressure on fees. Heconcluded that ultimately further consolidationis beneficial for shareholders and users, as itwill enhance liquidity and cut cost.

Peter Gromber (Deutsche Boerse)concentrated his talk on two main topics. Firsthe asked about the efficient level ofconsolidation. He observed that exchanges andsettlement systems in Europe are veryintegrated: the four big exchanges (DeutscheBoerse, Euronext, London Stock Exchange andNordic) represent approximately 78% of thetotal market. In his view, the next step forconsolidation will soon occur and be headed byone or more of the four exchanges. Costefficiency represents the main driving force ofconsolidation. The growth in transactionsdrives unit costs down and significantlyreduces the implicit trading costs for marketparticipants. Hence, Gromber concluded thatconsolidation will continue. He then addressedthe issue of internalisation of trades and itsregulation. He called for implementing aregulation that would protect investors’ rightsand establishes a high level of marketefficiency. In particular he suggested quotinginternalised trades in parallel in the open orderbook as a solution to the problems created byinternalisation.

Hannu Haltunen (Nordea) began hispresentation by describing Nordea. It is thelargest bank covering the four Nordic countriesand resulted from the merger of four nordiccountry banks. Linking these four nordic banksare four separate but inter-linked settlementsystems with links also to European entities.Haltunen then exposed his views onconsolidation. He predicted that furthereconomies of scale can be realised as well as ahigher level of liquidity. However, he askedthree issues related to consolidation. 1) Shouldconsolidation be vertical or horizontal?2) What should the governance structure be,for- or non-profit? 3) Should access beexclusive or open?

Tarkka then invited the panel participants todiscuss the different views expressed. Bohanreiterated that determining the risks ofinternalisation should be a priority. Went saidthat the current draft of the Investment ServicesDirective (ISD) is covering most of the issues


highlighted so far. He further claimed that thedebate on vertical versus horizontal integrationis secondary, and that corporate governanceissues are less relevant as long as services areprovided efficiently to customers. Gomber alsonoted that the debate on vertical versushorizontal integration had calmed down.Regarding internalisation, he emphasised thatit is key to find a solution that will preserve theefficiency and integrity of the price formationmechanism. Haltunen said that furtherconsolidation would entail high short termcosts that might temper the willingness of someinstitutions to actually consolidate. In spite ofthese costs, he expressed the view thatconsolidation will continue, although the paceand the new drivers of further consolidation areuncertain. He finally asked Bohan whether,from the point of view of customers, theregulatory authorities were taking a stance onthis issue. Bohan replied that on the tradingside, the regulatory environment is takingshape. However he saw a need to improve theclearing and settlement sides, where vestedinterests are very present. He called for furtheractions from the public authorities to removebarriers for cross-border clearing andsettlement, and to closely monitor the risksinherent to clearing and settlements systems.Tarkka then asked whether regulatory barrierswere of importance for further consolidation.Went replied affirmatively and Gomberhighlighted the difference of regulatorytreatment between types of institutionsregarding, for instance, liquidity provision.Philipp Hartmann (ECB) asked whetherinternalisation can be seen as an indirect way toreshuffle liquidity pools among exchanges,with the aim of fostering consolidation. Wentsaid that the ISD still allows for internalisationand expressed concerns that the priceformation process might be damaged if pre-trade transparency is not required. He finallyclaimed that pre-trade transparency iscompatible with stronger competition, so thatthis should not be a concern when adopting pre-trade transparency.

CLOSING REMARKS

Vítor Gaspar (ECB) closed the workshop byfirst thanking Sinikka Salo, Juhha Tarka andtheir team at the Bank of Finland for theirorganisation. He then reviewed the activities ofthe ECB-CFS Research Network so far,starting from the launching workshop whichtook place in April 2002 in Frankfurt, andlooking at where the network stands and whereit will go. He recalled that a main output ofthe launching workshop was the “roadmap”, adocument describing the scope of the networkand highlighting some areas that should receiveprimary attention over the first two years of it.He recalled that the chosen areas were (1) bankcompetition and the geographical scope ofbanking activities; (2) international portfoliochoices and asset market linkages betweenEurope, the United States and Japan; (3)European bond markets; (4) Europeansecurities settlement systems; and (5) theemergence and evolution of new markets inEurope (in particular start-up financingmarkets). The launching workshop coveredparts of areas (2) and (3), the present workshopaddressed areas (1) and (2). Finally, theforthcoming workshop at the Bank of Greece inAthens in November 2003 will resume on topic(3) and also cover areas (4) and (5). Heannounced that the current phase of thenetwork will be concluded by a conference inFrankfurt, in the late spring of 2004. He thendescribed some steps taken after the launchingworkshop. To further the interaction in thenetwork and to facilitate the flow ofinformation on the topics of interest to networkparticipants, a web-site was made availableat http://www.eu-financial-system.org. Ofspecial interest is an online library of someseminal published papers and many recentworking papers in the areas of the network.Finally he mentioned the creation of theLamfalussy Fellowship by the EuropeanCentral Bank to stimulate further research onthe five priority areas of the network.

95ECB


ANNEX E

Day 1 (Thursday November 20, 2003)

9:30-10:00 Welcoming Coffee

10:00-10:15 Nicholas Tsaveas (Bank of Greece) Opening Remarks

10:15-11:00 Tour of the Exhibition Celebrating the 75th Anniversary of the Bank of Greece

11:00-13:00 Parallel sessionsSession 1.1: Government bond market microstructures; liquidity andspilloversChair: Eli Remolona (Bank for International Settlements)

Avanidhar Subramahnyam (UCLA), An Empirical Analysis of Stock and BondMarket Liquidity (with Goizueta Taurn Chordia, Emory University and ArsaniSarkar, FRB New York)

Yui Chung Cheung (University of Amsterdam), Trading European SovereignBonds: The Microstructure of the MTS Trading Platforms (with Frank De Jong,University of Amsterdam and Barbara Rindi, Bocconi University)

Michael J. Fleming (Federal Reserve Bank of New York), Heat Waves, MeteorShowers, and Trading Volume: An Analysis of Volatility Spillovers in the U.S.Treasury Market (with Jose A. Lopez, Federal Reserve Bank of San Francisco)

12:30-12:45 Discussant: Albert Menkveld (Vrije Universiteit Amsterdam)


Session 1.2: Topics in the governance and integration of european financialmarketsChair: Charles Kahn (University of Illinois)

Mariassunta Giannetti (Stockholm School of Economics), Investor Protectionand Equity-holdings: An Explanation of Two Puzzles? (with Yrjo Koskinen,Stockholm School of Economics)

Leo Kaas (University of Vienna) Financial Market Integration and LoanCompetition: When is Entry Deregulation Socially Beneficial?

Cyril Monnet (European Central Bank) Guess What: It’s the Settlements! (withThor Koeppl, European Central Bank)

12:30-12:45 Discussant: Hans Degryse (University of Leuven and CentER)


ANNEX E PROGRAM AND SUMMARY OF THETHIRD WORKSHOPHOSTED BY THE BANK OF GREECE IN ATHENS,NOVEMBER 20-21, 2003


13:00-14:30 LunchP. Thomopoulos (Deputy Governor of the Bank of Greece)Lunch Speech on Financial Liberalisation, The Greek Experience

14:30-15:15 Marco Da Rin (University of Turin, ECGI, and IGIER)Key Lecture on European Venture Capital

15:15-15:30 Discussion


16:00-18:00 Parallel sessionsSession 2.1: Corporate bond financing and the cost of capitalChair: Joseph Bisignano (Bank for International Settlements)

John J. Puthenpurackal (Ohio University), Security Fungibility and the Costof Capital: Evidence from Global Bonds (with Darius P. Miller, IndianaUniversity)

Yigal Newman (Stanford University), Illiquidity Spillovers:Theory and Evidence from European Telecom Bond Issuance (with MichaelRierson, CitiGroup)

João Santos (Federal Reserve Bank of New York), Why Firm Access to the BondMarket Differs Over the Business Cycle: A Theory and Some Evidence

17:30-17:45 Discussant: Arnaud Mares (European Central Bank)


Session 2.2: IPOs in new marketsChair: Vicente Pons (Yale University)

Armin Schwienbacher (University of Amsterdam) Liquidity of Exit Marketsand Venture Capital Finance (with Grant Fleming, Australian NationalUniversity and Douglas Cumming, University of Alberta)

Tereza Tykvova (ZEW) Are IPOs of Different VCs Different? (with Uwe Walz,University of Frankfurt and CFS)

Vicente Pons (Yale University), Who Benefits from IPO Underpricing?Evidence from Hybrid Bookbuilding Offerings

17:30-17:45 Discussant: Wolfgang Bessler (Giessen University)


97ECB


ANNEX E

Day 2 (Friday November 21, 2003)

8:30-9 :00 Coffee

9:00-11:00 Parallel sessionsSession 3.1: Securities settlement systemsChair: Vítor Gaspar (European Central Bank)

Jens Tapking (European Central Bank), Raising Rival’s Costs in the SecuritiesSettlement Industry (with Cornelia Holthausen, European Central Bank)

Karlo Kauko (Bank of Finland), Interlinking Securities Settlement Systems:A Strategic Commitment?

Heiko Schmiedel (European Central Bank), Economies of Scale andTechnological Developments in Securities Depository and Settlement Systems(with Markku Malkamaki, Bank of Finland and Juha Tarkka, Bank of Finland)

10:30-10:45 Discussant: Charles Kahn (University of Illinois)


Session 3.2: The determinants of VC invesmentsChair: Jan-Pieter Krahnen (CFS)

Douglas Cumming (University of Alberta), The Legal Road to ReplicatingSilicon Valley (with John Armour, University of Cambridge)

Tuomas Takalo (Bank of Finland), Investor Protection and Business Creation(with Ari Hyytinen, University of California, Berkeley)

Giovanna Nicodano (University of Turin), What Drives the Structure ofPrivate Equity Investment? (with Marco Da Rin, University of Turin, ECGI, andIGIER and Alessandro Sembenello, University of Turin)

10:30-10:45 Discussant: Heather Gibson (Bank of Greece)




11:30-13:00 Policy Panel: European securities settlement systemsModerator: Alberto Giovanninni (Unifortune)

Kenneth Garbade (Federal Reserve Bank of New York)

Randy Kroszner (University of Chicago)

Joël Mérère (CEO, Euroclear France)

Gertrude Tumpel-Gugerell (Member of the Executive Board of the EuropeanCentral Bank)

13:00-13:15 Concluding remarks

13:15-14:15 Lunch

99ECB


ANNEX E


The third workshop of the network was hostedby the Bank of Greece in Athens on 20-21November 2003. The main priority areasanalysed in the course of the workshop were(3) European bond markets; (4) Europeansecurities settlement systems; and (5) Start-upfinancing and new markets. The workshopcombined research key lectures, research paperpresentations, and a plenary panel discussionon “European Securities Settlement Systems”,that included Kenneth D. Garbade (FederalReserve Bank of New York), Randy Kroszner(University of Chicago), Anso Thirè(Euroclear France), and Gertrude Tumpel-Gugerell (Member of the Executive Board ofthe European Central Bank). This documentsummarises the third workshop of the Network.

I N T RODUC T I ON


Nicholas Tsaveas (Bank of Greece) welcomedworkshop participants with some openingremarks on the three main topics of theworkshop, (1) venture capital finance, (2)securities settlement systems and (3) Europeanbond markets. Regarding venture capital heacknowledged appropriate timing of theworkshop, since it corresponds with the closureof the EU’s Risk and Capital Action Plan thataims at promoting precisely this form offinancing. In his views, Europe needs venturecapital firms for at least two reasons. First, theapplication of the Basle II accord will erodebank’s appetite for high risky high returnsinvestments. Second, with the forthcomingboom in retirements, pension funds will alsobecome less and less willing to invest in thesetypes of investments. To promote venturecapital, he asked whether Europe should aimfor a pan-European market to list new firms orwhether venture capital is better served bynational or regional stock exchanges.

Regarding securities clearing and settlementsystems, Tsaveas wondered about the best formof integration in this industry, calling for aform of organisation that would leave somespace for competition and provide sufficientincentives for service enhancements and lowercosts. To assess alternative models, heproposed that concepts of interoperability andharmonisation of procedures of the existinginfrastructures should be the guidingprinciples. He finally warned about thepotential danger of reaching too high a level ofconcentration, which although promotingefficiency could also cause systemic riskproblems, and called for adequate supervisoryprocedures.

Tsaveas then expressed his views on theEuropean bond markets. He first recalled thatbond markets are crucial for central banks’monetary policy operations. However hepointed out to some shortcomings of currentbond markets. First, bond markets are veryclose substitutes for bank lending when thingsgo well: they enhance liquidity and improve themonitoring of financial institutions by

allowing to “mark to market” their assets.However, when things go badly, it may becomeextremely difficult to raise funds in the bondmarkets. Second, with new bond derivatives, itis increasingly difficult to tell where bonds endand equities start. This of course createsdifficulties for investors to pin down the riskinvolved in each investment. Following thesefew remarks and before starting the workshop’ssession, Tsaveas invited workshop participantsto a tour of the exhibition for the 75th

anniversary of the Bank of Greece.

SESSION 1.1 GOVERNMENT BOND MARKETMICROSTRUCTURES; LIQUIDITYAND SPILLOVERS

The session was chaired by Eli Remolona(Bank for International Settlements). Thefirst speaker Avanidhar Subramahnyam(UCLA), presented the paper “An empiricalanalysis of stock and bond market liquidity”(co-authored with Goizueta Tarun Chordia,Emory University and Arsani Sarkar, FederalReserve Bank of New York). Recent financialcrisis suggest that in difficult marketconditions, liquidity can decline or disappear.To understand the determinants of marketliquidity, the authors consider the jointdynamics of liquidity, returns, and volatility instock and Treasury bond markets. Theobjective of the paper is to identify primitivefactors that induce correlated movements inliquidity, such as the monetary stance of theFed or the flow of funds into stock and bondmarkets. The authors use data from GovPX andTAQ/ISSM in a 7 years window (1991-1998) tomeasure quoted spreads, depths and orderflows in each of the two markets. To analyse thedata, they perform a vector autoregression withliquidity, order flow, returns and volatility.They find that stock and bond liquidity possesssimilarities such as common calendarregularities. Furthermore, there are cross-market dynamics flowing from volatility toliquidity. The authors also report that liquidityand volatility shocks are positively andsignificantly correlated across stock and bondmarkets at daily horizons, indicating that

DAY 1

101ECB


ANNEX E

liquidity and volatility shocks are often of asystematic nature. Finally, the authors find thatan unexpected loosening of monetary policy, asmeasured by a decrease in net borrowedreserves, is associated with a contemporaneousincrease in stock market liquidity.

Yui Chung Cheung (University of Amsterdam)presented the paper “Trading Europeansovereign bonds: The microstructure ofthe MTS trading platforms”, with Frank de Jong(University of Amsterdam) and BarbaraRindi (Bocconi University). This paperinvestigates the microstructure of the MTSGlobal Market System, the most importantEuropean interdealer sovereign bond tradingsystems. The institutional environment of the market for sovereign bonds can be dividedin 2 sectors. The primary sector decides onthe finance policy based upon the fundingrequirement of each government. This sectoracts as the ultimate provider of liquidity. Thesecondary market decides on the tradingenvironment: it determines the structure ofpayments and settlements and the tradingfacilities offered by brokers and market makers.The MTS system operates on the secondarymarkets. The MTS platform is constituted oflocal platforms in each country member of MTSand a pan-european platform, EuroMTS.Participants in the MTS trading platform aremainly investment banks, with a large variety ofbanks. All trades are anonymous. The authorsused a detailed data set covering everytransaction of Italian, French, German andBelgian government bonds being traded on theMTS platforms from January 2001 until May2002. They find that the spreads are smallest forthe most actively traded issues. The MTSdomestic trading platforms offer slightly betterspreads than EuroMTS. The authors alsoanalyse the price impact of trades and tradingduration (the time elapsed between two trades).They find that trading intensity plays a role: ahigher price impact of trades after longdurations, and lower price impacts when tradingactivity is high.

The third paper of the session “Heat waves,meteor showers and trading volume: ananalysis of volatility spillovers in the U.S.Treasury Market” by Michael J. Fleming(Federal Reserve Bank of New York) and JoseA. Lopez (Federal Reserve Bank of SanFrancisco) analysed the possibility for heatwaves and meteor showers across London, NewYork and Tokyo’s U.S. Treasury markets. Heatwaves’ refer to the fact that volatility has onlylocation specific autocorrelation, while meteorshowers’ refer to a situation when volatilityspills over from one centre to another. Theauthors test for volatility spillovers in the U.S.Treasury Markets using yields data fromGovPX, from a sample period running fromMarch 1992 to August 1994. They calculate thechange in intraday yield for each trading centreand use a GARCH model to test for some formof heteroskedasticity within centres. Assumingno conditional mean dynamics in yield changeand applying the Engel, Ito and Lin (1990)method, they examine whether informationfrom other trading centres impacts intra-marketvariance and test whether lagged volume helpsexplain volatility. They find that pricediscovery occurs in all three trading centres.Also, volatility persists in each of them. Theyfind evidence that the meteor showerhypothesis holds for Tokyo and London whileNew York is better characterised by heat wavehypothesis.

Albert Menkveld (Vrije UniversiteitAmsterdam), 2003 ECB Lamfalussy fellow,acted as discussant and started by pointing outthat there is much to learn from themicrostructure of government bond markets.He questioned Fleming and Lopez’sassumption that prices reflect efficient pricingin all trading locations on the basis thatLondon’s and Tokyo’s market are much thinnerthan New York’s. He encouraged the authorsto decompose prices to take into accounttemporary effects on London’s and Tokyo’smarkets. On the paper of Chordia, Sarkar andSubramahnyam, he remarked that the use oforder imbalance (dollar value of buys minusdollar value of sells) scaled by the total value of


buys and sells is uncommon in microstructure.He encouraged instead the author to use theabsolute value of imbalance as a driver ofliquidity. Finally regarding the paper byCheung, de Jong and Rindi, Menkveld waspuzzled by the fact that high intensity oftrading is associated with a smaller priceimpact and larger persistence in order flow,while the opposite is true in equity markets.Menkveld suggested to take a more integrativeapproach to the study of government bondmarkets in Europe by decomposing nationalyield changes into a European benchmark yieldlevel, yield spread for the various countriesrelative to the benchmark and countriesspecific innovations. The study of the effect ofcountries’ order imbalance on yields wouldthen be possible.

Subramahnyam replied that scaling imbalancestake care of trading volume effects. Also, heacknowledged the collinearity of absoluteimbalance and volatility and found similarresults when replacing volatility for absoluteimbalance. From the floor, Subramahnyam wasquestioned on alternative measures of liquiditysuch as yield maturity differential between onthe run and off the run bonds, and how thesewould relate to the bid-ask spread. He repliedthat although a very valid and insightfulcomment, the problem would be to find ananalogue in the stock market. GiovannaNicodano (University of Turin) criticisedSubramahnyam’s paper on the ground that adriving force of liquidity is information andalthough the authors exploit innovationsin returns in their paper, the analysis stillremains incomplete as they do not differentiatethe type of information (e.g. macro news).Subramahnyam answered that this concern ispartly taken care of by including dummies inthe estimation process. Cheung answered toMenkveld’s comment by using the example oflow trade intensity. In this case, traders knowthat if a trade occurs, it is more likely that it isfollowed by even more trades. Hence, in orderto capture gains from trade, the price will be setas highly as possible, thus increasing the bid-ask spread which in turn materialises in a high

price impact. Fleming replied that it wouldindeed be a good idea to remove temporaryeffects from prices. Menkveld further askedwhy previous day volume would affect currentinnovations and volatility. Fleming recognisedthat the measure of volatility based on closingyield may be too crude, in the sense that it doesnot take into account changes that occur withinthe day.

SESSION 1.2 TOPICS IN THE GOVERNANCE ANDINTEGRATION OF EUROPEANFINANCIAL MARKETS

Charles Kahn (University of Illinois) chairedthe session on “Topics in the Governance andIntegration of European Financial Markets”.Mariassunta Giannetti (Stockholm School ofEconomics) presented the first paper “Investorprotection and equity-holdings: Anexplanation of two puzzles” (joint with YrjöKoskinen, Stockholm School of Economics).The authors develop a theoretical model tosimultaneously explain two well documentedpuzzles in portfolio theory, limited stockmarket participation and home-equity bias. Themodel consists of two countries whereinvestors face a portfolio choice problembetween risky and risk-free assets in both, thehome and the foreign country. Participation inboth markets for risky assets is subject to thesame fixed cost, but offers benefits fromdiversification. Investors have the possibilityto acquire a controlling stake in the domesticrisky asset which allows them – depending onthe quality of investor protection – to extractprivate benefits to the detriment of minorityinvestors. Investment behaviour is driven bythe investor protection and wealth inequality.The degree of investor protection determinesthe attractiveness of participating in a market.Unevenly distributed wealth enables investorsto acquire a controlling stake in the domesticasset, which is the more profitable the weakerinvestor protection. The interaction betweenthe effects of the degree of investor protectionand of wealth inequality drives the followingresults. First, limited domestic stock marketparticipation arises if wealth in a country is

103ECB


ANNEX E

sufficiently unevenly distributed and investorprotection is poor. Second, home-country biasin equity investment arises due to two facts.Bad investor protection gives incentives toobtain a controlling stake in the domestic riskyasset. Furthermore, domestic investorsparticipate in the domestic market if investorprotection is sufficiently good. Third, wheninvestors face relatively poor protection athome they are more likely to access the foreignmarket (good country bias). The paperconcludes by providing new empiricalevidence for the latter two results. In theirsample, they find a positive correlationbetween stock market participation andshareholder rights measured by the antidirectorrights index constructed by La Porta et al.(1998). Also they find the mean level of foreignequity shares is lower in countries with highlevel of investor protection. The analysisshows that the full integration of Europeanmarkets will necessitate a convergence ininvestor protection.

