ECB policy during the crisis: “food for thoughts” for monetary policy research

ECB policy during the crisis: “food for thoughts” for monetary policy research

Lucrezia ReichlinLondon Business School

Based on joint work with D. Giannone (ULB), M. Lenza (ECB) and H. Pill (ECB)

The CEPR/ESI 14th Annual Conference on ‘How Has Our View of Central Banking Changed with the Recent Financial Crisis?’

Central Bank of Turkey, 28-29 October 2010

Outline

1. The crisis: symptoms and policy response2. Has and how ECB policy worked?3. Some key Issues4. Is there any lesson to learn?

Outline

1. The crisis: symptoms and policy response2. Has and how ECB policy worked?3. Some key issues4. Is there any lesson to learn?

0

1

2

3

4

5

6

2/7/2007 5/11/2007 14/03/08 23/07/08 26/11/08 6/4/2009 13/08/09 17/12/09 28/04/10

-50

0

50

100

150

200

250

Euro money market rates and spreads lhs: percent per annum; rhs: basis points

3-month repo ratesecured, lhs

3-month EURIBORunsecured, lhs

spreadrhs

turmoil

Bear Stearns

Lehman

Source: Bloomberg

Symptoms Euro money market rates and spreads lhs: percent per annum; rhs: basis points

Preliminaries

• Failure of Lehman increased perceived counterparty risk.• Adverse selection led to a freezing of the interbank money

market : heterogeneity, ‘red-lining’ of some banks in interbank market

• Banks are unable to refinance positions and maintain flow of loans to the private sector.

• Governments take various actions:– Fiscal stimulus– Support for financial sector (re-capitalisation, guarantees

for bank bonds, etc.)

Diagnosis

Monetary policy response

Standard monetary policy response : lower interest rates

Non standard measures:• Aim at restoring market functioning …• In the money market, replace interbank transactions with

transactions across the central bank balance sheet (i.e. act as an ‘intermediary-of-last-resort’);

• Improve the availability of bank funding, facilitating securitization and improve functioning of covered bond market …

monetary policy and financial stability policies go hand in hand

Monetary Policy Response: non-standard measures

• Various measures which expanded central bank intermediation …

– Liquidity transformation – accept broad range of collateral in fixed rate / full allotment operations;

– Maturity transformation – lengthen maturity of operations out to 1 year (absorbing at the (overnight) deposit facility);

– Facilitate payments – conduct operations with a large set of counterparties;

– Manage information issues – Eurosystem operations are anonymous, no stigma attached.

Key elements: fixed rate tenders with full allotment (FRFA) in monetary

policy operations

Non standard policies led to expansion of ECB balance sheet after Lehman

-- Asset side: expand scope of repo operations -- Liability side: allow increasing recourse to

deposit facilitiesWhile the ECB has not adopted the rhetoric of “quantitative

easing”, it has expanded it balance sheet, increasing reserves on the liabilities side of its balance sheet against (largely) conventional assets (repos) on the asset side

-2000

-1600

-1200

-800

-400

0

400

800

1200

1600

2000

2007W01 2007W21 2007W41 2008W09 2008W29 2008W49 2009W17 2009W37 2010W04 2010W24

-2000

-1600

-1200

-800

-400

0

400

800

1200

1600

2000

Eurosystem balance sheet EUR billions

9 August Lehman

Source: ECB

Comparison of central bank balance sheet sizepercent of GDP, June of each year

0

2

4

6

8

10

12

14

16

18

20

ECB FED BoE

2007 200720072008 2008 200820092009 2009

50

55

60

65

70

75

2006Jan 2006Jul 2007Jan 2007Jul 2008Jan 2008Jul 2009Jan 2009Jul 2010Jan 2010Jul

50

55

60

65

70

75

MFIs (incl. Eurosystem)to MFIs

MFIs (excl. Eurosystem)to MFIs

Source: ECB

Intra-MFI (bank) sector creditas a percentage of credit to the non-financial sector

0

1

2

3

4

5

6

2/7/2007 5/11/2007 14/03/08 23/07/08 26/11/08 6/4/2009 13/08/09 17/12/09 28/04/10

-50

0

50

100

150

200

250

The effect was a decrease in spreads: secured-unsecured rates lhs: percent per annum; rhs: basis points

3-month repo ratesecured, lhs

3-month EURIBORunsecured, lhs

spreadrhs

turmoil

Bear Stearns

Lehman

Source: Bloomberg

Excess Liquidity

• As a result of excess liquidity: the overnight money market interest rate (EONIA) moved systematically away from the policy rate (Main Refinancing Operation rate) and fell towards the rate on the deposit facility.