The second paper, “Financial marketintegration and loan competition: when isentry deregulation socially beneficial” wasgiven by the 2003 ECB Lamfalussy fellow LeoKaas (University of Vienna). This paperaddresses the consequences of entryderegulation in a closed banking sector oncompetition, bank stability and economicwelfare. In Kaas’ model, banks compete forborrowers that differ in their quality. In orderto reduce their risk, banks screen borrowers,but are nevertheless exposed to failure, asscreening yields imperfect information anddefault probabilities are subject to aggregaterisk. Banks incur a fixed cost of operation andscreening is costly. The paper assumes thatpotential entrants are more efficient inscreening than incumbent banks. Finally,capital requirements can secure banks againstfailure. Kaas presented first the benchmark ofan equilibrium where entry cannot take placeand no failures occur. Bank failures are ruledout as long as screening costs are high (strongmarket power), screening is relatively efficientand capital requirements are high. After

deregulation, entry occurs with or withoutcrowding out incumbent banks provided thescreening advantage is large enough andcapital requirements are not too high. Welfareeffects depend on whether entry causesincumbents to fail. If there is no failure ofincumbents, entrants cannot reap all benefitsfrom their ability to screen better. Then, thereis too little entry resulting in an inefficientmarket share for new banks. This result can bereversed if entry causes incumbent banks tofail. In this case, incumbents charge higherrates, thus driving too many new banks intothe market. Finally, Kaas expressed the viewthat, by limiting incumbents failure, capitalrequirements may be detrimental to overallwelfare. However, deregulating access to aclosed banking sector is always beneficial.

Cyril Monnet (ECB) presented the last paper,“Guess what: It’s the settlements!” (jointwith Thorsten Koeppl, ECB). The paperanalyses the role of vertical silos in securitiesmarket organisation for efficient horizontalconsolidation between components of the silo,i.e. exchanges and back-office operations suchas clearing and settlement. Monnet brieflydescribed the integration process among stockexchanges and settlement structures in the euroarea. All mergers between exchanges wereaccompanied by breaking up silos and followedby consolidating clearing and settling within asingle entity independent of the exchange.Starting from these facts the authors outline amodel where two firms that operate a silo canrealise gains from a merger. These gains arisefrom increases in overall demand as well ascost savings. Here, the authors assume that thecosts for settling transactions potentially differacross firms and are private information. Usingtools from mechanism design, the paper showsthat it is impossible to achieve an efficientmerger, i.e. a merger where after the mergersettlement takes place at the lowest cost. Thisis due to the fact that it is too costly to inducethe firms to truthfully reveal their settlementcosts. The second part of the paper presentstwo solutions to this impossibility result. First,a sufficiently high subsidy can realise all


benefits from lower settlement costs, but iscostly itself. Second, the authors offer a lesscostly market solution. They show that firmscan achieve an efficient merger by eachoutsourcing their own settlement operations toan agent. Competition between the agents forsettling all trades of the merged exchangereveals then the true cost of settlement. Hence,they argue that fostering competition and openaccess to securities settlement systems may berequired to enhance the efficiency of securitiesmarkets.

The discussant, Hans Degryse (University ofLeuven and CentER), stressed thecommonality of the three papers. All papersdiscuss different problems associated withfinancial integration, ultimately giving arationale for some form of public intervention.For the first paper, Degryse emphasised thatthe model is novel in jointly explaining twowell-known phenomena: home-equity bias andlimited stock market participation. The modelitself, however, seemed too much contrived bythe assumptions. He, therefore, suggested torelax some assumptions, in particular theimpossibility to take controlling stakes inforeign stocks. Finally, he expressed interestin a dynamic version of the model to study theinteraction of changes in investor protectionand wealth. For Kaas’ paper, the discussanthighlighted the result that increasedcompetition leads to too little entry, which runscounter to similar models of competition withentry. He conjectured that this result is drivenby the fact that incumbents take all theirdecisions before entry occurs. He also stressedthat – empirically – switching costs are morerelevant than efficiency differences betweenbanks. The final paper was criticised byDegryse as too stylised, not taking into accountthe full range of possible ways to consolidatefragmented infrastructures in securitiesmarkets. Furthermore, he suggested that theauthors should put more emphasis on theadvantages of vertical integration whendiscussion splitting silos to achieve horizontalconsolidation. Monnet replied that the paper’smain focus was not on reasons why exchanges

wish to consolidate but rather on the reasonswhy consolidation is difficult to achieve.

LUNCHEON ADDRESS

Panayotis Thomopoulos (Deputy Governor ofthe Bank of Greece) delivered a luncheonaddress on “Financial liberalisation, the Greekexperience”.

KEY NOTE LECTURE ON EUROPEAN VENTURECAPITAL

Vítor Gaspar (ECB) introduced Marco DaRin (University of Turin) who lectured about“European venture capital”. The author firstdefined venture capital firms as a specialisedform of financial intermediation whereby smalllocal partnerships finance new ventures viaequity-like instruments. Venture capital firmsusually offer, as a package, funds, advice andmentoring. The author then pointed to somekey facts of venture capital in Europe, using asa proxy for the number of VC firms, thememberships in the European Venture CapitalAssociation (EVCA). In terms of venturecapital fundraising over the 1990s, it seemsthat European venture capital firms arecatching up with the U.S. However Da Rinpointed out that this might be an illusion asduring the same period the level of investmentis at most 50% of the one of the U.S. Thedifference is due to the fact that fundraising forEurope also accounts for non-venture privateequity, which is not the case for investment.Within Europe, there is a high heterogeneity interms of market size and intensity – a measureof how important VCs are within an economy –and Nordic countries have the most intenseventure capital industry. In other words, theindustry is not represented or active in allcountries.

Then Da Rin moved on to describe key issuesin the European venture capital industry.(1) Based on a sample of 1200 VC-backedcompanies, Da Rin finds that the Europeanlandscape of VC firms is highly captive, with

105ECB


ANNEX E

an important presence of bank subsidiaries,corporate VC firms and public VCs. This is animportant fact, as these types of VC firmsinvest less in the early stages of projects and inhigh-tech projects, provide less soft support toprivate companies, and do not tend to monitorcompanies actively, as they sit less frequentlyon the board of companies relative toindividual VCs. (2) Da Rin then pointed to exitstrategies for VCs in Europe. In theory, exitstrategies are crucial for VCs, as additional exitopportunities create further incentives for VCsto invest. There is indeed evidence thatEuropean VCs aim at selling their firms asapproximately a quarter of VC firms have beeninvolved in firms listed on the “New Markets”.There is also evidence that the existence ofNew Markets fostered the creation of VC firms.(3) Based on a sample of 538 companies listedon Europe’s New Markets, Da Rin thenreported findings that – contrary to what theorypredicts – European VC might actually notmake a difference for the companies they arefinancing. While VC-backed companies raisemore capital at IPOs, they do not tend to growfaster than others. However, the author alsopointed to a difference in the quality of theVC firms, as some do better than others.(4) Finally, Da Rin addressed the link betweenhuman capital and venture capital in Europe.As predicted by theory, and controlling for thetype of VC firms, more experienced partnersprovide more support to companies, partnerswith higher university degrees invest in earlierstages of companies, and high-techinvestments are more often carried out bypartners graduated in sciences.

Da Rin then moved on to the state of financialintegration in the European VC industry. Hereported that this industry is not integratedacross borders. VC firms are investing verylocally and with very few exceptions have lessthan 10% of their partners from foreigncountries. Also, the cross borders investmentof VC firms is very low (less than 2% of totalinvestment). Finally, less than a third of fundsoriginate from foreign investors. Most foreigninvestors are from the US and are concentrated

in only a small number of firms. Da Rin arguedthat this may be explained by the nature of theVC industry, which is quite different from thatof other financial intermediaries: VC firmsmake localised and undiversified investments;VC is based on human rather than financialcapital; and VC has a small number ofinvestors. However Da Rin expects financialintegration to indirectly restructure the VCindustry through its effect on the allocation offunds and through changes in the economicstructures of EU countries, notably on thelisting ability of new firms. In this regard,financial integration may improve exitchannels for VC and reallocate talent andhuman capital.

Gaspar then opened the floor for questions.Joseph Bisignano (BIS) asked Da Rin aboutthe effects of pension funds as sources for VC,which is not present in Europe but very muchso in the US, following the advent of the“prudent man” rule. He also wondered aboutthe future of European VCs given the closure ofthe German Neue Markt and asked whetherVCs could survive in a bank dominatedenvironment. Then the issue of the use of fiscalpolicies to stimulate VCs was raised from thefloor. Da Rin reckoned the importance ofpension funds in having a larger pool ofinvestors going into VC. However, he arguedthat letting pension funds invest is rather aregulatory decision than an economic one. Hethen stated that exit is crucial for VCs and thatthere is definitely a need in Europe forinfrastructures allowing proper exit, whethernational or pan-european. He also elaboratedon fiscal issues, arguing that European fiscalpolicies should be designed to foster anentrepreneurial spirit by giving the rightincentives to start projects, as there are too fewgood firms in which VCs can invest. Finally, hedid not believe that bank VCs are a necessarilydesirable outcome, as they have very differentincentives from private VCs. However heargued that Europe is only in a transitory stageand might mature to a different VC industry atsome point.


SESSION 2.1 CORPORATE BOND FINANCINGAND THE COST OF CAPITAL

Joseph Bisignano (Bank for InternationalSettlements) chaired the session. JohnPuthenpurackal (Ohio University) wassupposed to present his paper “Securityfungibility and the cost of capital: evidencefrom global bonds”, co-authored with Darius P.Miller (Indiana University), but he was unableto travel to Athens due to visa issues. Thispaper was therefore presented by thediscussant, Arnaud Mares (ECB). Globalbonds have several characteristics. First, aglobal bond can be traded in multiple markets(such as the euro and dollar markets) withoutrestrictions. Second, they are soldsimultaneously at the same offer price. Third,they have extremely large supply, in multipletranches of different size and maturity. Finally,trading and settlement systems for globalbonds, are set up to encourage cross-markettrades. In particular, clearing and settlementsystems are integrated for global bonds, so thatcross-border transactions can occur more costefficiently. Using 87 global bonds issued byU.S. firms in 1996-2001, the authors find thatissuing globally tradable securities bringseconomically significant benefits. Inparticular, the borrowing costs – as measuredby the yield to maturity in excess of treasurybonds with the same maturity – is 15 basispoints lower than on comparable U.S. domesticbonds. Moreover, issuing costs of global bonds– as measured by underwriting spreads – are0.13% lower than U.S. domestic bonds.Therefore, global bonds reduce the cost of debtcapital. Furthermore, stock price reactions tothe announcement of global bond issuance arepositive and significant. The evidence thatglobal bonds lowers the cost of capital suggeststhat consolidation and integration of securitiesclearing and settlement systems and thecreation of a global clearing and settlementsolution would benefit firms’ capital raisingactivities.

Yigal Newman (Stanford University), 2003ECB Lamfalussy fellow, presented his paper

“Illiquidity spillovers: theory and evidencefrom European Telecom bond issuance”, co-authored with Michael Rierson (CitiGroup).Using data from 347 bond issues in theEuropean telecom industry between October 1,1999, and July 15, 2001, the authors documentthat a firm’s new issues of corporate bonds cantemporarily raise the yield spreads of otherbonds in the issuing firm’s sector. Theiranalysis is based on a measure of risk-adjustedissuance, which consists of multiplying themarket value of the issuance by its duration.They show that the yield spreads of other bondsincreases on average by 8 basis point on the dayof the issuance and then slowly decays overtime. Furthermore, the increase in yieldspreads begins well before the issuance date.Their result is robust to the riskiness ofoutstanding bonds and to the currency of thenew issuance. The temporary nature of theeffect and the fact that it picks up at the date ofissuance rather than at its announcement areevidence that this effect is not due to newfundamental information about the issuer thatmay be revealed during the issuance process.The authors also present a theoretical model toexplain this finding. In the model risk-averseintermediaries need to absorb some of thenewly issued bonds, while holding otherpositively correlated bonds. As intermediariesare risk-averse, their incentive to hold all bondsfalls when their inventory is high, so that theysell some of the already held bonds. As aconsequence, the price of these bonds declines.Furthermore, this decline precedes thescheduled issuance date by anticipation.

João Santos (Federal Reserve Bank of NewYork) presented his paper “Why firm access tothe bond market differs over the business cycle:A theory and some evidence”. The authorargues that the cost to access the bond marketvaries over the business cycle and that theimpact of recessions is not uniform acrossfirms, as access costs increase most for mid-credit quality firms. He first presented a modelwhere rating agencies have to evaluate low,medium and high firms that are subject to anadverse selection problem. These agencies are

107ECB


ANNEX E

assumed to be more likely to announce a splitrating for mid-credit firms than for either highor low credit quality firms. Also, when firmsissue in recessions, it is assumed that they are atleast as likely to get a split rating than whenthey do so in expansions. If firms do not userating agencies, investors are not able todistinguish their quality and they charge thesame average interest rate to each type of firm.This asymmetry of information obviouslybenefits the low type firms but makes otherfirms worse off. This asymmetry can bepartially solved by hiring rating agencies.However, as ratings become increasinglyuncertain in recessions, the informationbecomes more asymmetric thus affecting thecost of access to capital. Furthermore, lowquality firms will benefit from the additionalasymmetry, while mid-credit quality firms willsuffer most. The author then tests his theoryusing data between 1982 and 2002 on corporatebonds issued since 1970 in the U.S. byAmerican non financial companies. The resultsof the analysis confirm the validity of the twoassumptions in the model. Furthermore, theauthor finds that split ratings – the proxy forincreased asymmetric information – increasethe cost of bond financing in recessions. Also,this impact of recessions on credit spreads isnot uniform across firms.

Arnaud Marès (ECB) acted as discussant. Onthe first paper, he argued that underwriting feesare lower for global bonds simply becausecompetition in the underwriting business inEurope is fierce. Furthermore, he regretted thatthe costs of maintaining a client relationship,which are not small, are not taken into accountin the analysis. Finally, he argued that whileliquidity is higher, firms issuing global bondsseem to value the larger customer base inoverlapping trading hours between Europe andthe U.S., while they do not value so much therelatively longer trading hours. On Newman’spaper, Marès wondered whether there wouldnot be a significant change in perceptionregarding the liquidity risk of a firm issuinglong term debt to refinance short term debt.Finally, he speculated that Santos’s paper

could be interpreted as an explicit criticism ofrating agencies.

Bisignano then opened the floor for question.Regarding Miller and Puthenpurackal’s paper,he argued that 15 basis points may not beso large a decrease in borrowing costsfor global bonds. Alberto Giovannini(Unifortune) shared his experience with thefirst issue of global bonds in Italy in fall 1993.He explained that the benefits of issuing globalbonds was not so much the increased liquiditythan the perception that accessing the U.S.domestic market increases the value of theissuance. On Newman’s paper, AlbertMenkveld (Vrije Universteit Amsterdam)wondered why liquidity providers wereentering the market with positive inventories ofpositively correlated bonds. Newman repliedthat this assumption could probably be relaxedwithout affecting the results of the model, asthe change in incentives to hold bonds is whatmatters. On Santos’ paper it was argued fromthe floor that asymmetric information maydecrease in bad times, as bad firms are less ableto sustain shocks. Furthermore, it was arguedthat some assumptions are probably strongerthan necessary. In particular, it is most likelythat the result is robust to the assumption thatrating agencies can both agree on the wrongassessment. However, the assumption that splitratings are obtained randomly was questioned.Vítor Gaspar (ECB) then suggested that theauthor refers to the importance of moral hazardin his interpretation of the result, as the absenceof rating agencies would probably increase it.Santos replied that the data was dictating thechoice of assumptions. He also argued thatthere is no doubt that rating agencies areimproving information in the model, while alsoaffecting the cost of access to capital for firms.

SESSION 2.2 IPOS IN NEW MARKETS

Vicente Pons (Yale University) chaired thesession and introduced the first speaker,Armin Schwienbacher (University ofAmsterdam) who presented the paper“Liquidity of exit markets and Venture Capital


finance”, co-authored with Grant Fleming(Australian National University) and DouglasCumming (University of Alberta). The topic ofthe paper is the effect of exit opportunities onstrategic investment behaviour of venturecapitalists. Indeed, in venture capitalinvestments, liquidity risk represents the riskof not being able to sell the shares after a fewyears and thus being forced either to remainmuch longer in the venture or to sell the sharesat a high discount. Using data of VentureXpert,a dataset of Venture Economics, the authorsrandomly selected investments from 1985 to2001, including three stages of investments.They used the number of IPOs per year as aproxy for liquidity. Controlling for industry-specific risk, they found a positive relationbetween liquidity of exit markets and thelikelihood of investing in new projects, ascompared to follow-up projects. However, theyalso documented a negative relation betweenliquidity and investments in new early-stageprojects. In the author’s words, “when liquidityis high, they rush to exit by investing more innew later-stage projects”. Finally, theyreported a decrease in syndicate size whenliquidity of exit markets is high. In suchcircumstances, the investment is less risky andthus there is less need for syndication of deals.Similarly, when liquidity is low, venturecapitalists may prefer to syndicate more inorder to increase the screening of projects byinvesting only if other also join after havingdone their own screening of the proposal.

Tereza Tykvová (ZEW) presented the paper“Are IPOs of different VCs different”, co-authored with Uwe Walz (University ofFrankfurt). The paper analyses the impact ofdifferent types of VCs on the performance oftheir portfolio firms around and after IPOs.Performance is measured in terms of long-runreturns, volatility and under-pricing. Theanalysis is based on a hand-collected data baseincluding all IPOs on the Neuer Markt duringthe period 1997-2002, which means public VCsas well as independent and corporate VCs,German and non-German VCs. The authorsreported that firms backed by bank-dependent

and public VCs have significantly lowermarket value. Also, firms backed by bank-dependent VCs have significantly higher book-to-market ratios, while firms backed byindependent VCs have significantly lowerbook-to-market ratios. Independent VCs alsoimplies significantly larger average abnormalreturns for firms, where abnormal returns aredefined as the difference between individualreturns and market returns. This is in contrastwith firms backed by public VCs who enjoynegative abnormal returns. Finally, the authorsreport that independent VCs hold firms fora longer period before the IPO relative tobank-dependent and public VCs. Post IPOperformance of VC-backed firms are even morestriking. Firms backed by independent VCsperform significantly better than the firms ofother VCs or non venture-backed ones. Theyalso have less risk, i.e. they display a lowerreturn volatility. As such differences were notexpected by market participants, the authorssuspect that these results can be due to therather young and immature aspect of the NeuerMarkt.

The third paper of the session, “Who benefitsfrom IPO underpricing? Evidence from hybridbookbuilding offerings” was presented byVicente Pons (Yale University). In thisempirical paper, the author uses a sample of175 equity offerings, 137 IPOs and 38secondary equity offerings, that took place inSpain from 1985 to 2002. An important featureof the Spanish IPO market is that thedistribution of the IPO between class ofinvestors – retail, local institutional andforeign institutions investors – is stated in thepreliminary prospectus. That is the underwriterassigns the issue to class of investors beforeany investors is allowed to submit applicationsfor IPOs shares. Pons’ analysis concludes thatinstitutions receive nearly 75% of the profits inunderpriced issues, while they have to bearonly 56% of the losses in overpriced offerings.Since superior information regarding first dayunderpricing cannot completely explain theabnormal profits for institutions, he argued thatunderwriters are better informed about the

109ECB


ANNEX E

companies they take public and use thisinformation to favour their long term clients.However, he finds that the preferentialtreatment of institutions does not come at theexpense of retail investors, as they earnpositive profits from participating in the newissues market. In fact, retail investorssubscribe more heavily to underpriced issues,consistent with individuals being partiallyinformed. In particular, there is no evidence ofa winner’s curse on retail investors in the sensethat informed investors demand largerallocations of hot offerings and smallerallocations of those issues identified asoverpriced.

Wolfgang Bessler (Giessen University) actedas discussant. On the paper of Cumming,Fleming and Schwienbacher, he first wonderedabout the optimal investment strategies for VCfirms when the value of the exit market and thefirm valuation are high. He pointed out thatVCs may prefer investing in later stage projectsin order to avoid overvalued early ones andbenefit from good exit conditions. In his view,VCs should care more about the valuation ofthe stock market than about its liquidity. Healso wondered whether there exists a trade-offbetween the choice of technology and exit risk.Regarding the paper by Tykvová and Walz, hefirst pointed out that it is recommendable tocontrol for industry effects. Then he wonderedwhether analysing firms’ performance 2 yearsafter the IPO is not too long, as VCs influenceat Neue Markt is of limited duration(approximately 6 months). Finally, he pointedout that the results should be carefullyinterpreted, since there are only few publicVC-backed firms in the sample the authorsconsider. Then, on the paper by Pons, Besslerwondered whether underwriters are able torecognise underpriced and overpriced IPOs. Healso asked why preferential allocationsconcerned primarily foreign institutionalinvestors and not local investors? Finally, hepointed out that high initial returns on themarket may be a possible explanation for whyretail investors did not suffer losses. Tykvováreplied that a sector decomposition was not

possible at this stage due to lack of data. Shealso recalled that only long-run performancewas of interest to them.

SESSION 3.1 SECURITIES SETTLEMENTSYSTEMS

Vítor Gaspar (ECB) chaired the session andintroduced Jens Tapking (ECB) whopresented “Raising rival’s costs in thesecurities settlement industry”, joint workwith Cornelia Holthausen (ECB). Thistheoretical paper describes a model of pricecompetition between two settlement serviceproviders, a national CSD and a custodianbank. The CSD sets two prices, customers haveto pay a price q for having a securities accountwith the CSD and another price p for settling atransaction on such an account. The custodianbank does the same. There are many otherbanks that have to trade a security issued intothe CSD. Each of these “investor banks” tradesonce. To settle its transaction, each investorbank needs to have a security either with theCSD directly or with the custodian bank.Which of the two service providers an investorbank chooses depends on the prices and on thepreferences of the investor bank for theheterogeneous services offered by the CSD andthe custodian bank. The custodian bank itselfneeds to have a securities account with the CSDto settle transactions between an investor bankwith an account with the CSD and anotherinvestor bank with an account with thecustodian bank. It is shown that the CSD canraise the custodian bank’s costs in a subtle way:As mentioned above, each investor bank tradesonly once, i.e. investor banks with an accountwith the CSD have to pay the price q once andalso the price p once to the CSD. The custodianbank also has to pay q to the CSD once.However, it has to pay the price p many times.The CSD can raise the custodian banks costswithout changing the costs of its investor bankcustomers by increasing – and decreasing – bythe same amount. It is shown that by thisstrategy, the CSD can achieve a higher marketshare than the custodian bank. However, it isalso shown that this market share is not


necessarily too high from a welfare point ofview. The authors concluded saying thatregarding concerns about unfair practices byCSDs toward custodian banks, the ECB coulddiscourage regulatory interventions favoringcustodian banks, as long as CSDs are notallowed to price discriminate betweencustodian banks and investor banks.

Karlo Kauko (Bank of Finland) presented thepaper “Interlinking securities settlementsystems: A strategic commitment?”. This paperprovides an explanation for why links betweenCSDs are set up, but not often used. In the firstpart of the paper, it is assumed that a CSD firstsets a price for settling primary markettransactions. Then investors and issuers agreeon primary market transactions and thesetransactions are settled. Next, the CSD sets aprice for settling secondary markettransactions. Finally, secondary markettransactions are agreed upon and settled. Itis assumed that primary market transactionsand secondary market transactions arecomplementary goods, i.e. there is little benefitfrom primary market transactions withoutsecondary market transactions. Knowing this,the CSD will set a relatively high price forsecondary market transactions. In expectationof this, the investors and issuers trade on theprimary market only if the primary marketsettlement price is very low. In the second partof the paper, it is assumed that the CSD can setup a link to another CSD so that secondarymarket transactions can also be settled in theother CSD. In other words, if the issuer CSDchooses a very high secondary market price,settlement will take place in the other CSD.Hence, the issuer CSD can commit on lowersecondary market prices by setting up the link.It now can choose a relatively high primarymarket price and increase its profit. However,the link is hardly used.

The last paper of the session was “Economies ofscale and technological developments insecurities depository and settlement systems”,by Markku Malkamaki (Bank of Finland),Juhu Tarkka (Bank of Finland) and Heiko

Schmiedel (ECB). This paper investigates theexistence and extent of economies of scale indepository and settlement systems. The authorsshow that settlement in Europe is 33% morecostly than in the US, as the average cost persettled transaction is $3.86 in Europe and only$2.90 in the US. This difference is partlyexplained by the segmentation in the Europeanmarket as the average cost for operating aninternational Central Security Depository inEurope is $40.54, relative to $3.11 for adomestic one, while it is only $2.90 in the US.However, looking at the exploitation ofeconomies of scale in Europe and the US, theyalso show that the latter is operating at a muchmore efficient level. The European settlementinfrastructures show a strong potential for costsaving: cost will raise by a factor of only 0.68when the number of instructions increases by 1– while the same measure for the US is 0.944.Hence, Europe has a lot to gain from furtherconsolidation. However, given the level ofcomplexity in EU international securitiessettlement systems, the effectiveness ofsettlement industry infrastructure may benefitfrom further regulations simplifying theprocedure for cross-border settlement as forinstance advocated in the second Giovanninireport (2002).

Charles Kahn (University of Illinois) acted asdiscussant. In the discussion of Holthausen andTapking’s paper, he emphasised that if costs ofsettlements are mostly fixed costs (or cost byaccount or transaction) then CSDs should berestrained from limiting rivals’ size, as therewould be no networking externalities. He alsosuggested to calibrate the different coststructures to pin down the extent of the networkexternalities, which in turn would make thepolicy implications of the paper more precise.The major point in the discussion of Kauko’spaper was the question of whether the primarymarket really plays an important role in thepricing process, given that primary marketvolumes are much lower than secondary marketvolumes. In particular, Kahn questioned theassumption that CSDs make most of their profitfrom the primary issue of securities and,

DAY 2

111ECB


ANNEX E

for this, are willing to trade-off their profit onthe secondary market. Finally, on the lastpaper, Kahn stressed the high quality of theanalysis carried out, despite data limitations.He wondered whether conditioning on thetype of institutions owning the settlementinfrastructures could have affected the result.Finally, other types of robustness checks wereproposed, such as dropping ICSDs from thesample or measuring the output of CSDs usingthe number of securities they handle.

Gaspar suggested Malkamaki, Tarkka andSchmiedel to consider whether there iseconomies of scope inherent in securitiessettlement activities, which would make thisindustry a natural monopoly. He warned thatfor such an exercise the output would have to becarefully defined. He further commented on thefirst two papers that it would be interesting todistinguish the importance of the interaction ofthe primary and secondary markets. AlbertMenkveld (Vrije Universiteit Amsterdam)then asked about the relative importance ofsecurities settlement costs relative to othercosts of issuing a security, such as bid-askspreads, broker commissions, etc. From thefloor, comments were made on the level ofcompetition in the settlement industry, whichprovides a service to brokers, and on thecompetition from international CSDs. In thisrespect, the question of the value-added ofinternalisation was raised. On the last paper,the audience also asked for clarification as towhether fixed or variable costs are importantfor economies of scale.

Related to the issue of competition, Tapkingreplied that if there is more than one custodianbank, the equilibrium market share of the CSDmight be too small from a welfare point ofview. In response to Kahn, he suggested that inreality the variable costs of settlement mightnot be as low as suggested by Kahn, as there areseveral million transactions per year. However,if these variable costs are assumed to be zero,then the market share of the CSD will be higherthan socially optimal. He also replied thatendogenising behaviour on the primary market

would not affect their results. Finally he notedthat settlement prices can indeed matter,especially when CSDs are profit maximisers.Then Kauko commented that it is not clear thatCSDs’ strategy would be modified if they wereuser owned, rather than for profit institutions.Also, he clarified that investors always havethe choice whether to participate or not.Therefore, issuers have to offer an interestingenough deal to investors. Finally, Schmiedelreckoned that GDP per capita was a very roughproxy for inputs, but had to be used due tolimited data availability. He also took note ofthe propositions for other output proxies.

SESSION 3.2 THE DETERMINANTSOF VC INVESTMENTS

Jan-Pieter Krahnen (CFS) chaired the sessionand introduced Douglas Cumming (Universityof Alberta) who presented the paper “The legalroad to replicating Silicon Valley”, joint withJohn Armour (University of Cambridge). Theauthors consider whether legal reforms canmake a significant difference in the structure ofthe VC industry. To carry out the analysis, theyuse 13 years (1990-2002) of data from theEuropean Venture Capital Association(EVCA), Venture Economics and the CanadianVenture Capital Association, regarding privateequity investment from 15 developedcountries, with 195 overall observations. Inorder to test for how law matters for the supplyof VC finance, the authors use an index of legaland fiscal variables – such as the taxtransparency for domestic investors, or theability to avoid paying VAT – set out by theEVCA. Using this index, they find thatfavourable tax and legal environments increasethe supply of venture capital and facilitate thecreation of VC firms. Furthermore the authorsstudy the impact of personal bankruptcy lawson the demand for VC finance. They find thatsevere personal bankruptcy laws discourageearly stage entrepreneurs and thereforesignificantly reduces the demand for VCfinance. Finally, Douglas Cumming reportedthat government programs in favour of earlystage projects crowd out venture capital and


private equity investment by significantlyreducing overall industry profits per GDP. Heconcluded that the road to establishing anactive private equity market is paved withfavourable tax laws and legal structuresfavourable to VCs, appropriate bankruptcylaws and only very small direct governmentinvestments programs.

Tuomas Takalo (Bank of Finland) thenpresented the paper “Investor protection andbusiness creation” (with Ari Hyytinen,University of California, Berkeley). Theobjective of this theoretical paper is to studythe effects of investor protection on thecreation of start-ups. More precisely, theauthors study the existence of a trade-offbetween investors protection and businesscreation. The source of the trade-off is thatinvestor protection, while increasing thewillingness of investors to lend, potentiallyreduces the incentives of entrepreneurs to setup firms. To make this point clear, the authorsuse a theoretical model of search with moralhazard. Search frictions can delay financing ofprojects. There are two types of moral hazard:entrepreneurs can choose between a projectyielding only private benefits with certaintyand a project yielding transferable benefitswith some probability; also, in caseentrepreneurs chose the latter, they canmisrepresent the outcome of this project.Investors possess an audit technology at a costto monitor the result of the project. Audit costsand private benefits are both declining in thedegree of investors protection chosen by thegovernment. By increasing investor protection,entrepreneurs have diminished incentives toenter the search market to set up a firm, thusreducing the creation of companies. However,investor protection favours a reduction in thecost of finance, as audit costs are lowered.Tuomas Takalo concluded stating that reformsimproving investors’ position can have verydifferent consequences depending on whetherthey imply a reduction in audit costs or alimitation in the choice of projects. In thisregard, the authors proposed that smallcompanies be liberated from many of the

investor protection regulations, but perhaps notfrom transparency regulation.

Giovanna Nicodano (University of Turin)presented the last paper of this session, “Whatdrives the structure of private equityinvestment?” co-authored with Marco Da Rin(University of Turin, ECGI and IGIER) andAlessandro Sembenelli (University of Turin).The paper studies the determinant of thestructure of private equity investments. Theauthors provide first a model where demandand supply factors affect the distribution offinancing between venture capital and non-venture private equity, and between early stageand late stage venture capital investments. Themodel consists of a moral hazard problem à laHolmstrom and Tirole (1997), whereentrepreneurs can either invest in a goodproject or a bad project that yields them privatebenefits. Entrepreneurs have collateral theycan pledge and they can ask two types ofinvestors for funds. Uninformed investors andinvestors that have access to an audittechnology at a cost. The authors associate thelatter type of investors with private equityinvestors. They prove that the private equityratios fall in response to a reduction of bothproject returns and private equity funds, whenthe supply of uninformed capital is infinitelyelastic. Also, an increase in the supply ofprivate equity funds, when there is excesssupply of private equity, affects neitheraggregate nor private equity investments, northe private equity ratios. Using data fromEVCA for 17 Europeans countries for 14 years(1988-2001), the authors find evidence thatwas the possibility of an excess supply ofventure capital funds in Europe in the 1990s.Also, they find that the opening of NewMarkets in some countries in the mid 1990shelps explain the increase in the share of earlystage and high-tech venture capitalinvestments. Finally, they find that publicexpenditure in R&D is significantly positivelyrelated to the share of high-tech investments.

Heather Gibson (Bank of Greece) acted asdiscussant. On the paper by Takalo and

113ECB


ANNEX E

Hyytinen, she wondered to what extent theprincipal-agent problem exists betweenentrepreneurs and venture capitalists who havea more “hands on” approach. She argued thatthe information asymmetry could be morestringent at the search stage, where VCs have toidentify the quality of the project before theyinvest funds. On the paper by Armour andCumming, she noticed that the EVCA legalindex was only from 2003 and therefore no timeseries analysis was possible. She wonderedwhether this would matter given the recentvarious initiatives, e.g. from the EuropeanCommission. She also asked the authors toclarify their interpretation that governmentfunds crowd out private investors because thelatter have to commit to projects before theyknow how much public funds will becomeavailable in the market. Finally, she wonderedwhether results of the effect of the legalenvironment and government finance on theindustry profit would differ if instead ofevaluating profit through stock market returns,profit was evaluated using accounting data, asin the M&A literature. On the last paper,Gibson wondered to what extent the results aredriven by the period of interest which ischaracterised by a rising stock market. She alsonoted that no variables was capturing changesin the legal environment that occurred duringthe sample period. Finally, she asked Nicodanoabout survey evidence pointing to the lack offunds as being the problem rather than to thelack of good projects.

Krahnen then opened the floor for furtherquestions. He started by asking Nicodano for apossible explanation for the positive effects ofgovernment funds, which contrasts with theresult of Cumming. Regarding the first paper, itwas asked from the floor whether local factorscould explain the phenomenon observed inthe Silicon Valley. Then Nicodano askedCumming to what extent measuring privateequity returns using stock market returns affectthe results.

Takalo acknowledged Gibson’s comment andargued that information asymmetries related to

entrepreneurs’ behaviour could still persist ex-post. Cumming argued that a time seriesanalysis on the basis of the legal index wouldbe quite involving as they would have tocompute the index for 15 countries across 13years. As far as returns are concerned, heexplained that using alternative measures ofreturns does not significantly affect the results.Interestingly, he finds that the effect of returnsmatter much for different stages of projectdevelopment. Finally, he explained that thedata are aggregated at such a level that it is notpossible to compare VC industries at theregional level. Nicodano confirmed that theycontrolled for the effect of the market bubbleand therefore would reject the claim that theresults on excess supply is driven by thebubble. She reckoned that institutions havebeen varying through time but explained thatthey could not find any index capturing suchchanges. She explained the claim that thegovernment subsidies reduce the supply offunds: in equilibrium the supply of funds is notobserved, hence, this claim can be understoodas the government subsidies reducing theequilibrium investment level.


Alberto Giovannini (Unifortune) chaired thePolicy Panel of the workshop and made a fewintroductory remarks. First he noted a changein attitudes of policy makers and marketparticipants with regard to financialintegration, which is not any more consideredas a threat to domestic financial systems.However he noted that legislation put intoplace to protect local providers are nowworking in favour of large global providers,emptying out local regions of financialintermediaries that are needed to fosterprogress there. Hence he remarked a movetoward legal harmonisation that would allowequal access to capital across all regions.Giovannini then recalled the three EuropeanCommission’s objective in reforming post-trading in Europe: (1) to maximise the benefitsof scale by creating one single post-tradingarea, (2) to insure that the gains from scales are


available to all and (3) to maximise the risk-benefits of scale by designing the mostefficient structure resilient to systemic risk. Hewarned that while governments are takinginitiatives, it is also the responsibility ofprivate players to achieve a market solutionconsistent with the three objectives. He thenintroduced the four panel participants.

Kenneth Garbade (Federal Reserve Bank ofNew York) asked whether financial marketreforms are only undertaken as a response tocrisis, elaborating on the case of the USTreasury markets. He underlined threeepisodes of failures that concentrated theattention of policy makers on settlement issuesand prompted policy reforms in two of theseinstances. (1) He mentioned first a back-officecrisis in the late 1960s caused by an unforeseenincrease in trading volume and settlementprocesses based on settlement of trade withphysical delivery of bonds. Dealers werereluctant to make trades with same daysettlement which threatened the execution ofU.S. monetary policy. As a response the Fedcreated a security holding facility andexpanded its book entry system. (2) The secondepisode is the destruction of several inter-dealer brokers and the temporary loss of asettlement system during the events of 09.11.In his view, these events highlighted theimportance of back up facilities in dispersedlocations. Also, the loss of information onanonymous trade commitments created theneed to maintain this information in anelectronic format that could be disseminatedrapidly. Most of the disrupted settlementduring these events where repurchaseagreements, illustrating the significance of themarket for Treasury securities. In addition,failures can spread when investors do not lendsecurities any more. The reluctance to lend onthe run issues to dealers that needed them tocure settlement failures was only solved whenthe Treasury re-opened the on the run 10 yearnotes in early October, thus demonstrating itswillingness to take un-precedent actions tosolve the problem. (3) He finally mentioned aproblem involving the 10 year note auction of

May 2003. The demand to borrow the note todeliver against short sales rose dramatically inlate June, when investors realised that the Fedwas unlikely to pursue any effort to reducefuture long term interest rate. The specialcollateral repo rate for the May 10 year note fellrapidly near zero, thus giving little incentivefor sellers to borrow the note to avoid failure.Increasing failures left investors reluctant tolend the note. The problem here originatedfrom the equilibrium dynamics of the marketand not from an exogenous shock. The bettersolution would then seem to be a change in themarket structure than an action from theTreasury.

Randy Kroszner (University of Chicago)focused his remarks on the trade-off betweenfragmentation or consolidation of securitiessettlement systems and the resiliency of thesystem. Fragmentation can provide sources ofdiversification and competition as well asredundancy in case of unforeseen events, whileconsolidation can exploit economies of scalebut also can lead to a monopoly. With respect tothis trade-off, he emphasised that futureattempts to articulate or integrate Asian,European and North American systems willcreate important challenges for regulators.However, Kroszner stressed that market forcesshould not be underestimated in coming upwith solutions. To make his point, he took theexample of the early Chicago derivativesmarket where the clearing house played animportant role in ensuring consistency ofcontract terms and homogeneity in credit riskin being the counterparty to all traders. Also,market participants set up a regulatorystructure such as requiring capital minima. Toconclude, he pointed out that while regulatory,fiscal and legal uncertainty barriers should bebrought down, it might not be desirable to forceexcess uniformity from regulators as theoptimal solution might vary with time.

Anso Thiré (Euroclear) who replaced JoelMérère commented on the fragmentation ofthe European securities infrastructures. Herecalled that the total cost of CSD and ICSD for

115ECB


ANNEX E

Europe represents 1.5 billion euros inoperating expenses, i.e. 1.3 times the cost ofDTCC in the U.S. This proves that significantscale economies can be realised. However, hepointed out that it is difficult to implement suchgains, as most stock exchanges and CSDs aremonopolies in their own markets. Obviously,costs could be brought down via furthercompetition. However Thiré said that most ofcross-border settlements take place in books ofa few large banks that operate for profit and thatCSDs are not active in this market. As aconsequence, he argued that a commonplatform accessible by all could be a solution.The drawback of this solution is that nobodycould force CSDs to participate in such aplatform. In the absence of competition, someform of price regulation would be needed toprevent service providers (including largecustodians) from abusing potentially dominantpositions. Finally, Mr Thiré highlighted somepositive features of user-owned governancestructures which, in his view, ensure that theusers’ needs are taken into account fairly well.

Gertrude Tumpel-Gugerell (ECB) firstrecalled the lack of integration in Europeansecurities infrastructure. This is at the sourceof many costs, such as linking differentsettlement systems, that further consolidationand/or more competition could reduce if noteliminate. In her view, a process of reshapingthe current state of European infrastructure isinevitable. In this regard, she argued thatpublic action may be warranted to foster theprocess of integration in this industry. Sheclaimed that while it is clear that theelimination of the barriers identified by theGiovannini group is a necessary condition foran efficient infrastructure to emerge, it is maynot be sufficient. Market failures can preventmarket forces to freely develop and an efficientmarket infrastructure to emerge. For instance,Ms. Tumpel-Gugerell pointed to the difficultyin allocating the overall benefits that can bereaped from the shift to a consolidatedinfrastructure, or the vested interests ofcustodian banks to maintain fragmentation.She then recalled the ECB’s view that market

forces should drive the process ofconsolidation. In this respect, she mentionedthat the ECB has been active as a catalyst forimprovement by encouraging discussionsamong the relevant players, on harmonisingcentral banks procedures and operations, onsetting standards and co-operating with theCommittee of European Securities Regulatorsto ensure an integrated regulatory andoversight framework. She finally called forfurther research on several topics, such asfindings evidence of market power, ordetermining the level of competition in theEuropean securities infrastructures.