(ECB chose to reabsorb this excess liquidity by having banks with excess cash make recourse to the (remunerated) deposit facility)

• Money market rates at all maturities adjusted downward

Page 13

Impact on overnight interest rate: wedge between policy rate and eonia percent per annum

With FRFA, excess liquidity conditions emerged in the overnight money market, and the EONIA dropped systematically to the deposit facility rate

Source: ECB

Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-100

1

2

3

4

5

6

0

1

2

3

4

5

6

EONIA

MRO rate

rate on deposit facility

marginal lending rate

Emergence of money market

tensions

Failure of Lehman Bros.

Outline


Has it worked? Exercise 1

Lenza, Pill and Reichlin, Economic Policy 2010

• Characterize the effect of the introduction of non-standard measures in terms of its impact on a variety of money market spreads:– Narrowing of the spread between secured and unsecured term rates;– Reduction of market overnight rate relative to the “policy” MRO rate;– Flattening of the money market yield curve through 1-year LTRO.

• Characterize the (partial) macroeconomic impact of non-standard measures as the difference between two counter-factual exercises (conditional forecasts) constructed using a model of the euro area economy, based on different interest rate assumptions

The model

• Developed and evaluated by Giannone, Lenza and Reichlin (2009): establishes and documents ‘stylised facts’ about monetary dynamics in the euro area.

• The model is a Bayesian vector autoregressive model (B-VAR), estimated over the sample period January 1991 to August 2008 using monthly data.

• The model consists of 32 macroeconomic variables:• Macro variables – economic activity (IP); prices (HICP); unemployment;

etc.• Monetary and credit variables – monetary aggregates; sectoral credit

by use / maturity; and• Money market rates and bond yields …

• We avoid the “curse of dimensionality” by using Bayesian shrinkage techniques (see De Mol, Giannone and Reichlin, 2008, Banbura, Giannone and Reichlin, 2010, Giannone, Lenza and Primiceri, 2010 for setting of priors)

Policy scenario (P) with non-standard measures– Euribor 3 and 12 month rates as observed between November 2008

and August 2009– Simulation → EA(L)(Xt…T│ X0…t-1; P)

No Policy scenario (NP) without non-standard measures– Euribor 3m = MRO + [Spread Euribor 3m/MRO 10/2008 level +

[Spread MRO/EONIA from 11/08 to 08/09]– Euribor 12m = MRO + [Spread Euribor 12m/MRO(10/08)] + Flattening

of the yield curve due to non-standard policy– Simulation → EA(L)(Xt…T│ X0…t-1; NP)

Effect of non-standard measuresImpactns = EA(L)(Xt…T│ X0…t-1; P) - EA(L)(Xt…T│ X0…t-1; NP)

Of course, this all assumes model is stable (we come back to that) …

Exercise 1

The macro effects of non-standard measures:

scenariosEuribor Three months

0

1

2

3

4

5

6

7

Plus EU Spread Actual MRO NP Scenario

Policy No Policy

The macro effects of non-standard measures:

scenariosEuribor Twelve months

0

1

2

3

4

5

6

7

Plus EU Spread Actual MRO NP Scenario

PolicyNo Policy

Exercise 1 - Resultsimpact of non-standard measures (EA(L)(Xt…T│ X0…t-1; P) - EA(L)(Xt…T│ X0…t-1; NP), percentage points on annual growth rates (excl. unemployment)

Source: Lenza et al, 2010

-0.8

-0.6

-0.4

-0.2

0.0

0.2

Jan-07 Jan-08 Jan-09 Jan-10 Jan-11 Jan-12

-0.8

-0.6

-0.4

-0.2

0.0

0.2

-2

-1

0

1

2

3

4


-1.5

-0.5

0.5

1.5

2.5

3.5

Unemployment rate M1

Exercise 1 - Resultsimpact of non-standard measures (EA(L)(Xt…T│ X0…t-1; P) - EA(L)(Xt…T│ X0…t-1; NP), percentage points on annual growth rates

Source: Lenza et al, 2010

Loans to non-financial corporationsLoans for house purchase

-5

-4

-3

-2

-1

0

1

2

3

4


-5

-4

-3

-2

-1

0

1

2

3

4

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6


0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Comments on exercise 1

• Not surprisingly effects on real economy consistent with what we know about the effect of monetary policy shocks (Lenza, Giannone and Reichlin, 2009)

• Strong effect on M1 suggests that M1 multiplier did not collapse (see next slides): overnight market preserved with ECB replacing the market (intermediary of last resort)