Vítor Gaspar (ECB) asked the panelparticipants to elaborate further on the balancebetween the prospects for market failure andthe prospects for public intervention failure,which could arise from e.g. informationconstraints. He also wondered whether themultiplicity of authorities with overlappingcompetencies in Europe is a relevant problem.Kroszner replied that characterising the trade-off is an impossible task. He argued that this isthe reason for promoting flexibility as neitherthe market nor the government have enoughinformation to design the exact solution.Tumpel-Gugerell then intervened on the issueof multiple regulators in Europe. She arguedthat the problem is about the speed with whichdecisions are taken and would favour publicaccountability related to the decision process.On the public intervention failure, shementioned that one problem is the actualenforcement of regulations rather thanregulations. However she stressed the need forharmonisation of regulations. On theregulatory aspects, Thiré took the case ofEuroclear that has five subsidiaries in each ofthe member countries, facing five differentregulators. Since there is no Europeansregulator, national regulators sign amemorandum of understanding, which relieson the sharing of information. Arnaud Marès(ECB) then asked whether there could be waysto lower costs, such as user ownership.Giovannini explained that advocating aparticular market model would appear


premature at this stage, although theGiovannini Group developed three interestingdifferent scenarios. He also argued that givingthe right incentives to managers to reduce costsis possible without a for profit structure. Thirérecalled that markets have a private agenda andregulators might have difficulties to have thecorrect feedback on how the industry reallyworks. Related to this topic, Giovanniniproposed that a more formal and systematicprocedure of consultation between theregulator and market participants be used.Kroszner then advocated that incentives tomanagers can be and must be set up rightwhether in a corporation or a user ownedstructure. Then Thiré reported that the bonusesgranted to managers in his company aredetermined on the basis of a variety of factors,out of which profit is only one element and noteven the most important one. More importantare factors such as customer satisfaction,timely and effective implementation ofprojects, staff turnover etc. Jan-PieterKrahnen (CFS) then asked the panelparticipants their views on the tension betweencompetition and consolidation and whycustomers could not choose between twoclearing systems. Thiré replied that the answerlies in cost savings for banks only, where oneprovider suffices to settle. Regardingcompetition, he argued that opening upclearing and settlement systems will createmore costs for customers as nettingpossibilities will be reduced and collateralrequirements increased. Related to the USclearing situation where two big banks clearedinternally trades prior to the 09.11 events,Joseph Bisignano (BIS) asked Ken Garbadewhether these events constituted a test todetermine which of these two banks were best –given that one of this bank fared poorly in itstrading activities. Garbade replied that oneback-up facility indeed failed, but none of thedealers that cleared through this bank switchedto the other bank. He also elaborated oncompetition arguing that it does not appear asprice competition in this sector.

CLOSING REMARKS

Vítor Gaspar (ECB) closed the workshop withsome remarks. He first reviewed some of theimportant findings of the papers presented inthe workshop in the three main areas on theprogram, namely (1) European securitiessettlement systems, (2) Start-up financing andnew markets and (3) European bond markets.He mentioned that further consolidation iswarranted by the evidence of significanteconomies of scale in this industry, noting thefact that vertical integration can prevent theefficient consolidation of trading andsettlement platforms as a striking result. Hesingled out Marco Da Rin’s claim that there isvery little integration of the European venturecapital industry and wished that further effortsbe made in this direction. Finally, on corporatebond markets he highlighted the differingaccess of firms to the bond markets over thebusiness cycle, and the effects of monetarypolicy on the liquidity of (government) bondmarkets, as particularly relevant for centralbankers.

He then announced the Network will concludeits first 2 years of existence with a finalsymposium in Frankfurt on 10th and 11th of May2004, covering all the three main areas ofinterest to the Network, namely (1) theimplications of European financial integration,especially for banking, the development ofimportant financial market segments, theirregulation and infrastructures, (2) Financialsystem structures in Europe, in particular assetsecuritisation, corporate governance andcontrol, and economic performance, and(3) Financial linkages between Europe, the USand Japan. The best papers presented at thefinal symposium will be invited for submissionto a special issue of the “Review of Finance”.

Vítor Gaspar also announced that the nextphase of the workshop is already underpreparation. In this next phase, more attentionwill be geared toward financial systemmodernisation and its impact on economicgrowth in Europe. Also, the Network will seek

117ECB


ANNEX E

to foster research on the relationship betweenfinancial integration and financial stability.Finally, the development of the financialsystems of countries acceding to the euro areawill be of particular interest.

Finally, he thanked the local organisers,George Tavlas, Panayotis Thomopoulos,Nicholas Tsaveas, Heather Gibson and TinaKourkoutsaki for having hosted the workshop.


Day 1 (Monday 10 May, 2004)

8:30-9:00 Registration

8:45-9:15 Welcoming coffee

9:15-9:30 Vítor Gaspar (European Central Bank)Opening Remarks

9:30-10:15 Alexandre Lamfalussy (Université Catholique de Louvain)Keynote Speech on European Bond Markets



11:00-13:00 Parallel sessionsSession 1.1: European government bond market microstructureChair: Ignazio Angeloni (European Central Bank)

Bruno Biais (University of Toulouse) Liquidity and the cost of funds in theEuropean treasury market (with Antoine Renucci and Gilles Saint-Paul,University of Toulouse)

Marco Pagano (University of Salerno) Valuation, liquidity and risk ingovernment bond markets (with Carlo Favero, Università Commerciale LuigiBocconi and Ernst-Ludwig von Thadden, University of Lausanne)

Albert J. Menkveld (Vrije Universiteit Amsterdam) Euro-area sovereign yielddynamics: The role of order imbalance (with Yiu C. Cheung and Frank de Jong,University of Amsterdam)

12:30-12:45 Discussant: Eli Remolona (Bank for International Settlements)


Session 1.2: Bank competition and its implicationsChair: Nicholas Tsaveas (Bank of Greece)

Steven Ongena (Tilburg University) The impact of competition on bankorientation and specialization (with Hans Degryse, Tilburg University)

ANNEX F PROGRAM AND SUMMARY OF THESYMPOSIUMHELD AT THE ECB IN FRANKFURT, MAY 10-11, 2004

119ECB


ANNEX F

Elena Carletti (University of Mannheim and Center for Financial Studies)Multi-bank lending: diversification and free-riding in monitoring (with VittoriaCerasi, Università degli Studi di Milano Bicocca, and Sonja Daltung, SverigesRiksbank)

Stefan Arping (University of Amsterdam) Playing hardball: relationshipbanking in the age of credit derivatives

12:30-12:45 Discussant: Bruno Parigi (University of Padova)


13:00-14:30 Lunch

14:30-16:30 Parallel sessionsSession 2.1: International financial linkages – capital flowsChair: Holger Wolf (Georgetown University)

Rui Albuquerque (University of Rochester) International equity flows andreturns: A quantitative equilibrium approach (with Gregory H. Bauer,University of Rochester and Martin Schneider, New York University)

Jonas Vlachos (University of Chicago) Does regulatory harmonizationincrease bilateral asset holdings?

Michael Halling (University of Vienna) Where is the market? Evidence fromcross-listings (with Marco Pagano, University of Salerno, Otto Randl,University of Vienna and Josef Zechner, University of Vienna)

16:00-16:15 Discussant: Philip Lane (University of Dublin, Trinity College)


Session 2.2: Financial structures, competition and growthChair: Luigi Guiso (University of Sassari)

Stijn Claessens (University of Amsterdam) Competition in the financial sectorand growth: A cross-country perspective (with Luc Laeven, World Bank)

Solomon Tadesse (University of South Carolina) Financial architecture andtechnology

Enisse Kharroubi (DELTA and Banque de France) Financial integration: Forwhom can it be a wrong medicine?


16:00-16:15 Discussant: Yrjö Koskinen (Stockholm School of Economics)



17:00-17:30 Measures of financial integration (Lieven Baele, Ghent University, AnnalisaFerrando, Peter Hoerdahl, Elizaveta Krylova and Cyril Monnet, EuropeanCentral Bank)

17:30-17:45 Discussant: Michael Melvin (Arizona State University)


19:30 DinnerOtmar Issing (European Central Bank)Dinner Speech on Asset Prices and Monetary Policy

Day 2 (Tuesday May 11, 2004)

8:15-8:45 Registration

8:30-9:00 Coffee

9:00-9:45 Robert Flood (International Monetary Fund)Key Lecture on New Approaches to Assess Financial Integration



10:30-12:30 Parallel sessionsSession 3.1: Topics in financial integrationChair: Francesco Drudi (European Central Bank)

Mark Carey (Federal Reserve Board) Is the corporate loan market globallyintegrated? A pricing puzzle (with Greg Nini, Federal Reserve Board)

Thomas Harr (University of Copenhagen) Branch or subsidiary? Capitalregulations of multinational banks (with Thomas Roende, University ofCopenhagen)

Jens Tapking (European Central Bank) Horizontal and vertical integration insecurities trading and settlement (with Jing Yang, Bank of England)

12:00-12:15 Discussant: Harry Huizinga (Tilburg University)


121ECB


ANNEX F

Session 3.2: International financial linkages – comovementsChair: Michael Binder (University of Frankfurt and Center for FinancialStudies)

Clara Vega (University of Rochester) Real-time price discovery in stock, bondand foreign exchange markets (with Torben G. Andersen, NorthwesternUniversity, Tim Bollerslev, Duke University and Francis X. Diebold,University of Pennsylvania)

Robert Connolly (University of North Carolina – Chapel Hill) Commonality inthe time-variation of stock-bond and stock-stock return co-movements (withChris Stiversm, University of Georgia and Licheng Sun, Penn State Erie)

Philipp Hartmann (European Central Bank) The breadth of currency crises(with Stefan Straetmans, Maastricht Universiy and C.G. de Vries, ErasmusUniversity of Rotterdam)

12:00-12:15 Discussant: Paolo Pasquariello (University of Michigan)


12:30-13:45 Lunch

13:45-15:15 Policy Panel: Drivers of european financial integration – markets orpolicy?Moderator: Tommaso Padoa-Schioppa (European Central Bank)

Mario Draghi (Goldman Sachs)Alexander Schaub (European Commission)Jens Thomsen (Danmarks Nationalbank)

15:15-15:45 Gertrude Tumpel-Gugerell (European Central Bank)Speech on The Role of the ECB in Financial Integration


16:15-18:15 Parallel sessionsSession 4.1: Firm financing and corporate governanceChair: Jan-Pieter Krahnen (University of Frankfurt and Center for FinancialStudies)

Mihir Desai (Harvard University) Institutions, capital constraints andentrepreneurial firm dynamics: evidence from Europe (with Paul Gompers andJosh Lerner, both Harvard University)


Renée Adams (Stockholm School of Economics) A theory of friendly boards(with Daniel Ferreira, SITE, Stockholm School of Economics)

João A. C. Santos (Federal Reserve Bank of New York) Identifying the effect ofmanagerial control on firm performance (with Renée Adams, StockholmSchool of Economics)

17:45-18:00 Discussant: Giancarlo Spagnolo (Sveriges Riksbank andUniversity of Mannheim)


Session 4.2: Systemic riskChair: Garry Schinasi (International Monetary Fund)

Reint Gropp (European Central Bank) Bank contagion in Europe (with JukkaVesala, European Central Bank)

Grégory Nguyen (National Bank of Belgium) Interbank exposures: Anempirical examination of systemic risk in the Belgian banking system (withHans Degryse, Tilburg University)

Giulia Iori (Kings College, London) An analysis of liquidity and systemic riskin alternative securities settlement architectures

17:45-18:00 Discussant: Kostas Tsatsaronis (Bank for InternationalSettlements)


123ECB


ANNEX F


The Symposium took place at the ECB inFrankfurt on 10 and 11 May 2004, and aimed toconclude two years of work under the network.It combined research key lectures, researchpaper presentations on the five priority areas,and a plenary panel discussion on “Drivers offinancial integration: market or policy”, whichwas chaired by Tommaso Padoa-Schioppa(European Central Bank) and included MarioDraghi (Goldman Sachs), Alexander Schaub(European Commission) and Jens Thomsen(Danmarks Nationalbank) as panellists. Thisdocument summarises the networkSymposium.

I N T RODUC T I ON


Vítor Gaspar (ECB) opened the Symposiumwith some brief remarks. He first recalled thatthe original idea of the ECB-CFS Researchnetwork was to form a coherent group ofresearchers with different backgrounds, linkedby their interest in the integration of Europeanfinancial markets. He noted that in his view thishad been achieved. He also noted that a majorachievement of the network was to kick-startnew research in the area of securitiessettlement systems. He then went on todescribe the role of the euro in fosteringfinancial integration. It forced nationalgovernments to adopt economic policiesgeared toward price stability and balancedfiscal positions. This led to a generalconvergence of the economies of the euro area,which was clearly reflected in the impressiveconvergence of bond and money market yieldsin the run-up to EMU. With the introduction ofthe euro, a euro-wide money market wascreated almost instantaneously. Countryspreads in the overnight interbank rate becamenegligible and were driven to zero shortlyafterwards. He recalled that similardevelopments occurred in the government bondmarkets, although significant cross-countrydifferences remain. The elimination ofcurrency risks and the accompanyingharmonisation of financial regulations have ledat the same time to a drastic reduction intransaction costs across euro area financialmarkets. Also, he mentioned that the increasein the number of potential investors and thereduction in underwriting costs led to asubstantial increase in the amount of corporatebond net issues.

He continued with the role of banks in theEuropean financial system. The traditionalfunction of intermediation between depositsand loans has diminished, given the increasedimportance of investment funds, pension fundsand insurance companies. For investmentbanks, geographical national boundaries thatonce limited their scope of activities have lostmeaning. Mergers and acquisitions, instead,have been used to gain access to the retailsector of foreign countries. Finally, he

concluded that the euro was a major forcefostering the development of Europeanfinancial markets, and that these profoundtransformations have important realimplications in terms of growth.

Gaspar then introduced AlexandreLamfalussy (Université Catholique deLouvain) who delivered a keynote Speech onthe European Bond Markets.

Mr. Lamfalussy tackled three problems in hiskey lecture: (1) the impediments to furtherintegration and efficiency of the eurodenominated bond markets; (2) the evolvingregulatory structures in the European securitiesmarkets; and (3) the existing co-movementsbetween US and euro area long-term interestrates.

Regarding the first topic, Mr. Lamfalussy firstrecalled that while the unsecured overnightinterbank market is almost perfectlyintegrated, equity markets suffered the mostfrom a lack of integration, the eurodenominated government debt markets, lyingin between these two extremes. Despite theprogress toward financial integration, hementioned two barriers to further integration ofthe government bond markets: thesegmentation in post trading arrangements andthe difficulty of creating a single issuerinstrument for government bonds. On the firstimpediment, Mr. Lamfalussy evoked the workof the Giovannini group, mentioning tax andlegal structures as a barrier to furtherintegration of securities clearing andsettlement systems. According to him, thiscalls for further government action, as themarket cannot act upon these issues. He alsoexpressed his reluctance to adopt a model ofintegration that would lead to a single CCP anda single settlement platform in Europe. Theremoval of legal barriers would be too lengthy aprocess and full centralisation is not necessaryto achieve the full benefits of consolidation, astechniques to interface the different systemsare now available. He would rather favour theinterconnection of the different settlement

DAY 1

125ECB


ANNEX F

systems in real time. Although a difficultendeavours, he would first recommendspreading the techniques of electronic tradingplatforms.

On the second impediment, Mr. Lamfalussyrecalled that a single issuer instrument forgovernment bonds would enhance theefficiency of the government bond markets byguaranteeing a much better liquidity than isnow the case. However he admits that a singleissuer for medium to long-term bonds is neitherfeasible nor desirable, as there may besubstantial differences between countries.However, he noticed that there are no ratingdifferences in the short-term end of the marketbetween countries. As a result, short-termgovernment bonds could be merged into asingle instrument. According to Mr.Lamfalussy, this would substantially increaseliquidity and enhance the overall efficiency ofthe government bond market.

Mr. Lamfalussy then moved onto describingthe evolving regulatory structure in Europeansecurities market. He recalled that theCommittee of Wise Men advised to adopt anopen-ended solution on the issue of a Europeanregulator. This choice was driven by twofactors: first proposing a single regulatorwould have slowed down the Europeanintegration process, as some countries wouldhave opposed it strongly. Second, there was nocore European legislation in 2000-2001 onwhich the action of a single regulator could bebased. However, things are now evolving fast.The implementation of the FSAP will almostcertainly be completed in 2005 and will thusprovide a body of Europe-wide core legislationproviding the basic framework for thefunctioning of financial markets.

However Mr. Lamfalussy recalled theuncertainty surrounding the implementation ofthe FSAP measures at the national level. Thesuccessful implementation of the FSAP nowdepends on CESR. He recalled that CESR(Committee of European Securities

Regulators) worked out the implementationdetails when dealing with the FSAP and willnow be in charge of overseeing the nationalimplementation. However, CESR will have toovercome very heterogeneous regulatory andsupervisory practices at the national level,which will complicate its work as muchcoordination among supervisors and regulatorswill be needed.

Nonetheless, he noticed that reforms of theregulatory structures at the national levels havebeen taking place recently. New developmentsare tackling the issue whether all financialindustry regulators should work on the sameroof or be separated, and the power that shouldbe granted to these bodies. Reforms revealnational differences and create difficulties, asthey are not all going in the same direction.Finally, Mr. Lamfalussy noted that theECOFIN extended the four level regulatoryprocess to banking and insurance. However,further difficulties may appear, as in bankingand insurance prudential issues are key forregulations while they are not so important forsecurities. This raises the issue of micro versusmacro prudential policies, where crisisprevention stops and crisis solution begins.Difficulties may also be created by thedifferent roles of National Central Banks in thisprocess, as some are full macro-prudentialregulators while others are not. The basicquestion is therefore: who should be in chargeand in what area? Regarding integration, herecalled that better functioning markets bringmore growth but also tend to be less stable. Hewould therefore encourage discussions on asingle regulator.

Finally, Mr. Lamfalussy tackled the question ofco-movements between US and euro area long-term interest rates. Recent studies showevidence that these long-term interest rates areco-varying. He saw two possible explanationsfor this phenomenon: 1) financial integration atthe world level and 2) an increased correlationof the real economies and the lead of the USeconomy over Europe’s. Mr. Lamfalussy


expressed the views that inflation expectationtriggered by much broader developments suchas oil prices may explain this phenomenon.

Gaspar then opened the floor for questions.Christian Upper (Deutsche Bundesbank)commented that a single issuer might causeproblems when or if the ratings for short-termbonds, which are for the moment similar acrosscountries, diverge. Vítor Gaspar followed upon this issue, questioning the ability of marketsto discriminate sovereign borrowers in terms ofcredit risk. Marco Pagano (University ofNaples) pointed out that a way to monitorindividual countries in case a single issuer is inplace is to require them to issue T-bills inparallel to the common issuance. In case theratings of the individual country issuedecrease, then the single issuer could choosewhether to keep this country in the commonissuance. Bruno Biais (University ofToulouse) remarked that it was not evennecessary to have a parallel issue for short-terminstruments as long-term instruments could aswell be used as monitoring device. GarrySchinasi (IMF) asked what is the role of centralbanks in crisis prevention and resolution. Healso asked whether central banks have a naturalrole in ensuring financial stability. Mr.Lamfalussy answered that neither is there littledoubt that the ECB should play a role in case ofa systemically important event, nor is theremany questions about its responsibility inensuring the proper functioning of the paymentsystem. However, he pointed out twodifficulties that may emerge in crisisprevention/resolution. First, information needsto flow without constraint from nationalsupervisory authorities (if any) to NationalCentral Banks to the ECB, for appropriatemeasures to be taken. Second, crisis resolutionmay require specific lending to take place, thusrequiring the involvement of governments. In aworld with multinational banks, proper andspeedy co-operation of governments maytherefore be needed.

Jing Jang (Bank of England) wonderedwhether a fully integrated securities settlement

system always brings more systemic risk, andwhether a pure delivery versus payment systemwould not to a large extent take care of this risk.Mr. Lamfalussy underlined that to make asystem systemic-risk-proof is just a matter ofworking out the details of the functioning of thesystem. Rather, the reason why it looks riskieris that it becomes the system, which mayrequire stronger security arrangements.

One question from the floor related toconsolidation in the European banking sector.Mr. Lamfalussy is of the view that muchprogress on consolidation has been made at thedomestic level in some countries, and there isnow an increasing potential for cross-borderinitiatives. However these initiatives might bediscouraged if there is regulatory uncertainty.Finally, related to a question on the leadsupervisor, he mentioned that this is a difficulttopic as the lead supervisor has an immediateconsequence on central banking assistance andgovernment involvement, for instance in termsof deposit guarantees.

SESSION 1.1 EUROPEAN BOND MARKETMICROSTRUCTURE

Eli Remolona (Bank for InternationalSettlements) chaired the session. Heintroduced the first paper of the session“Liquidity and the cost of funds in theEuropean treasury bills market” by BrunoBiais, Antoine Renucci and Gilles Saint-Paul(all IDEI, University of Toulouse). Using dataon yields and the amounts issued for Treasuryauctions in Belgium, France, Germany,Greece, Italy, Portugal and Spain between 2001and 2003, the authors study how differences inmarket microstructure across countries andthrough time affect short-term yields ingovernment treasury auctions, and thus the costof funds for governments. To study their impacton yields, they use macroeconomic variables,such as the volatility of the stock market beforethe auction and the ratio of public debt ratio.They find that these variables have asignificant impact on yields. In particular,large public deficit raise yields. Also, high

127ECB


ANNEX F

volatility in the stock market before the auctionreduces yields, thus allowing governments tosell Treasury bills at relatively high price. Theyalso find that market microstructure affectsyields. For example, regularly issuing billssignificantly reduces yields. Also, when billsare traded on a centralised, transparentelectronic limit order book such as MTS, theirliquidity rises and the yields significantlydecline. However, the authors find no supportfor the hypothesis that the amount of billsoutstanding matters for yields. The main policyimplication of the paper is that governmentscould enhance liquidity and reduce yields andthe costs of their funds by efficiently designingthe Treasury markets and securities. Moreprecisely, improving the microstructure of theTreasury market – by using regular issuanceand a limit order book – increases yield spreadsrelative to Euribor (from 2.9% to 6.2% for the3 months maturity, for instance). This impliesthat governments would raise euro 350.19million more at the time of the auction.

Marco Pagano (University of Naples)presented the paper “Valuation, liquidity andrisk in government bond markets” joint withCarlo Favero (University of Bocconi) andErnst-Ludwig von Thadden (University ofLausanne). This paper aims to explore thedeterminants of observed yield differentialsbetween long-term sovereign bonds in the euroarea. The authors use daily data from Euro MTSGroup’s European Benchmark Market tradingplatform for 5-year and 10-year maturities forthe period January 1992 to December 2003.They show that there is a strong co-movementamong countries’ yield differentials betweensovereign bonds and German bonds. Thiscommon trend appears to be highly correlatedwith the differential between high-risk U.S.corporate bonds and U.S. government bonds atthe corresponding maturity, a measure of theinternational risk factor. In contrast, liquiditydifferentials – as proxied by bid-ask spreaddifference between the local and the Germanrelevant spread – display sizeableheterogeneity and no common factor. Thissuggests that liquidity is unlikely to have a

direct impact on yield differentials, while itmay have an impact through its interaction withrisk. The authors present a model in whichyield differentials should increase in bothliquidity and risk, with an interaction termwhose magnitude and sign depends on the sizeof the liquidity differential with respect to thereference country. Testing these predictions,they find that the international risk factor isconsistently priced, especially for high-yieldcountries and for longer maturities. However,liquidity differentials are priced only for asubset of five countries (out of a total of eight)and their interaction with the risk factor iscrucial to detect their effect.

Albert J. Menkveld (Free University ofAmsterdam) then presented his paper “Euro-area sovereign yield dynamics: the role oforder imbalance”, joint with Yiu C. Cheungand Frank de Jong (both University ofAmsterdam). The project addresses thebehaviour of yields in the government bondmarkets of Germany, France, Italy andBelgium. Using data from Euro MTS, theystudy daily changes in euro-area ten-yearsovereign yields by decomposing them intobenchmark (German) yield changes, commonyield spread changes, country-specificchanges, and temporary changes. They findthat none of the national order imbalancesinfluences benchmark yield innovations.However, common yield spread innovationsare only driven by Italian imbalances. Theauthors argue that dealers across Europe offsetany common yield spread exposure through thehighly liquid Italian market.

The discussant, Eli Remolona (Bank forInternational Settlements) started with anoverview of the three papers, proceeding inreverse order. On the paper by Menkveld, henoted the striking result that order imbalanceshave no effects on the benchmark yieldinnovation. He also asked the author toelaborate further on the role of information foryield changes. In particular, how are yieldchanges affected by the arrival of publicinformation? In particular, he mentioned the


well known reaction of European yields to USannouncements. Also, for order imbalances tohave a permanent effect, they must containinformation on fundamentals. Remolonatherefore concluded that it would be interestingto know which fundamentals are driving theItalian, Belgian and French imbalances. On thepaper presented by Pagano, the discussantnoted that the concept of liquidity used in themodel is not risky. He therefore suggested toendogenise market liquidity by introducingmarket makers in the model. Finally, on thefirst paper, Remolona wondered whether thedesign of the auction itself mattered for theamount of funds that can be raised bygovernments. He mentioned in particular thecase of the auction performed by ‘Google’.Contrary to the usual practice of relying ondiscriminatory auctions (where the winnerpays what she bids), they used a Dutch auction(where the winner pays the cut-off bid).Finally, he asked whether bank risk matters. Tocontrol for market conditions, the authors usedthe percentage spread against the Euribor,which is an unsecured interbank rate. Mostbanks in this market are AA rated banks, whichinvolves some risk in the long term. Hence,controlling for bank risk may be advisable.

Eli Remolona then opened the floor forquestions. Clara Vega (University ofRochester) remarked that in the paper byPagano, the authors only look at countryspecific risk and asked whether using acommon factor analysis to look at thesystematic liquidity risk would modify theresults. On the same paper, Christian Upper(Deutsche Bundesbank) wondered whether theempirical implementation of the model wouldbe robust having a risky instead of a risk-freebenchmark bond. Finally, on the paperpresented by Menkveld, the result that Italianorder imbalances are important wasquestioned. In particular, it was argued thatBelgian imbalances should have a moreprominent effect as this market is relativelyless liquid than the Italian one. Pagano repliedthat the model and the results would generaliseto an environment with a risky benchmark

because the predictions are about the impact ofchanges in relative risk. Menkveldacknowledged the importance of publicinformation and reported that adding dummieson announcements by the ECB did not changethe result.

SESSION 1.2 BANK COMPETITION AND ITSIMPLICATIONS

Nicholas Tsaveas (Bank of Greece) introducedthe first paper of the session “The impact ofcompetition on bank orientation andspecialisation” by Hans Degryse (K.U.Leuven) and Steven Ongena (TilburgUniversity). Using a data set containing bankloans to over 13,000 Belgian firms, comprisingthe entire loan portfolio of an important bank inBelgium, the authors investigate the effect ofbank competition on bank branch orientationand specialisation. Around 83% of the firms inthe portfolio are single-person business andmost borrowers obtain just one, relativelysmall loan from this bank. To carry though theanalysis, the authors define relationshipbanking as the situation where the bank is themain bank of the firm (i.e. it has a monthlyturnover on its current account of at least€2500 and it possesses at least two productsfrom the bank) and the relationship has beenongoing for at least one year. Competition in agiven zone is proxied by the Herfindahl-Hirschman Index (the summed squares of bankmarkets shares divided by the number ofbranches in each zone of interest). The authorsfind that an increase in the banks’concentration index from 0.06 to 0.5 decreasesthe probability of observing relationshipbanking by almost 10%. Hence they find thatbank branches facing stiff local competitionengage relatively more in relationship-basedlending. The effect of competition on industryspecialisation is much less pronounced, as theauthors conclude that branches of the bankengage somewhat fewer borrowers in the sameindustry if local market concentrationdecreases. The effects appear rather modest,both in statistical significance and economicrelevance. Finally, the authors report that the

129ECB


ANNEX F

probability of observing relationship bankingdecreases significantly with the distanceseparating the borrower from the lender.

Elena Carletti (University of Mannheim andCFS) presented the paper “Multiple-banklending: diversification and free-riding inmonitoring” with Vittoria Cerasi (Universitàdegli Studi di Milano Biocca) and SonjaDaltung (Sveriges Riksbank). The paperanalyses theoretically banks’ choice betweenlending to firms in exclusive relationships andsharing financing with other banks in a contextwhere both firms are subject to moral hazardproblems and bank monitoring is essential forfinancing to take place. Firms need funds toundertake projects and can decide whether toexert effort and increase project successprobabilities. Banks can improve firms’ moralhazard problem via costly and non-observablemonitoring. Banks cannot commit to monitor.Their incentives to monitor depend on whetherthey lend exclusively or share financing withother banks. Multiple-bank lending entails twoeffects. First it improves banks’ monitoringincentives by allowing banks to finance moreprojects and therefore achieve greaterdiversification. Second, it entails free-ridingproblems and duplication of efforts, thusreducing banks’ incentives. Multiple-lendingis optimal whenever the first effect dominatesthe second. The authors find that multiple-lending is more attractive as the cost ofmonitoring increases and less attractive forlarge amounts of inside equity and projectprofitability. Thus the model predicts a greateruse of multiple-bank lending when banks aresmall relative to the projects they finance,when firms are less profitable and when poorfinancial integration, strict regulation andinefficient judicial systems make monitoringmore costly.

Stefan Arping (University of Amsterdam)presented the paper “Playing hardball:relationship banking in the age of creditderivatives”. Credit derivatives may reduce theincentive for banks to monitor firms theyfinance, as it reduces banks’ exposure to risky

ventures. In this paper, the author provides atheory for the widespread use of creditderivatives. He argues that credit derivativesallow banks to improve on their oversightduties by giving them an exit option, therebymaking it less costly to penalise misbehavingborrowers by letting them fail. Therefore creditderivatives strengthens banks’ commitment toengage in timely intervention, thus givingborrowers the proper incentives ex-ante. Inaddition, he shows that by taking derivativesthat expire before loan maturity, banks providea termination threat. However, this threat isonly credible for those banks that are highlycapitalised as they are most able to sustainlosses. In this context, credit derivativesfacilitate the optimal dynamic management ofclient relationships in banks’ core loanbusiness, thereby promoting value creation inthe real sector. In his setting, the introductionof a viable credit derivatives market can createvalue on purely incentive related relationshipmanagement grounds that are unrelated tocapital or financial constraints at the banklevel.

Bruno Parigi (University of Padua) discussedthe three papers. On the first paper, he notedthat the sample was formed of very small firmswith relatively few alternatives to relationshipbanking. An interpretation of the results couldtherefore be that when competition increases,the bank seeks to retain customers bystrengthening its relationship with them. Heregretted that loan rates were unavailable, as anadjustment in loan rates would be expectedfrom an increase in competition. However,research from the authors (Distance, lendingrelationships and competition) suggests that anincrease in distance or competition decreasesloan rates. Together with the result thatrelationship lending suffers from increasingdistance, one could surprisingly conclude thatrelationship lending is associated with higherloan rates.

On the second paper, Parigi noticed that bankswere modeled as being risk neutral, whichappears at odds with the interpretation of the


results in terms of diversification. He furthersuggested making capital constraints and otherfactors that might generate concavity in banks’objective functions explicit. Finally, he alsopointed out that one implication of the modelmay be that more leveraged banks may be safer.In particular, bank’s lending was limited by theintroduction of capital constraints. Bydecreasing this constraint, banks are able tofinance more projects and become moreleveraged. More leveraged banks are morelikely to enter into multiple lending and as aconsequence become more diversified andsafer. Simulations with different levels ofcapital constraints were therefore encouraged.

On the third paper, Parigi noted that the modelassumed the payment of a fixed cost by thebank to observe fully the effort level, so that theproper incentive on effort can be provided at nomarginal cost. As a consequence, banks andfirms form a coalition at no cost. He thereforeasked whether a richer framework allowingfirms to choose riskier projects with higherreturns would modify the bank-firm coalitionchoice of effort. In particular, he stressed that abank protected by credit derivatives may preferthe firm to choose a riskier project. Moregenerally, he argued that, as shown in Arping’spaper, credit risk transfer modifies problems ofasymmetric information between borrowersand lenders in a way that deserves moreattention. Parigi concluded his discussion bycommenting on the recent scandals from Enronand Parmalat. According to him, these eventspoint to pervasive conflicts of interest fromrelationship banking, which are not yet fullyunderstood.

SESSION 2.1 INTERNATIONAL FINANCIALLINKAGES – CAPITAL FLOWS

The session was chaired by Holger Wolf(Georgetown University). Rui Albuquerque(University of Rochester) presented the paper,“International equity flows and returns: Aquantitative equilibrium approach”, co-authored with Gregory Bauer (University ofRochester) and Martin Schneider (New York

University). The paper tries to explain threestylised facts typically observed in equitymarkets. The flow momentum, according towhich a net capital inflow into a countryanticipates a net future inflow. The burst andgross trading activity, which amounts to apositive contemporaneous correlation betweengross purchases and sales. The return chasing,i.e. if prices increase, the average foreigninvestor buys shares from the average localinvestor. They show that investorsheterogeneity is crucial in order to explain thedata on international portfolio choices. Theypropose a model of international portfoliochoice with heterogeneous investors bothwithin country and cross-country. Their mainfinding is that within country heterogeneity ismuch more important than cross-countryheterogeneity as a model that match the datawell must have the property that cross-countrydifferences between average trades are muchsmaller than within-country differencesbetween trades of sophisticated andunsophisticated investors.

The paper, “Does regulatory harmonisationincrease bilateral asset holdings?”, waspresented by Jonas Vlachos (The ResearchInstitute of Industrial Economics, Stockholm).The study explores the importance ofdifferences in securities regulation on assetmarket integration. The estimation of anempirical gravity model, where data recentlymade available by the IMF are used, revealsthat cross-border asset holdings do increasesubstantially with the harmonisation ofsecurities regulation. More precisely, theauthors construct an index of regulatorydifferences, building on the work of Laporta,Lopez-de-Silanes, Schleifer (2003). The sizeof the coefficient multiplying an indexmeasuring the improvement of domesticsecurities regulation suggests that an increasein this index by one standard deviation (3.77points) will increase foreign portfolio holdingsby about 40 percent. Also a country thatimproves its regulation and moves itssecurities regulations by one standarddeviation towards all other countries’

131ECB


ANNEX F

regulatory framework would experienceanother 150 percent increase in foreign assetholdings. Although institutional or culturaldifferences have negative impacts on bilateralasset holdings, results for regulatorydifferences are robust even after taking theseeffects into account. Moreover, it seems thatregulation is used to protect domestic markets.

Michael Halling (University of Vienna)presented the paper “Where is the market?Evidence from cross-listings”, co-authoredwith Marco Pagano (University of NaplesFederico II), Otto Randl and Josef Zechner(both University of Vienna). Conventionalwisdom says that the integration of capitalmarkets has been steadily increasing over time.However, this is inconsistent with the growingnumber of companies listing their shares notonly in their domestic stock exchange but alsoon foreign exchanges. The study investigateswhether frictions such as trading costs,informational barriers or regulatory obstaclesencourage cross listing. The research finds thatcross listing initially raises trading volume inforeign markets, but a trend decline follows(the so-called “flow-back” phenomenon).Although this would suggest a return to thedominance of the domestic market, for certaincompanies the decline of foreign trading tendsto be quite slow. Foreign trade volume turns outto be higher for export-oriented companies andfor companies which cross-list into foreignexchanges with lower trading costs and betterinsider trading protection. Also small, high-growth and high-tech firms tend to haverelatively higher foreign trading activity.Finally, the presence of strong inertia in tradingis also detected.

The discussant, Philip Lane (University ofDublin, Trinity College) stressed that thecommon denominator of the three papers wasthe recognition of the limited degree ofinternational financial integration. The firstand third study paid special attention toinvestor behaviour. Lane evidenced that thefirst paper provides an interesting theoreticalunderpinning, but the determination of asset

prices is entirely left to domestic investors,while foreign investors are not influential. Infact, national security prices are affected byinternational factors. Moreover, the modelcould incorporate the national differences inprivate opportunity, consumption pattern,labour income risks, and tax system. As for thesecond paper, the discussant pointed out thatalthough it tackles the harmonisation ofregulations in a novel way, when measuringharmonisation it applies weights to neitherindividual regulation, nor to other controlvariables. Furthermore, the role of factors suchas exchange rate volatility, currency unionsand tax treaties could also enter the analysis.Lane highlighted that the third paper is animportant contribution to the extensiveliterature on cross-listed shares. He wouldinclude in the model time-zone differences, thedegree of return correlation in foreign versushome markets and perhaps country fixedeffects to understand why firms list their sharesin foreign stock exchanges.

SESSION 2.2 FINANCIAL STRUCTURES,COMPETITION AND GROWTH

The session was chaired by Luigi Guiso(University of Sassari). Stijn Claessens(University of Amsterdam) presented the paper“Competition in the financial sector andgrowth: A cross-country Perspective” with LucLaeven (World Bank). The authors testempirically whether competition in thebanking sector is beneficial to economicgrowth. Banking competition affects the accessto external finance, which has a positive effecton economic growth. In theory, however, it isnot clear how competition affects the access toexternal finance. On the one hand, morecompetition may lead banks to offer morecredit and lower lending rates. On the otherhand, more competition decreases incentivesfor banks to invest in information acquisitionor relationships with borrowers and, by doingso, it tends to limit the access to externalfinance. The authors use cross-country andcross-sector data. They estimate the effects ofan index of the degree of competition in the


banking industry on the average industrialgrowth rate over the period 1980-1997 for 29countries. They control for heterogeneityacross countries and sectors and, in particular,for the degree of financial development at thecountry level, and for the degree of dependenceon external finance at the industry level. Theirresults depend upon the degree of financialdevelopment. In under-developed countries,sectors that are financially dependent growslower when the financial system is morecompetitive, while in developed countriesmore competition is associated with highergrowth. More precisely, financially dependentfirms will grow by 1.5 percent per annum moreif the country’s financial sector is morecompetitive. These findings support the viewthat market power in banking systems might bebeneficial to less developed countries but notfor industrial countries.

In “Financial architecture and technology”Solomon Tadesse (University of SouthCarolina) explores empirically the impact offinancial structure (bank-based or market-based) on technological innovation. The role offinancial architecture in fostering innovationand technology is theoretically controversial.On the one hand, the proponents of the bank-based system argue that banks encourageinnovation by facilitating the financing ofsmall firms and long-term projects. On theother hand, the proponents of the market-basedsystem underscore the advantages of marketsover banks in financing projects with uncertainviability. The author uses a broad cross-sectionof countries with a panel of industry. Heestimates the effects of an index of marketversus bank orientation on the rate oftechnological progress over the period 1980-1995 for 34 countries. He controls forheterogeneity across countries and sectors and,in particular, for the average size, age, andlevel of financial dependence of the firms. Thefindings suggest a nontrivial impact offinancial architecture on industrial innovativeactivities. Market-based financial systemshave an overall positive effect on technologicalprogress. Regardless of their financial

dependence, this effect is common to allsectors of the economy. However, the studyalso finds evidence that financial architecturehas a heterogeneous effect across industries. Inparticular, industries whose small and youngfirms are relatively more dependent on externalfinance fare better in bank based financialsystems. Hence, financial architecture not onlymatters for long-term growth of a country, butalso plays an important role in shaping itsindustrial structure.

Enisse Kharroubi (DELTA and Banque deFrance) presented “Financial integration: forwhom can it be a wrong medicine?” This paperstudies theoretically how financial integrationaffects economic growth. The author developsa theoretical model of a small open economy,where firms can undertake either a safe or arisky project, which can be financed by localand/or foreign banks. The two main differencesbetween these banks are that local banks canobserve which project is financed but have alimited supply of funds, while foreign bankscannot observe the quality of projects but havean infinitely elastic supply of funds. The authorfinds that financial integration may, undersome conditions, lower the growth rate of theeconomy. The reason is the following. Localfirms undertake the safe projects when theyborrow from local banks, because the latterhave information about the quality of theirprojects. This happens as long as firms do notneed to resort to foreign funds. When localfirms borrow (a relatively large amount offunds) from foreign banks, however, an adverseselection problem between foreign banks andlocal firms arises, where firms take more risksand the external finance premium is high. Inthis case, local banks’ assets become riskier aswell, and local depositors require higherdeposit rates, therefore increasing further thelocal lending rates and depressing economicactivity. It follows that financial integrationmight be detrimental to growth for countrieswhose domestic supply of funds is too low.

Yrjö Koskinen (Stockholm School ofEconomics) discussed the three papers. On the

133ECB


ANNEX F

first paper, he mentioned that the degree ofcompetition may be determined by the ease ofobtaining information about firms, andtherefore on firm opaqueness. He thereforesuggested controlling for this effect, forinstance be adding a variable on accountingstandards. He also wondered how the presenceof stock markets would affect the results. Onthe second paper, he suggested that the authorgoes beyond the effect of financial architectureon technology to study the interaction betweenfinancial architecture, dependence anddevelopment. He also proposed to usealternative measures of how efficient afinancial system is, such as the relative numberof IPOs taking place in a single year. On thepaper by Kharroubi, he suggested two possibleextensions to the version that was presented.First, the results may be modified if the returnon projects is dependent on exchange rates.Second, as it is assumed that domesticintermediaries cannot borrow from abroad, heasked whether the results would hold if foreignborrowing was allowed.