StocksUS $ billion

0

1

2

3

4

5

6

7

8

1929 1931 1933 1935 1937 1939

0

1

2

3

4

5

6

7

8

0

5

10

15

20

25

30

35

40

1929 1931 1933 1935 1937 1939

0

2

4

6

8

10

12

14

16

M1(lhs)

currency(rhs)

Source: Rasche (1987)

Money multiplierdefined as M1 / currency

M1 multiplier - Now versus ThenThe thirties

StocksEUR billion

Source: ECB BSI data

Money multiplierdefined as Mx / monetary base

0

1000

2000

3000

4000

5000

6000

7000

8000

9000

10000

2000 2002 2004 2006 2008 2010

0

400

800

1200

1600

2000

2400

2800

3200

3600

4000

monetary base(rhs)

M3(lhs)

M1(lhs)

0

2

4

6

8

10

12

14

2000 2002 2004 2006 2008 2010

0

2

4

6

8

10

12

14

euro cashchangeover

M1

M3

M1 multiplier - Now versus Then2008-2010

Exercise 2 Same model as exercise 1Giannone , Lenza and Reichlin, 2009• Compare actual path of macroeconomic variables with those of

model forecasts conditional on the observed path of economic activity (as captured in the evolution of the IP series);

• Addresses question: Have the non-standard measures prevented a “breakdown” / disruption to the pre-crisis regularities seen in the data (and, by implication, the behaviour of the economy)?

• Conditional forecasts start in Jan. 1999 (but the model is estimated using sample to August 2008 – now updated to 2010) …

Exercise 2 – Loans to NFC stable - “business as usual”annual growth rates, sa; 87% confidence interval

Source: Lenza et al, 2010; Giannone et al, 2010

Jan-

99

May

-99

Sep-

99

Jan-

00

May

-00

Sep-

00

Jan-

01

May

-01

Sep-

01

Jan-

02

May

-02

Sep-

02

Jan-

03

May

-03

Sep-

03

Jan-

04

May

-04

Sep-

04

Jan-

05

May

-05

Sep-

05

Jan-

06

May

-06

Sep-

06

Jan-

07

May

-07

Sep-

07

Jan-

08

May

-08

Sep-

08

Jan-

09

May

-09

Sep-

09

Jan-

10

May

-10

-25

-20

-15

-10

-5

0

5

10

15

20

Observed annual growth rates

Exercise 2 – M1 stable - “business as usual”annual growth rates, sa; 87% confidence interval


Jan-

99

May

-99

Sep-

99

Jan-

00

May

-00

Sep-

00

Jan-

01

May

-01

Sep-

01

Jan-

02

May

-02

Sep-

02

Jan-

03

May

-03

Sep-

03

Jan-

04

May

-04

Sep-

04

Jan-

05

May

-05

Sep-

05

Jan-

06

May

-06

Sep-

06

Jan-

07

May

-07

Sep-

07

Jan-

08

May

-08

Sep-

08

Jan-

09

May

-09

Sep-

09

Jan-

10

May

-10

-10

-5

0

5

10

15

20

25

30

Observed annual growth rates

Exercise 2 – 3-month-euribor stable - “business as usual”

Jan-

04

Apr-0

4

Jul-0

4

Oct-04

Jan-

05

Apr-0

5

Jul-0

5

Oct-05

Jan-

06

Apr-0

6

Jul-0

6

Oct-06

Jan-

07

Apr-0

7

Jul-0

7

Oct-07

Jan-

08

Apr-0

8

Jul-0

8

Oct-08

Jan-

09

Apr-0

9

Jul-0

9

Oct-09

Jan-

10

Apr-1

0

Jul-1

0

-6

-4

-2

0

2

4

6

8

10

for house purchase for consumption


Exercise 2 – some heterogeneity in loansannual growth rates, sa; 68% confidence interval

-5

0

5

10

15

Jan-96 Jan-98 Jan-00 Jan-02 Jan-04 Jan-06 Jan-08 Jan-10

-5

0

5

10

15

Simulated

Actual

-5

-3

-1

1

3

5

7

9

11

13

15


-5

0

5

10

15

Simulated

Actual

Comments on exercise 2

1. Loans to corporate, 3-month euribor and M1 remarkably stable [same message as exercise 1 – this is ``indication’’ that ECB policy worked to some extent]

2. The euribor path is evidence of stability of ECB policy rule

3. Some heterogeneity on loans behaviour

Remarks on the exercise

It is a counterfactual exercise: no clean interpretation in terms of causality

But if you are willing to assume that no large financial shock materialized from 1992 to 2007, this is sort of a natural experiment: we are conditioning on past correlations (no financial crisis) and business conditions observed during the crisis

Page 32

What does this say?