MEASURING FINANCIAL INTEGRATION IN THEEURO AREA

Annalisa Ferrando and Peter Hördahl (ECB)presented the paper ”Measuring FinancialIntegration in the Euro Area” (co-authoredwith Lieven Baele, Ghent University,Elizaveta Krylova, and Cyril Monnet, bothEuropean Central Bank), which prior to theSymposium had been released as an ECBOccasional Paper. In the presentation, it waspointed out that the purpose of the paper hadbeen to provide a comprehensive overview ofthe state and evolution of financial integrationin five key euro area markets, namely themoney, government bond, corporate bond,bank credit, and equity markets. In order tomeasure the degree of integration, a carefuldefinition of the concept “financialintegration” was provided, according to whicha market is considered fully integrated if allpotential participants in this market faces asingle set of rules, have equal access to it, andare treated equally when active in the market.

In the case of integration of euro area markets,this would imply that there should be nodiscrimination among market participantsbased on their location/country.

A methodological framework for measuringfinancial integration was then presented, inwhich three types of measures were proposed:price-based measures, which include simpleyield differentials and unadjusted or risk-adjusted return differentials; news-basedmeasures, which measure the systematicresponse of returns to common shocks orfactors; and quantity-based measures, whichinclude e.g. flows and home-bias indicators.The first two types of measure, which make upthe bulk of the analysis in the paper, restslargely on a key implication of the adopteddefinition of financial integration, namely thelaw of one price. The empirical evidence thatwas presented confirmed that different marketsectors have attained different levels ofintegration. The money market enjoys near-perfect integration (even if e.g. the reposegment shows signs of being less integratedthan the unsecured segment). The degree ofintegration in the government bond market wasfound to be high, but not perfect, although itwas pointed out that it is difficult to distinguishbetween a remaining marginal lack ofintegration and small but systematic effectsarising from differences in perceived creditrisk or liquidity risk among markets in differentcountries. The corporate bond market wasdeemed to be reasonably well integrated, whilethe equity and bank credit markets were foundto be less integrated, although improvementshad taken place to some extent since theintroduction of the euro.

The discussant, Michael Melvin (ArizonaState University), started by commending theauthors for providing a comprehensive anduseful study of euro area financial integration.One aspect of the study that he foundparticularly appealing was that it used data thatis not publicly available, such as interest ratetransactions data reported by banks belongingto the EONIA panel. He pointed out that due to


the confidentiality of this data, studies such asthis one provided an important source ofinformation. He urged the ECB to continuework with this kind of data along the lines in thepresent paper. The discussant then continuedby providing some questions and suggestionsrelating to the analysis in the paper. He firstasked whether it was reasonable to place asmuch emphasis on the law of one price as theauthors had done. He pointed out that even ifmarkets are fully integrated, returns need notbe the same across assets, and that countryeffects can remain due to e.g. liquiditydifferences, regulatory differences, or evenbehavioural biases. Regarding the news-basedmeasures of integration, he called forrobustness checks, in which the effect ofimportant exogenous news (such as monetarypolicy announcements) could be used tomeasure the response in different countries,rather than relying only on asset returns of abenchmark country. Finally, he voiced theopinion that capital flows may perhaps be amore useful source of information on financialintegration than price-based measures.However, in the ensuing discussion followingthe presentation, Bob Flood pointed out that hewas sceptical of the usefulness of flow data forthis purpose. The reason, he explained, wasthat he felt that it is very difficult to interpretflow data in terms of integration. Sometimes avery high degree of country-activity may signalthat markets are integrated, while in other casesmarkets can be fully integrated even thoughsuch activity remains very low or even non-existent.

135ECB


ANNEX F

Jan-Pieter Krahnen (CFS) introduced RobertFlood (IMF) who gave a Key Lecture on Newapproaches to assess financial integration,based on work with Andrew Rose (Universityof California, Berkeley). The objective of theLecture was to propose an intuitive measure ofasset-market integration. According to theauthors, financial markets are integrated whenassets are priced by the same stochasticdiscount rate. That is, security markets areintegrated if all assets on those markets arepriced according to the usual pricing equationpt = Et (MRSt+1

χt+1), where MRS stands for

marginal rate of substitution and where pricesof a portfolio equal to the expected incomereceived by the owner of this portfolioweighted by the intertemporal marginal rate ofsubstitution for income accruing in the future.According to the law of one price for assets,portfolios with the same risk and returncharacteristics should have the same price.This implies that the MRS for both portfoliosshould be the same. Therefore if two assetmarkets are integrated, any portfolios with thesame risk/return characteristics should havethe same MRS. As a consequence, a necessarycondition for asset markets integration is thatthe two values of the first moment of the MRSused to price the two portfolios are identical.Flood and Rose propose a new test to checkwhether this is the case. They only concentrateon the first moment, as they argue it is rathersimple to measure, and cross-marketdifferences in estimated values of the firstmoment allow them to use standard risk pricingmodels.

Flood then described their methodologyand how it compares to the usual financeapproach. The asset pricing equation abovecan be easily rewritten as

,

where ( ) zxEx ttt )( 111 +++ −=ε is a normalisedprediction error, and δ t = 1/Et (MRS t+1). In anintegrated market, δ t is identical for all assets.The traditional finance approach uses z = pt,and ads two assumptions to the resulting

DAY 2equation; 1) rational expectations, according towhich the residual is uncorrelated withinformation available at time t and 2) acovariance factor model such that,

where f is a

vector of time varying factors. With these twoassumptions, the equation can be estimated.However, currently available estimationtechniques do not allow identifying both theasset-specific intercept and δt. Instead, theusual finance approach sets one of these twoparameters to a fixed value. This is obviouslyunsatisfactory. Flood and Rose rather proposeto set z � pt, so as to be able to identify andestimate δ t (for their purpose they choosez = pt-1,). The idea is to choose something thatstabilises the data and preserves theinformation in the current price, while stilldelivering moments in the covariance term thatcan be modelled as stable functions of a fewaggregate sources of risk. Then the test ofintegration is simple. Estimating the equationfor a set of assets J and then repeating theanalysis for the same period of time with adifferent set of assets K gives two setsestimates of δ, a time-series sequence ofestimated discount rates. These can becompared directly, using conventionalstatistical techniques, either one by one, orjointly.

Flood then presented results for the integrationof the S&P stocks priced in the NYSE andNASDAQ, using daily data from the “USPricing” database of Thomsen Analytics,covering April-May 1999. Using 20constructed portfolios from the S&P, Flood andRose built two sets of 10 portfolios each toestimate discount rates. Two striking featuresemerge form their analysis. First the time seriesvariation in δ is high, which can be used toreject the hypothesis that the MRS is equal tothe short t-bill rate as assumed by the financeliterature. Second, the estimates of δ from thetwo sets of portfolios are statistically similar,consistent with the integration of the S&P.Several other periods were considered as a

11

11 ),( +

++

+ +−= t

t

t

t

t

t

z

xMRSCOV

z

p

z

x εδ

COV (MRS t+1,χ t+1) = β0 + ∑ βi fi,t,z i


robustness check, reaching the sameconclusion. The same procedure wasconducted for the NASDAQ, grouping datafrom 100 NASDAQ firms into 20 portfolios of10 firms each. While the NASDAQ appearsgenerally integrated, integration is rejected forOctober and November 1999, shortly before thecollapse of the NASDAQ. Then, the authorschecked whether the market for large stocks(S&P 500) is integrated with the NASDAQ,comparing the δ when estimated with twentyS&P portfolios, twenty NASDAQ portfoliosand all forty portfolios pooled together. Thetest results are grossly inconsistent with thehypothesis that S&P and NASDAQ are twointegrated markets.Jan-Pieter Krahnen then opened the floor forquestions. Garry Schinasi (IMF) wonderedhow the distribution, from which Flood wasextracting moments, differs from thedistribution generally used in finance. Fromthe floor, it was remarked that the estimated δmoves a lot over time and it was asked whetherthe choice of daily data could generate theselarge movements. Flood replied that, unlike infinance, the noise is actually a good thing sinceit helps them to nail down precisely the δ. Itwas then asked whether the authors consideredstocks that were traded on both NYSE andNASDAQ and whether Flood could elaboratemore on the distribution that the δ is capturing,since the results are difficult to believe. MarcoPagano (University of Naples) wonderedwhether this methodology could be applied toan environment with heterogeneous investors,which would imply that the marginal investoris different for different sets of instruments.Flood replied that heterogeneity is perfectlyconsistent with their methodology. Althoughhe reckoned that an important assumption isthat there is no restriction in trading, so that animportant first step in testing whether twomarkets are integrated is to check theinstitutional environment. He then continuedexplaining that these measures of integrationare to be taken as an upper bound on the degreeof integration two markets can achieve, as it isdifficult to conceive two markets that would bemore integrated than NYSE and NASDAQ.

However, Flood also admitted that theirmethodology could not tell the differencebetween integration and liquidity.

SESSION 3.1 TOPICS IN FINANCIALINTEGRATION

Francesco Drudi (ECB) chaired this sessionand introduced the first speaker. Mark Carey(Federal Reserve Board) presented the paper“Is the corporate loan market globallyintegrated? A pricing puzzle”, with Greg Nini(Federal Reserve Board). This empirical paperprovides evidence that prices of syndicatedcorporate loans differ between the Europeanand U.S. markets. More precisely, using datafrom Dealogic’s Loanware database from 1992to 2002, the authors find that interest ratespreads are, on average, 30 basis points smallerin Europe than in the U.S. The authors testedseveral hypotheses that could explain this gap.First, different characteristics of loans andborrowers differ across the sub-samples.However, controlling for such differences, andother factors such as non-price terms of loans,asymmetric information or moral hazard, legalregime, multi-product package pricingpractices and regulation, does not explain thedifference fully. The puzzle subsists even afterchecking for potential errors in the data.However, the authors cannot reject thehypothesis that there is market inefficiency inthe sense of myopic behavior by marketparticipants. However, given the type oflenders and borrowers participating in thismarket, the authors disregard this possibility asbeing very unlikely. The conclusion from theresults of the paper is that the syndicated loanmarkets in Europe and the US might not befully integrated, and the authors presume thatthe explanation for the home bias they observein the data may be an important factors tounderstand the pricing difference.

Thomas Harr (University of Copenhagen)presented a theoretical paper “Branch orsubsidiary? Capital regulations ofmultinational banks” with Thomas Roende(University of Copenhagen). It looks at optimal

137ECB


ANNEX F

capital requirements for multinational bankswith an information advantage over theregulator. (The bank knows better than theregulator how risky it is.) The bank can choosea branch structure or a subsidiary structure.Before the bank selects its structure, theregulator chooses a capital requirement forbanks that select a branch structure and anothercapital requirement for banks that select thesubsidiary structure. It is shown that thereexists an optimal screening mechanism inwhich the regulator chooses a higher capitalrequirement for banks with a subsidiarystructure. Since a subsidiary structure limitsthe liability of the bank and thus is preferablefor riskier banks, banks with subsidiaries needto have higher capital requirements.

Jens Tapking (ECB) presented a theoreticalpaper “Horizontal and vertical integration insecurities trading and settlement”, joint withJing Yang (Bank of England). It looks atwelfare implications of two different types ofconsolidation in the securities trading andsettlement industry, mergers of an exchangewith a Central Security Depository (CSD)(vertical integration) and mergers of two CSDs(horizontal integration). Though the model ofthe paper is quite complex, the results arestrikingly simple: Both types of integrationincrease the economic welfare compared to noconsolidation, but the welfare improvement isgreater in case of a horizontal merger. Thepresenter emphasised that this result dependson two crucial assumptions of the paper.Firstly, it is assumed that investors have strongpreferences for securities issued abroad. Ifinstead investors preferred home securities,vertical integration would outperformhorizontal integration. Secondly, it is assumedthat all exchanges are bound to settle in a givenCSD and cannot choose among different CSDs.The results of the paper could change if thisassumption was not used.

The discussant was Harry Huizinga (TilburgUniversity). On the first paper, he raised doubtsthat the authors really controlled for allrelevant variables. For example, it was pointed

out that it was not possible to control preciselyfor the nationality of lenders and borrowers. Onthe second paper, he stressed that in realitymany banks have a hybrid structure withbranches and subsidiaries. It may therefore bedifficult to apply the results of the paper.Furthermore, he argued that larger banks aresafer and it might therefore be more efficient tolet the capital requirements depend on sizeinstead of branch versus subsidiary structure.On the third paper, Harry Huizinga made twosuggestions. Firstly, he asked if the high levelof economic welfare achieved by a horizontalmerger could also be achieved if regulatorsforced the CSDs to charge a price of zero for thetransfer of securities through links. Secondly,he suggested that the authors analyse in theirmodel if it can be expected that market forceslead to the optimal form of consolidation.

On relation to the Carey-Nini paper, Jan-Pieter Krahnen (CFS) wondered whetherusing only companies on which only externalratings are available could put structure on thedata set that should be considered in theregression. Another explanation for the pricingpuzzle might be the extent of relationshiplending that is known to have effects on loanprices. Carey agreed that there might be aselection bias, since loan size is bigger inEurope than in the US but not significantly so.On relationship lending, he replied that theyaddressed this issue by considering both thenumber and nature of lenders in a syndicate, onthe presumption that when there is a largenumber of lenders, the marginal lender in asyndicate is less likely to have relationshiplending. They find that this factor is priced aswell, but not enough to solve the puzzle.

SESSION 3.2 INTERNATIONAL FINANCIALLINKAGES – COMOVEMENTS

Michael Binder (University of Frankfurt andCFS) chaired the second session on“International Financial Linkages”, dedicatedto asset price co-movements. Clara Vega(University of Rochester) talked as the firstspeaker about her paper “Real-time price


discovery in stock, bond and foreign exchangemarkets” (joint with Torben Andersen,Northwestern University, Tim Bollerslev,Duke University, and Francis Diebold,University of Pennsylvania). Using intradayfutures data over the last 5 to 10 years, thepaper estimates in the first place the effect ofmacroeconomic news releases on differentasset prices. News releases are limited to theUnited States, whereas euro area, UK and USmarkets are covered for stocks and bonds, andthe Japanese yen is also added for exchangerates. In a first step the impact effect of thesurprise component of news releases onindividual asset returns is assessed. In a secondstep, returns are estimated simultaneously asfunctions of contemporaneous and laggedreturns of other assets and news surprises. Afirst result of this analysis is that by usingintraday data the effects of macroeconomicvariables on asset returns can be identified in amuch clearer way than in the previous literatureusing lower-frequency data. Second, positivenews on the economy can have different effectson stock market returns, depending on the stageof the business cycle. During the expansion of1998 to 2001 “good news” tended to increasestock returns, whereas during the downturn of2001 to 2002 they tended to decrease stockreturns. The same does not apply to bond yieldsand dollar exchange rate returns, whichunambiguously increase in response to positivenews. Finally, the authors identify significantcross-country linkages between stock marketsthat are not explained by US macroeconomicnews.

The second paper of this session on“Commonality in the time-variation of stock-bond and stock-stock return comovements”was presented by Robert Connolly(University of North Carolina). The paper,which is co-authored by Chris Stivers(University of Georgia) and Licheng Sun (PennState Erie), discusses how cross-country stock-stock and domestic stock-bond return co-movements vary with stock marketuncertainty. One approach used comparesreturn correlations across the different

quintiles of the distribution of stock marketuncertainty, as indicated by implied volatilitiesin major stock markets (measured by theChicago Board of Trade Option ExchangeVolatility Index (VIX) and the German OptionVolatility Index (VDAX)). Another approachuses a GARCH model to estimate changes incorrelations. And, finally, a two-state regime-switching model is applied, in which transitionprobabilities between the different states canvary as a function of stock market uncertainty.The data is daily, covering Germany, theUnited Kingdom and the United States between1992 to 2002. The results for cross-assetreturns provide evidence in favour of flight-to-quality and flight-from-quality behaviour.When implied stock market volatility is high orincreases, then stocks and bonds tend to benegatively related. When stock marketuncertainty is low, then stocks and bonds arepositively related. Moreover, in times of highuncertainty cross-country stock marketcorrelations tend to be particularly strong.Overall, these asset linkages seem to be drivenby global rather than domestic developments.

The last paper of the session on “The breadth ofcurrency crises” focused on co-movementsamong industrial country exchange rates,among emerging market exchange rates andbetween industrial country and emergingmarket exchange rates. It was presented byPhilipp Hartmann (ECB), who co-authors thepaper with Stefan Straetmans (MaastrichtUniversity) and Casper de Vries (RotterdamUniversity). In contrast to the first two papersof the session, this paper puts all the emphasison very extreme returns and, hence, crisislinkages. To do so, it presents a newmethodology how to estimate semi-parametrically multivariate asset return co-movements based on extreme-value theory.From this methodology the authors derive,inter alia, a “contamination function” thatillustrates the probability of having manymarkets crash, given that a specific number ofmarkets crash. Weekly returns for 5 industrialcountry currencies and 10 emerging marketcurrencies between 1987 and 2003 are

139ECB


ANNEX F

considered. One first result suggests that –contrary to conventional wisdom – the breadthof extreme currency spillovers are not greateramong emerging market currencies than amongindustrial country currencies. (Univariately,however, emerging market currencies crashmuch more often than industrial countrycurrencies.) Second, lagged extreme spilloverstend to be generally weaker thancontemporaneous extreme spillovers. Third,extreme spillovers between industrial countryand emerging market currencies are relativelyweak. Finally, there are a number of currenciesthat constitute “hot spots” in the foreignexchange market, i.e., they are more oftenassociated with widespread currency turmoilthan other currencies.

The discussant, Paolo Pasquariello(University of Michigan), pointed out that co-movements of asset price are a very importanttopic in financial economics, with implicationsfor risk management and portfolio allocations.He felt that all three papers addressed differentfeatures of this topic in a rigorous way. As all ofthem are primarily of an empirical nature, hethrew a more theoretical perspective oninternational financial linkages, focusing on“excess co-movements”. Based on one part ofthe financial contagion literature, he definedexcess co-movements as a situation in whichasset prices move together beyond the degreejustified by their economic fundamentals.While the Vega et al. paper was identifyingsuch excess co-movements, the Connolly et al.and the Hartmann et al. papers do not controlfor fundamentals. On the Vega paper thediscussant cautioned that ignoring non-USnews could bias the coefficients in theestimations. He also pointed out that basing theinterpretation of the stock market results ononly one upturn and one downturn constitutedrelatively limited evidence. On the Connollypaper he criticised that the implied stockmarket volatilities used may not be a goodmeasure of asymmetric information, whichaccording to the paper is one theoreticalexplanation for the spillovers observed.Moreover, referring to some of his own work,

changes in risk aversion are not sufficient forexcess co-movements in rational expectationsmodels, and he was interested in hearing whichother frictions are responsible for thecorrelations identified. On the Hartmann paperhe challenged the association of very extremereturns with crisis situations, as a full-blowncrises should also be associated with largesocial and economic costs. He also suggestedthat the authors would account for exchangerate regimes in the different currencies.

Michael Binder opened the floor for question.Hartmann made two comments on the Vegapaper. First, he suggested allowing theparameters to change with the introduction ofthe euro, as the strength of return and newsspillovers may have changed with EMU.Second, a discussion of some recent work byCarlo Favero (Bocconi University, Milan)could be useful, as this author argues that US-European stock market spillovers seem to berelated to inefficient propagations of USinnovations, whereas the same does not holdfor bond market spillovers.

POLICY PANEL – DRIVERS OF EUROPEANFINANCIAL INTEGRATION: MARKETS ORPOLICY?

Tommaso Padoa-Schioppa (ECB) opened thepanel session on “Drivers of EuropeanFinancial Integration – Markets or Policy?”After briefly outlining the current status offinancial integration, he turned to the centralquestion to which extent markets or policymakers should determine the degree ofintegration of markets. In his view, this is anaspect that should be left to markets. However,the market should not face any legislative orregulatory barriers in its development towardsa more integrated market. To this aim, it wasessential that policy created a framework inwhich markets could operate towards theoptimal level of integration. Mr. Padoa-Schioppa mentioned two areas in which thepolicy framework needed improvement. Thefirst concerned the implementation of acommon rulebook. He stressed that it was


crucial that the FSAP measures were translatedefficiently into a consistent EU rulebook andenforced coherently across the EU. Hecriticised that so far, policy makers had placedemphasis on rule-making, but put little focuson competition policy aspects. The latter couldensure that national authorities would not try toprotect national markets. The second aspectconcerned the enforcement of rules. Here, Mr.Padoa-Schioppa mentioned two objectives: onthe one hand, a convergence of supervisorypractices, and second, a higher level of co-operation between different supervisoryagencies.

Mario Draghi (Goldman-Sachs) assessed thestatus of integration from the perspective of awholesale market institution. The degree ofintegration was very diverse in the differentmarket segments – while equity marketsshowed a strong home-bias, bond markets andderivative equity markets, for instance, werebasically unified. Pan-European activity wasincreasing, but hard to quantify (because largeparts of the markets are done over-the-counter)and concentrated in very few market placessuch as London, Frankfurt, or Paris. Mr. Draghipointed out the conventional wisdom that awholesale institution already operating at aEuropean level may be content with thesegmentation of markets because of its uniquepotential to profit from European-widearbitrage. However, he asserted that it was stillmore likely to profit from more integrationbecause it would benefit from deeper and moreliquid markets. He criticised that none of theFSAP proposals had so far been implemented atthe national level. In his view, responsible forthis delay was mainly the strong desire ofnational regulators to maintain competencies.

Alexander Schaub (European Commission)stressed that both markets and policy processeswere important in the development ofintegration, but that market forces had beenplaying a more dominant role. Especially untilthe launch of the FSAP in 1999, policy was notthe driving force, as the development offinancial markets was not on the political

agenda. Market forces, on the other hand, didinfluence policy, among other factors, becauseof increasing globalisation – an example wasthe case of Enron, which led to areconsideration of EU rules.

Mr. Schaub then proceeded to comment on thefuture of the FSAP. In his view, the mostimportant challenge was to eliminate obstaclesto integration by eliminating or harmonisingnational rules. As examples, he mentioned agrowing need for broader market access rights,to address new sources of European-widefinancial risk, for consolidated supervision ofconglomerates, and to address loopholes in thelegislative framework. The degree ofharmonisation, however, was an aspect thatshould not be determined ex-ante but ratherdevelop over time. He criticised thatpractically none of the 42 FSAP measures hadso far been effectively implemented in nationalrulebooks. However, the mere announcementof introducing these measures already had aneffect because of the expected implementation.He asserted that the future of the FSAP wouldbe driven further by market forces. Fourworking groups consisting of market playershave already issued their first reports, whichwill be an important input for future work.

Jens Thomsen (Danmarks Nationalbank) firstaddressed the lack of integration in retailmarkets. He stressed that reasons were mainlydue to differences in language and consumer’srestricted knowledge of information on foreigncompanies. This led to a domestic attitude formany consumers. He then proceeded to thetopic of market or policy driven integration. Asan example for market forces leading to moreintegration, he mentioned the MTS electronicplatform. Finally, he commented on the SecondBanking Directive. He criticised that theDirective remained silent about the regulationof bank branches located in different countries.For a small country like Denmark, this topicwas of great relevance since the market share offoreign banks in Denmark was extremely high –about 25% of the market being covered by theSwedish bank Nordea alone.

141ECB


ANNEX F

Mr. Padoa-Schioppa then opened the floor forquestion. Garry Schinasi (InternationalMonetary Fund) mentioned differencesbetween Europe and the US in the regulatoryframework. In his view, the degree of politicalunion was the main difference between the twoareas. He furthermore questioned whether lackof integration was really the biggestinefficiency in EMU. Similarly, Jesper Berg(Danmarks Nationalbank) conjectured thatlarge European banks might not be interested indiversifying across Europe, but rather on aglobal level. Philipp Hartmann (ECB) askedSchaub about the main messages that hademerged in the FSAP expert groups. He alsoinquired what may have led Nordea toannounce the adoption of a branching structure.Mr. Thomson replied that most likely, Nordeahas difficulties achieving scale economieswhen operating with subsidiaries. It was moreefficient to also unify the banks’ headoperations. He also mentioned that this willhave an impact on supervision, sinceconsolidated supervision of the bank will nowbe necessary.

Mr. Padoa-Schioppa asked Mr. Schaubwhether competition policy may be allowed toplay a greater role. Mr. Schaub answered thatpressures on the Commission fromgovernments and lobbyists were very strong, sothat that it was likely that “weak”commissioners not urging for morecompetition policy were put in place.

Several remarks from the floor addressed thedrivers of both policy makers and marketplayers for enhancing integration. Jan-PieterKrahnen (CFS) asked Mr. Draghi whetherintegrated or separated markets were moreprofitable for an institution like Goldman-Sachs. Also, Mr. Padoa-Schioppa’s wasinterested in the reasons why pan-Europeanmarket players did not push policy makersmore for removing obstacles to integration. Mr.Draghi replied that it was hard to judge whethermost of these market players actually benefitedfrom fragmentation of markets, for instanceby profiting from tax arbitrage. Still, he

acknowledged that if anyone profited fromsegmentation, it was likely to be large players.

Gertrude Tumpel-Gugerell (ECB) thenclosed the Policy Panel with some remarks on“The role of the ECB in financial integration”.Ms. Tumpel-Gugerell explained that the role ofthe ECB in promoting European financialintegration is threefold. First, it co-operatesand acts as a facilitator or catalyst betweenseveral parties. Second, in fields where themandate of the ECB applies directly, itimplements structural reforms that fosterfinancial integration. Third, it raises awarenessby doing specific analysis and assessment ofthe current state of financial integration. Shethen went on explaining each role in moredetail.

Regarding the first role, the ECB regards theco-operation with the European Commission asof particular importance. The ECB participatedactively in the production of the two reports ofthe Giovannini Group, about the structure ofthe security settlement industry in Europe. Shereported that while the ECB fully supports theconclusions of the reports, consolidation isslow and market forces may be insufficient inthis sector.

Regarding the second role, Ms. Tumpel-Gugerell highlighted several examples of ECBmeasures that foster financial integration.Obviously, a uniform interest rate for the shortterm interbank market, was a majorcontribution of the ECB to the integration ofthe money and bond markets in January 1999.Also, she explained that the ECB supports astandardised and safe contractual basis that canbe used in different national jurisdictions toimprove the integration of the euro repomarkets. In particular, the ECB will introduce a“Single List” in the collateral framework toreplace the current two-tier system of eligiblecollateral, in order to ensure a level playingfield in the euro area, to further promote equaltreatment of counterparties and issuers, andto increase the overall transparency of thecollateral framework. Finally, she mentioned


the next generation of the Eurosystem paymentsystem, TARGET 2, which will contribute tofurther financial integration by consolidatinginfrastructures in the market for large-valuepayments and by homogenising features at theuser end.

Regarding the third role of the ECB consistingof raising awareness, Ms. Tumpel-Gugerellstated that research is more than just raisingawareness and welcomed the ECB-CFSResearch Network on “Capital Markets andFinancial Integration in Europe” as the mostextensive forum through which the ECB carriesout and stimulates its research work onfinancial integration.

She then turned to the past and future work ofthe ECB-CFS Research Network, and how theyrelate to the objectives of the ECB in terms offinancial integration. She mentioned andelaborated further on three achievement of thenetwork. First, the network successfullyestablished itself as a “network of people”. Inher view, the Network provided the necessarystructure to exploit synergies and cross-fertilisation between researchers fromdifferent institutions. The results of this type ofrepeated interaction is a higher productivityand output that goes beyond what can beexpected from the simple one-off organisationof traditional seminars and workshops.

Second, the network kicked off a new researchfield on securities settlement systems, one ofthe big challenges for further financialintegration in Europe.

Last, Ms. Tumpel-Gugerell mentioned thatresearch conducted within the networkimproved the knowledge about Europeanfinancial integration. First, regarding theEuropean banking sector, first integrationappears not to be very advanced in retailbanking markets. Second, some of the inherentcharacteristics of traditional loan and depositbusiness constrain the cross-border expansionof commercial banking, even in a commoncurrency area. Hence, the implementation of

some policies to foster cross-border integrationmay be ineffective. There is also increasingevidence that the introduction of the euro hascontributed to a reduction in the cost of capitalin the euro area, for example in the form ofcorporate bond underwriting fees. Monetaryintegration therefore improved access tofinance for investment and prospects forgrowth. Finally, on securities settlementsystems, she reminded the audience of findingsby the Bank of Finland, presented in thenetwork, that securities settlement in Europe ismore than 30% more costly than in the US, andhighlighted two reasons to explain thesenumbers: market power and a lack ofconsolidation. With the absence ofcompetition, there is no pressure to developsystems operating at unit costs. Beyond that,she mentioned another paper presented in thethird workshop that showed that fragmentationand the still relatively limited consolidation insecurity settlement systems, which is a remnantof the epoch with national currencies, preventthe full exploitation of scale economies asshown in the Finnish study.

Ms. Tumpel-Gugerell then announced that theECB Executive Board decided to continue theECB-CFS Research Network for three moreyears. In this new phase, three areas will beadded to the list of network priorities. 1. Therelationship between financial integration andfinancial stability. 2. EU accession, financialdevelopment and financial integration, and3. Financial system modernisation andeconomic growth in Europe. These three areashave become particularly important at thecurrent juncture, but have not receivedparticularly strong attention in the past twoyears of the network. She then completed herremarks saying that the research network willorganise two workshops per year until 2007.

SESSION 4.1 FIRMS FINANCING ANDCORPORATE GOVERNANCE

Jan-Pieter Krahnen (University of Frankfurtand CFS) chaired this session. The sessionstarted with Mihir Desai (Harvard University)

143ECB


ANNEX F

presenting the paper “Institutions, capitalconstraints and entrepreneurial firmdynamics: evidence from Europe” (joint withPaul Gompers and Josh Lerner (both HarvardUniversity)). The paper investigates theimportance of the institutional framework forentrepreneurial activity in Western and Centraland Eastern Europe (CEE region). Measures ofentrepreneurial activity include rates of firmentry and exit and different moments of firms’size and age distribution, such as average size,skewness and industry vintage, a size-weightedmeasure of firm age. Controlling for industryfixed effects and different levels of economicdevelopment, the authors identify a particularsensitivity of entrepreneurial activity toinstitutional factors (corruption/fairness,protection of property rights, well-functioninglegal system) for countries of the CEE region.In particular, less corruption and betterprotection of property rights increase entry andreduce exit. While the average size of the firmdecreases, there is an increase in industryvintage. Finally, descriptive statistics show forthe CEE region that the firm size distribution isskewed to the left with decreasing skewness forolder firms. The authors interpret their resultsas supporting the view that well-designedinstitutions foster entrepreneurial activitypartly through the positive impact on relaxingcapital constraints, which can potentiallyaccount for the results on the skewness ofyoung firms.

In the second paper, “A Theory of friendlyboards” (joint with Daniel Ferreira, StockholmSchool of Economics), Renée Adams(Stockholm School of Economics) addressedthe question why – contrary to the US –European governance structures rely on a“dual” board system. Such a system separatesthe monitoring and advising roles of the boardof directors. The analysis is based on a modelwhere a manager is hired to make an investmentdecision for a risky project from which hederives a private benefit. His ability to identifygood projects is unknown to the board and evento himself. Both the board and the managerreceive signals concerning the quality of the

project. After receiving his signal the managerdecides to reveal his information to the board inexchange of a recommendation which is basedon the board’s signal. Finally, before signalsare received, the board chooses a probability ofmonitoring the manager ex-post. Monitoringenables the board to learn the manager’s signaland to obtain information about the manager’sability. For a given probability of monitoring,the manager faces a trade-off for revealing hissignal. Revealing information and gettingadvice enables the manager to make betterdecision, but this might increase his chances ofgetting fired when his information changes theboard’s opinion about his ability. This trade-off provides a rationale for the board to reduceits monitoring activity up-front whenever it isnot too costly to induce the manager to revealhis information. The authors show further thatthe first-best level of monitoring cannevertheless be attained when the two roles(advice and monitoring) of the board areseparated.

João Santos (Federal Reserve Bank of NewYork) presented the third paper of the session“Identifying the effect of managerial control onfirm performance” (joint with Renée Adams(Stockholm School of Economics). The goal ofthis paper is to re-evaluate the effect ofmanagerial control through voting rights onfirm performance by focusing on voting rightsrather than on share ownership of managers.Shares provide managers not only with controlrights, but also cash-flow rights making itdifficult to disentangle incentive effects frompotentially positive effects of voting rights.The authors use data from a historic sample oflarge banking institutions in the US that holdtheir own stock – as a result of their trustactivities – in the capacity of a fiduciary. Thisenables the authors to construct a proxy formanagerial control rights to measure cleanlythe effects of control, since the fiduciary has nocash-flow rights attached to the shares. Theauthors regress the performance of banks (asmeasured by Tobin’s q) on the proxies formanagerial control rights (and their squares)controlling for firm size, leverage, uncertainty


and historic performance. They find a negativerelationship for small voting stakes, but apositive one for larger stakes, thus confirmingseveral findings in the theoretical literature.Quantitatively, the result seems to beeconomically significant. A change in controlequal to one standard deviation causes changesof up to a third of the standard deviation inTobin’s q.

Giancarlo Spagnolo (University of Mannheimand Sveriges Riksbank) discussed the papers ofthis session. For the first paper, he stressed thatthe analysis provides novel and interestingevidence on the importance of institutionalvariables for firms. Most importantly,institutions seem to matter more for transitioneconomies than for advanced economies. Hestressed, however, that the link betweeninstitutions (and other legal variables) andentrepreneurial activity is still lacking athorough theoretical explanation, which couldpartly account for the partly ambiguous effectof legal variables. Moreover, given the longlags between institutional reforms andinvestment activities, he expressed concernthat the real test of these relationships can onlybe assessed seriously when more data becomeavailable over the next few years. Spagnolopraised the importance of the second paper inestablishing a novel trade-off when boardssimultaneously advise and evaluate managers.As his main theoretical criticism he pointed outthat the results could change when there is acareer concern for the manager. This would addreputation into the model inducing the managerto reveal his information more readily oncerecognised as being of high quality. As minorpoints, Spagnolo pointed out that theseparation of board functions might be relatedto specialised skills, and that preventing theflow of information might be impossible evenin separated boards. For the final paper, heproposed to control for two more factors. First,large shareholders might matter, since theyhave more control rights than managers orfiduciaries have. Second, cash flowconsideration could matter nevertheless, sincethe overall level of the fiduciary’s cash flow

rights could influence the importance of hiscontrol rights.

Jan-Pieter Krahnen then opened the floor forquestions. João Santos was first asked whetherthey had access to firm specific performancemeasures, and if they tried to use thisinformation rather than the measures ofperformance of the bank itself. On the secondpaper, the assumption that supervisory andmanagement boards do not communicate witheach other was questioned, as in reality it is notclear how these boards are synchronised.Renée Adams was then asked whether theyhave evidence of commitment problems fromboards and whether these boards use acommitment device.

João Santos replied that they do not control forwhoever is controlling the rest of the bank, asthey do not have access to such data. Regardingthe positive link between large voting rightsand the value of the institution, Santos repliedthat there are two possible theoreticalexplanations. First, when control increases,individuals are willing to invest more humancapital in the institution, which in turnincreases its value. Second, in case of a hostiletake-over, increased control would imply that ahigher bid is asked for. Renée Adams agreedthat they needed to address the point of careerconcerns. However, she does not think thatboards generally share information. Finally,regarding the commitment technology, shementioned soft (but not direct) evidence forcultural norms in the boardroom that managersare not asked questions on their policy.

SESSION 4.2 SYSTEMIC RISK

The session was chaired by Gary Schinasi(IMF). The first paper, entitled “Bankcontagion in Europe” was presented by ReintGropp (ECB). The paper analyses contagion ina sample of European banks during 1996-2003.The paper builds on the recent literatureexamining large movements in financialmarkets. The paper examines large movementsof the weekly first differenced distance to

145ECB


ANNEX F

default of European banks. It estimates a twostage model: In the first stage, in a Poissonmodel the number of banks experiencing alarge shock at the same time is explained with aset of macro and sectoral risk exposurevariables. The paper shows that the sector riskexposure variables improve the fit of the modeldramatically. Using the unexplained portionfrom this regression, the paper estimates theprobability of an individual bank experiencinga large shock. This approach is intended togenerate a contagion variable that is orthogonalto common shocks affecting more than onebank simultaneously. The paper findssignificant domestic and cross-bordercontagion, although there is no cross-bordercontagion from the smaller banks in thesample. There is, however, significantcontagion from small banks within countries.Overall, the results are consistent with a tieredinterbank market structure, in which due toasymmetric information only the largest banksperform cross-border transactions. Further, thepaper finds that the introduction of the euromay have reduced domestic contagion, whilethere is no increase in cross-border contagion.

In the second talk of the session, GregoryNguyen (National Bank of Belgium) presented“Interbank exposures: An empiricalexamination of systemic risk in the Belgianbanking system”. The paper investigates theevolution of contagion risk for the Belgianbanking system from 1993 to 2002, usingdetailed information on aggregate interbankexposures of individual banks and on largebilateral interbank exposures. The papersimulates the effect of the failure of one bankon defaults of other banks. Overall, the resultssuggest relatively little contagion. The paperfinds that the structure of the Belgian bankingsystem changed from a complete structure (inwhich all banks have symmetric links) to a“multiple money centre bank structure” (wherethe money centres are symmetrically linked tosome banks, which themselves are not linked).The paper suggests that this change may haveresulted in a reduction in the risk and extent ofcontagion. Also, the increase in foreign

interbank assets and liabilities has reduced therisk of domestic contagion but potentiallyincreased the risk of contagion from foreignbanks.

The third paper of the session “An analysis ofsystemic risk in alternative securitiessettlement architectures” was presented byGiulia Iori (Kings College). The papercompares systemic risks in net and grosssecurity settlement systems. It studies thesettlement risk arising from exogenousoperational delays and compares settlementfailures as a function of the length of thesettlement interval under different marketconditions. The paper finds that settlementfailures are non-monotonically related to thelength of settlement cycles under botharchitectures and that there is no clear cutranking of which architecture (net or gross)delivers greater financial stability.

The discussant, Kostas Tsatsaronis (BIS)thought that the three papers were well matchedin one session, as they provided differentperspectives on contagion as a source ofsystemic risk. He emphasised that all threepapers focused on idiosyncratic shocks thatgenerate systemic losses. Clearly, this wasonly one of at least four scenarios: commonexposures, dynamic interaction between realand financial sectors, information contagionand inter-linkages. He doubted that the lattersource, which the papers focused on, was thedominant source of systemic risk. He thenproceeded to discuss the papers in reverseorder. On Iori’s paper, the discussant suggestedthat an objective function clearly specifyingthe trade off between output of the system andrisk may be useful, but felt that overall thepaper was a good first step to model thecomplex interaction between strategicbehaviour and risk in settlement systems.Regarding Nguyen’s paper, the discussantsuggested that the worst case scenario used inthe paper was a bit extreme and that the paperwould benefit from discussing moreextensively the change in the structure of thebanking system that took place in 1996/97. The


discussant also thought that the paper wouldbenefit from a better proxy for the prevalenceof the money centre structure of the bankingsystem. Finally, regarding Gropp’s paper, thediscussant questioned some of theinterpretations of the results in the paper andargued that the weekly frequency of the datamay be too high, given that many macrovariables are only available at a quarterlyfrequency. He also wondered about thepersistence of shocks in the paper.

In the general discussion, Steven Ongena(Tilburg University) argued that in the Gropppaper given the left censoring of the first stagedependent variable a negative binomial modelmay be more appropriate than the Poissonmodel used in the paper. He also stressed that inthe set up chosen in the paper, it was difficult tobe sure that all macro/common shocks had beenaccounted for. If they had not, the results of thepaper may be biased in favour of findingcontagion, especially cross-border contagion.Regarding the Nguyen paper, it was questionedwhether the reduction in the risk of contagionwas due to the change in market structuredescribed in the paper or due to an increase inTier 1 capital, which could be observed in mostbanks in OECD countries during the period.Further, it was argued that the presence ofinternational exposures of banks may bias theresults in favour of finding cross-bordercontagion. On Nguyen’s paper, someparticipants defended the relatively high lossgiven default assumptions used in the paper, aseven if the eventual recovery is higher, thepaper is concerned with the immediate impactand not the long term consequences of a bankfailure.

147ECB


ANNEX GA NN EX G PA P ER S O F T H E L AM FA LU S S Y F E L LOWS

International Equity Flows and Returns:A Quantitative Equilibrium Approach

Rui Albuquerque Gregory H. Bauer Martin Schneider

November 2003

Abstract

This paper considers the role of foreign investors in developed-country equitymarkets. It presents a quantitative model of trading that is built around two new as-sumptions: (i) both the foreign and domestic investor populations contain investorsof di erent sophistication, and (ii) investor sophistication matters for performancein both public equity and private investment opportunities. The model delivers aunified explanation for three stylized facts about US investors’ international equitytrades: (i) trading by US investors occurs in bursts of simultaneous buying andselling, (ii) Americans build and unwind foreign equity positions gradually and(iii) US investors increase their market share in a country when stock prices therehave recently been rising. The results suggest that heterogeneity within the foreigninvestor population is much more important than heterogeneity of investors acrosscountries.

JEL Classification: F30, G12, G14, G15.Keywords: Asymmetric information, heterogenous investors, asset pricing, in-

ternational equity flows, international equity returns.

We thank Monika Piazzesi for detailed comments and seminar participants at the WEGMANSconference in Rochester, Simon School of Business, the Federal Reserve Bank in Cleveland, the 2003SAET conference, the 2003 SED meeting and the 2003 ES Winter Meetings. Albuquerque thanks theLamfalussy Fellowship program sponsored by the European Central Bank. Any views expressed areonly those of the authors and do not necessarily represent the views of the European Central Bank orthe Eurosystem. Albuquerque and Bauer are a liated with the William E. Simon Graduate School ofBusiness Administration. Address: Carol Simon Hall, University of Rochester, Rochester, NY 14627.Schneider is at the Department of Economics at NYU, 269 Mercer St., 7th floor, New York, NY 10003.


� � � � � � � � � � � � � � � � � � � � � � � � � " $ � & ' � � ) � � � � � " ' � � � � / " � " & � ' � � � " �

� " � � � " $ " � � � ' & 8 � � " & � � ' " � ;

< > @ A > D F G H >

J K M N P Q S K U Q V X Z N Q [ K S N Q ] _ ` a b ] U c ` d V f f K c Ki Q P N U k a m V U k V U p d q r q m i a t u b u

w x S N ] f z c ] { f ] N u ] V P ] | } _ f u N _ u { }

� � � � � � � � H � � � � � � �

� � � � � � � �

� � � � � � � � � � � � � � � � � � � � � � ¡ � � � � � ¡ ¡ £ � � � ¤ ¡ ¥ � � � � � ¤ ¦ ¤ � ¡ ¡ � ¤ ¥ � � � � � ¡ � � ¡ � � � © ª ¡ « � � ¡¬ � � � � � � ¤ £ � � ¦ � ¡ ¯ � � � ° � ¤ ¦ ¦ � � � ± � � � � ¡ � � � £ £ � � ¦ � ¡ ¯ � � � � � � � ¤ ¡ � ¬ � � � � � � � � � � � ¡ � � ³¡ � � ¤ ¦ � � ± � � � � � � � � ¤ � � � � � � � ¥ � � ¦ � ¤ ¥ ¡ � � � � � ¦ � ¡ � � � ± � � � � ¤ � � ± � � £ � ´ £ � ¡ � � � � ¡ � � � � � ¤ ¡ � ©ª ¡ ¡ � � ¤ � ¡ ¦ � � � � � ¡ ¡ £ � � � ¤ ¡ � � � ° � � � � � ¤ ¥ ¬ � � � � µ � ¥ � ¤ � � � � � � � ¡ � � ¤ � £ ¦ � £ � ¯ � � ¤ ¦ � � � ³� � � � � � � ¡ ¡ £ � � � ¤ ¡ ¬ � � £ � � � ´ � ¡ · � � ¤ ¡ � � ¡ · � � � � � � ¡ � � ¡ � � � � � ¬ ¤ � ¡ � � ¤ � � ¬ ¡ � � £ � ¤ ¥ ¡ � � ¬¡ � � � � ¡ ¡ £ � � � ¤ ¡ � ¤ ¡ � � ± � £ © ¸ � � £ � � � ¡ ¡ £ � � � ¤ ¡ ¬ � � £ � � � � � � ¤ � ¤ ³ � � ¤ � ¡ � ¤ � � � £ £ ¯ � � £ � ¡ � ¦ ¡ �¡ � � £ � ¤ ¥ ¡ � � ¬ � � ¡ ¡ £ � � � ¤ ¡ � ¯ � £ � � ¤ ¦ � � ´ � ¡ � � � � � � ¡ � � ¡ � � � » ¡ � � � � � � ¤ � � £ � � � � ¡ � � ¤ ° � ¤ ¥� ¬ · � � � � � � � � � ¡ � � ¡ � � ¦ � £ � ± � � � ¥ � � � ¡ � � � ¡ � ´ � £ � ¡ ¯ ©

í � � � � � � � � � � � � ´ � � ¤ � � � � � � � ¦ ´ ¯ ¡ � � � ¡ � � � ¤ ¦ � � ¡ � � î � � ¬ � £ � � ¯ ï � £ £ � · � � � � ð � � ¥ � � � � � � ¤ � � � � ¦ ´ ¯ ¡ � �ò � � � � � ¤ ó � ¤ ¡ � � £ ô � ¤ ° © õ ¤ ¯ ± � � · � � µ � � � � � � ¦ � � � � ¤ £ ¯ ¡ � � � � � ¬ ¡ � � � ¡ � � � � ¤ ¦ ¦ � ¤ � ¡ ¤ � � � � � � � � £ ¯ � � � � � � � ¤ ¡

¡ � � ± � � · � � ¬ ¡ � � ò ó ô � � ¡ � � ò � � � ¯ � ¡ � � ©

149ECB


ANNEX G

Financial Market Integration and LoanCompetition: When is Entry Deregulation Socially

Beneficial?∗

Leo Kaas†

January 28, 2004

Abstract

The paper analyzes how the removal of barriers to entry in banking affectloan competition, bank stability and economic welfare. We consider a model ofspatial loan competition where a market that is served by less efficient banksis opened to entry by banks that are more efficient in screening borrowers.It is shown that there is typically too little entry and that market shares ofentrant banks are too small relative to their socially optimal level. This isbecause efficient banks internalize only the private but not the public benefitsof their better credit assessments. Only when bank failure is very likely orvery costly, socially harmful entry can occur.

JEL classification: D43; D82; G21Keywords: Entry deregulation; Bank competition

∗This paper has been prepared under the Lamfalussy Fellowship Program sponsored by theEuropean Central Bank. Any views expressed are only those of the author and do not necessar-ily represent the views of the ECB or the Eurosystem. I wish to thank an anonymous referee,Hans Degryse, Philipp Hartmann, Thorsten Koeppl, Cyril Monnet, Bruno Parigi, participants ofthe Joint Lunchtime Seminar of the ECB, the CFS and the Bundesbank, and at the ECB–CFSworkshop on “Capital markets and financial integration in Europe” hosted by the Bank of Greece(November 20–21, 2003) for many helpful comments.

†Department of Economics, University of Vienna, Hohenstaufengasse 9, 1010 Vienna, Austria.E-mail: [email protected]


Euro-Area Sovereign Yield Dynamics:

The Role of Order Imbalance

Albert J. Menkvelda, Yiu C. Cheungb, Frank de Jongb

February 3, 2004

AbstractWe study sovereign yield dynamics for ten-year government bonds in thelargest euro-area markets: Italy, France, Belgium, and Germany. We exploitas of yet unused transaction data to explain daily yield changes. We findthat these changes are caused by (i) a “benchmark” yield innovation, (ii)a common yield spread innovation, (iii) country-specific innovations, and(iv) (microstructure) noise. We relate changes in each of these factors toorder imbalance and find that Italian order imbalance explains yield spreadinnovations, French and Belgian order imbalance explain country-specificinnovations, and German order imbalance only changes yields temporarily.Order imbalance, however, does not have explanatory power for the mostimportant factor: benchmark yield innovations.

aVrije Universiteit AmsterdambUniversiteit van Amsterdam

Corresponding author is Albert J. Menkveld, [email protected], Vrije UniversiteitAmsterdam, FEWEB, De Boelelaan 1105, 1081 HV Amsterdam, Netherlands, phone:+31 20 4446130, fax: +31 20 4446020, web: www.albertjmenkveld.org. Yiu C. Cheungcan be reached at [email protected]; Frank de Jong at [email protected]. Forthis project, Menkveld received a 2003 Lamfalussy fellowship from the European CentralBank, for which he is very thankful. And, we are grateful to MTS Spa and Barbara Rindifor providing the data, to an anonymous referee at the European Central bank for usefulcomments, and to Patricia van Dam for research assistance. The usual disclaimer applies.

151ECB


ANNEX G

Illiquidity Spillovers: Theory and Evidence

From European Telecom Bond Issuance

Yigal S. Newman

Michael A. Rierson

January 22, 2004

ABSTRACT

In a study of the European telecommunication-sector bond market, we find empirical

evidence that a firm’s new bond issue can temporarily inflate yield spreads of other bonds in

its sector. We show that this effect seems unrelated to new fundamental information about

the bond’s issuer. Our results imply that an issuance of 15.5 billion Euros by Deutsche

Telekom temporarily depressed the mark-to-market value of 100 billion Euros in outstanding

European telecom debt by approximately 273 million Euros. This study is supported and

motivated by a stylized model of a risk-averse liquidity-provider in which supply shocks,

such as new issues, place price pressure on correlated securities.

†Newman (corresponding author) is at the Graduate School of Business, Stanford University, Stanford,CA 94305-5015. Rierson is at CitiGroup’s Global Equities. Email addresses: [email protected] [email protected]. The current version of this paper is available athttp://www.stanford.edu/∼ynewman. A previous version of this paper was circulated under the title“The Volume of New Issuance and Its Impact on Market-Wide Credit Spreads.” We are deeply indebted to DarrellDuffie for providing guidance, feedback, and countless useful suggestions. We also thank Peter DeMarzo, KenSingleton, Ming Huang, Anat Admati, Mark Ferguson, Jeremy Graveline, Philipp Hartmann, Ilan Kremer, PaulPfleiderer, Jeff Zwiebel, seminar participants at Stanford University, the Hebrew University, Tel-Aviv University, theInter-Disciplinary Center Herzliya, the European Central Bank, the 3rd ECB-CFS workshop on “Capital Marketsand Financial Integration in Europe,” the Salomon Smith Barney Fixed-Income Quantitative Research Group, and,especially, Yakov Amihud, for useful comments and suggestions. All errors and omissions are our own. This paperhas been prepared under the Lamfalussy Fellowship program sponsored by the European Central Bank. Any viewsexpressed are only those of the authors, and do not necessarily represent the views of the ECB or Eurosystem.

152ECBCapital markets and financial integration in europeDecember 2004

OXFORD REVIEW OF ECONOMIC POLICYVOL. 20 NO. 4, WINTER 2004

EUROPEAN FINANCIAL INTEGRATION(EDITED BY: XAVIER FREIXAS,PHILIPP HARTMANN, AND COLIN MAYER)

CONTENTS

1 The Assessment: European FinancialIntegration. Xavier Freixas, Pompeu FabraUniversity, Philipp Hartmann, EuropeanCentral Bank, and Colin Mayer, OxfordUniversity

2 Financial systems in Europe, the US, andAsia. Franklin Allen, University ofPennsylvania, Michael Chui, Hong KongMonetary Authority, and Angela Maddaloni,European Central Bank

3 Measuring European Financial Integration.Lieven Baele, Ghent University and TilburgUniversity, Annalisa Ferrando, PeterHördahl, Elizaveta Krylova and CyrilMonnet, all European Central Bank

4 The Impact of Technology and Regulationon the Geographical Scope of Banking. HansDegryse, University of Leuven, and StevenOngena, Tilburg University

5 The European Bond Markets Under EMU.Marco Pagano, University of Naples, andErnst-Ludwig von Thadden, University ofMannheim

6 Equity Returns and Integration: Is EuropeChanging? Kpate Adjaoute, HSBC PrivateBank (Suisse) SA, and Jean-Pierre Danthine,University of Lausanne

7 Bank Regulation and MacroeconomicFluctuations. Charles Goodhart, LondonSchool of Economics, Boris Hofmann,Deutsche Bundebank, and MiguelSegoviano, London School of Economics

ANNEX H OUTLINE OF THE SPECIAL ISSUE OF THE“OXFORD REVIEW OF ECONOMIC POLICY”ON “EUROPEAN FINANCIAL INTEGRATION”

153ECB

Capital markets and financial integration in europeDecember 2004

ANNEX I

ANNEX I LIST OF CONTRIBUTORS TO THE ECB-CFSRESEARCH NETWORK (2002-2004)

Renée AdamsStockholm School of Economics

Rui AlburquerqueUniversity of Rochester

Franklin AllenUniversity of Pennsylvania

Ignazio AngeloniEuropean Central Bank

Stefan ArpingUniversity of Amsterdam

Lieven BaeleGhent University

Marco Da RinUniversity of Turin, ECGI, and IGIER

Olivier de BandtBanque de France

Jesper BergEuropean Central Bank

Wolfgang BesslerGiessen University

Bruno BiaisUniversity of Toulouse

Joseph BisignanoBank for International Settlements

Michael BinderUniversity of Frankfurt and CFS

Roberto BlancoBanco de Espana

Niall BohanEuropean Commission

Emilia Bonaccorsi di PattiBanca d’Italia

Robin J. BrooksInternational Monetary Fund

Claudia BuchKiel Institute of World Economics

Giorgio CalcagniniUniversity of Urbino

Mark CareyFederal Reserve Board

Elena CarlettiUniversity of Mannheim andCenter for Financial Studies

Yiu Chung CheungUniversity of Amsterdam

Stijn ClaessensUniversity of Amsterdam

Giancarlo CorsettiUniversity of Rome III

Robert ConnollyUniversity of North Carolina – Chapel Hill

Douglas CummingUniversity of Alberta

Jean-Pierre DanthineUniversity of Lausanne

Hans DegryseUniversity of Leuven and CentER

Giovanni Dell’AricciaInternational Monetary Fund

Mihir DesaiHarvard University

Mario DraghiGoldman Sachs

Francesco DrudiEuropean Central Bank

Paul EhlingUniversity of Lausanne

Michael EhrmannEuropean Central Bank

Marina EmirisNational Bank of Belgium

Annalisa FerrandoEuropean Central Bank

Miguel Angelo FerreiraISCTE School of Business

Michael J. FlemingFederal Reserve Bank of New York

Robert FloodInternational Monetary Fund


Xavier FreixasUniversitat Pompeu Fabra

Kenneth D. GarbadeFederal Reserve Bank of New York

Vítor GasparEuropean Central Bank

Thomas GehrigUniversity of Freiburg

Mariassunta GiannettiStockholm School of Economics

Heather GibsonBank of Greece

Alberto GiovanniniUnifortune Asset Management

Peter GomberDeutsche Boerse AG

Peik GranlundBank of Finland

Luigi GuisoUniversity of Sassari

Reint GroppEuropean Central Bank

Leo de HaanDutch Central Bank

Michael HaliassosUniversity of Cyprus

Michael HallingUniversity of Vienna

Hannu HalttunenNordea

Sirkka HämäläinenEuropean Central Bank

Philipp HartmannEuropean Central Bank

Thomas HarrUniversity of Copenhagen

Harald HauINSEAD

Cornelia HolthausenEuropean Central Bank

Peter HördhalEuropean Central Bank

Harry HuizingaTilburg University

Giulia IoriKings College, London

Otmar IssingEuropean Central Bank

Jos JansenDe Nederlandsche Bank

Tullio JappelliUniversity of Salerno

Leo KaasUniversity of Vienna

Charles KahnUniversity of Illinois

Karlo KaukoBank of Finland

Enisse KharroubiDELTA and Banque de France

Stefanie KleimeierUniversity of Maastricht

Daniela KlingebielWorld Bank

Thorsten KoepplEuropean Central Bank

Yrjö KoskinenStockholm School of Economics

Jan Pieter KrahnenCenter for Financial Studies

Randy KrosznerUniversity of Chicago

Alexandre LamfalussyUniversité Catholique de Louvain

Philip LaneTrinity College Dublin

Gyongyi LoranthLondon Business School

Simone ManganelliEuropean Central Bank

155ECB


ANNEX I

Arnaud MarèsEuropean Central Bank

Colin MayerOxford University

David MayesBank of Finland

Michael MelvinArizona State University

Albert MenkveldVrije Universiteit Amsterdam

Joël MérèreEuroclear France

Cyril MonnetEuropean Central Bank

Yigal S. NewmanStanford School of Business

Grégory NguyenNational Bank of Belgium

Giovanna NicodanoUniversity of Turin

Steven OngenaTilburg University

Charlotte OstergaardNorwegian School of Management

Tommaso Padoa-SchioppaEuropean Central Bank

Marco PaganoUniversity of Salerno

Bruno ParigiUniversity of Padova

Paolo PasquarielloUniversity of Michigan

Maria Fabiana PenasFree University of Amsterdam

Vicente PonsYale School of Management

John J. PuthenpurackalOhio State University

Eli RemolonaBank for International Settlements

Helene ReyPrinceton University

Geert RouwenhorstYale School of Management

João A. C. SantosFederal Reserve Bank of New York

Antonio Sáinz de VicuñaEuropean Central Bank

Sinikka SaloBank of Finland

Alexander SchaubEuropean Commission

Garry SchinasiInternational Monetary Fund

Heiko SchmiedelEuropean Central Bank

Armin SchwienbacherUniversity of Amsterdam

Giancarlo SpagnoloSveriges Riksbank and Universityof Mannheim

Benn SteilCouncil on Foreign Relations

Rune StenbackaSwedish School of Economics

Avanidhar SubrahmanyamUniversity of California at Los Angeles

Salomon TadesseUniversity of South Carolina

Tuomas TakaloBank of Finland

Jens TapkingEuropean Central Bank

Juha TarkkaBank of Finland

Eric TheissenUniversity of Bonn

P. ThomopoulosBank of Greece


Jens ThomsenDanmark Nationalbank

Kostas TsatsaronisBank for International Settlements

Nikolaos TsaveasBank of Greece

Gertrude Tumpel-GugerellEuropean Central Bank

Tereza TykvovaCenter for European Economic Research

Christian UpperDeutsche Bundesbank

Clara VegaUniversity of Rochester

Jukka VesalaEuropean Central Bank

Xavier VivesINSEAD

Jonas VlachosUniversity of Chicago

Axel WeberUniversity of Cologne

Andre WentEuronext

David WrightEuropean Commission

Holger WolfGeorgetown University

Xiaoyun YuIndiana University

157ECB


ANNEX J

The Research Network is organised by a Steering Committee. This is composed of representativesof the two organising institutions (ECB and CFS) and of outside scientific experts from academia.

ANNEX J COMPOSITION OF THE STEERINGCOMMITTEE (2002-2004)

For 2004- :Roberto Mario Billi

Center for Financial Studies

Lorenzo CappielloEuropean Central Bank

Franklin AllenWharton School, University ofPennsylvania

Giancarlo CorsettiUniversity of Rome III and Yale University

Jean-Pierre DanthineUniversity of Lausanne

Vítor GasparEuropean Central Bank

Philipp HartmannEuropean Central Bank

Jan Pieter KrahnenCenter for Financial Studies and Universityof Frankfurt

Marco PaganoUniversity of Salerno

Axel WeberCenter for Financial Studies and Universityof Frankfurt

Secretariat:

For 2002:Bernd Kaltenhauser


Simone ManganelliEuropean Central Bank

For 2003-2004:Christina Metz


Cyril MonnetEuropean Central Bank


Chair Hartmann, Philipp DG-Research (Financial Research Division)

Secretary Capiello, Lorenzo (2004- ) DG-Research (Financial Research Division)

Manganelli, Simone (2002) DG-Research (Financial Research Division)

Monnet, Cyril (2003-2004) DG-Research (Financial Research Division)

Members Decker, Doris (2002-2003 ) DG-Market Operations (Operations AnalysisDivision)

Egea, Carlos (2003- ) DG-International and European Relations(European Relations Division)

Grittini, Sergio (2003- ) DG-Market Operations(Operations Analysis Division)

Huemer, Stefan (2002-2003 ) DG-International and European Relations(Institutions and Fora Division)

Ferrando, Annalisa (2003- ) DG-Economics(Capital Market and Financial Structure Division)

Kerjean, Stéphane (2004- ) DG-Legal Services(Financial Law Division)

Leclercq, Guillaume (2002) DG-Secretariat and Language Services(Secretariat Division)

Maddaloni, Angela DG-Economics

(2002-2003) (Capital Market and Financial Structure Division)

Marès, Arnaud (2002-2003) DG-Market Operations(Operations Analysis Division)

Mayerlen, Frank (2004- ) DG-Statistics(Statistics Development and Coordination Division)

Ohlerich, Christoph (2004- ) DG-Operations(Risk Management Division)

Pronk, Carin (2002-2004) DG-Statistics(Statistics Development and CoordinationDivision)

Scheicher, Martin (2004- ) D-Financial Stability and Supervision(Financial Supervision Division)

ANNEX K COMPOSITION OF THE ECB INTERNALCONTACT GROUP

159ECB


ANNEX K

Stenström, Mikael (2002- ) DG-Legal Services (Financial Law Division)

Tabakis, Evangelos (2002- ) DG-Market Operations(Risk Management Division)

Tapking, Jens (2002- ) DG-Payment Systems and Market Infrastructure(Securities Settlement Systems Policy Division)

Vesala, Jukka (2002-2004) D-Financial Stability and Supervision(Financial Stability Division)


National central banks have nominated a contact person for network related matters. Thecomposition of the NCB contact group is:

COMPO S I T I ON O F T H E N CB CONTA C T G ROUP

Oesterreichische Nationalbank Martin Scheicher

Nationale Bank van België / Marina EmirisBanque Nationale de Belgique

Suomen Pankki – Finlands Bank Tuomas Takalo

Banque de France Henri Pagès

Deutsche Bundesbank Hannah Hempell

Bank of Greece Heather Gibson

Central Bank of Ireland John Frain

Banca d’Italia Carmello Salleo

Banque Centrale du Luxembourg Patrick Lunnemann

De Nederlandsche Bank Leo de Haan

Banco de Portugal Goncalo Ribeiro

Banco de España Juan Ayuso

RESEARCH NETWORK ON CAP ITAL MARKETSAND F INANC IAL INTEGRAT ION IN EUROPE

DECEMBER 2004

CENTER FOR

FINANCIALSTUDIES

RESULTS ANDEXPERIENCE AFTER TWO YEARS

Documents

Research network on capital markets and financial integration in