• ECB has achieved some results without massive QE and no recourse to zero bound

• Has done so by manipulating spreads and using more than an instrument in a situation of pervasive frictions

• The stated ECB objective has been the preservation of the transmission mechanism – non standard monetary policies have acted as complements to lowering interest rates

• Emphasis on the fact that EA is bank based financial system

Bank versus market centered financial systems percent of total debt financing of corporate sector

0

10

20

30

40

50

60

70

80

90

100

loans debt sec loans debt sec

euro area United States

0

10

20

30

40

50

60

70

80

90

1001999 2009

Outline


A. The corridor and liquidity policy: has the ECB done the right thing ?

The ECB has used two instruments, policy rate and deposit rate

Why introducing a costly wedge between the money market rate and the “policy rate”?

• this reflects the existence of excess liquidity in aggregate (which drives EONIA towards DF rate) and the desire of the CB not "penalise" banks making recourse to the (perfectly elastic) supply of liquidity at the MRO with FRFA

The right thing to do during normal times

• If itRR = it

P and RR >> md(normal) (the demand by banks for CB liquidity to settle net interbank positions), then arbitrage at the margin will ensure that it

D = itRR (= it

P)

• This is the case in a “well-functioning” money market

• In other words, sufficiently high remunerated RRs can satiate the banks demand for CB liquidity at zero cost to the banks (Friedman rule)

Frozen money market

• Say the money market “freezes”• (In extremis,) banks have to settle all gross

positions in CB money• By implication md(exceptional) >> md(normal) • … and it is possible that md(exceptional) > RR

Frozen money market

• In that case:– with unchanged supply of CB liquidity

itD >> it

RR (= itP )

– with increased / elastic supply of CB liquidity (to re-establish satiation) ms(exceptional) ≥ md(exceptional) > RR

itD < it

RR (= itP)

falls to zero or to rate on deposit facility in a corridor system

Problem 1

• Situation just described mimics the Bagehot rule -- lend freely against good collateral at a penalty rate.

• But is the collateral really "good" and is the penalty sufficiently large to create an incentive to correct underlying structural problems?

• That is what the problem of "dependence" from CB liquidity is about.

• Collateral quality is also a policy variable

Problem 2

• The idea of using the corridor is motivated by banks’ heterogeneity

• Evidence ......

• But if it is the value of the instrument that matters and not the value of the institution, better to use the collateral to discriminate

2007W012007W022007W032007W042007W052007W062007W072007W082007W092007W102007W112007W122007W132007W142007W152007W162007W172007W182007W192007W202007W212007W222007W232007W242007W252007W262007W272007W282007W292007W302007W312007W322007W332007W342007W352007W362007W372007W382007W392007W402007W412007W422007W432007W442007W452007W462007W472007W482007W492007W502007W512007W522008W012008W022008W032008W042008W052008W062008W072008W082008W092008W102008W112008W122008W132008W142008W152008W162008W172008W182008W192008W202008W212008W222008W232008W242008W252008W262008W272008W282008W292008W302008W312008W322008W332008W342008W352008W362008W372008W382008W392008W402008W412008W422008W432008W442008W452008W462008W472008W482008W492008W502008W512008W522009W012009W022009W032009W042009W052009W062009W072009W082009W092009W102009W112009W122009W132009W142009W152009W162009W172009W182009W192009W202009W212009W222009W232009W242009W252009W262009W272009W282009W292009W302009W312009W322009W332009W342009W352009W362009W372009W382009W392009W402009W412009W422009W432009W442009W452009W462009W472009W482009W492009W502009W512009W522009W532010W012010W022010W032010W042010W052010W062010W072010W082010W092010W102010W112010W122010W132010W142010W152010W162010W172010W182010W192010W202010W212010W222010W232010W242010W252010W262010W272010W282010W292010W302010W312010W322010W332010W342010W352010W362010W372010W382010W392010W402010W412010W420

50

100

150

200

250

300

350

400

0

50

100

150

200

250

300

350

400

Failure of Lehman Bros.15 September 2008

Volume alloted inMain refinancing operation

Recourse to deposit facility

EUR billions

The existence of excess liquidity in aggregate (as reflected in recourse to DF) co-exists with individual bank demand for liquidity at MRO, which points to (severe) segmentation given the difference in rates (width of the corridor).

0

5

10

15

20

25

30

02/01/07 02/07/07 02/01/08 02/07/08 02/01/09 02/07/09 02/01/10

0

5

10

15

20

25

30

GBP LIBOR

EURIBOR

USD LIBOR

Failure of Lehman Bros.15 September 2008

Dispersion of individual bank contributions to EURIBOR average mean absolute deviation from fixing, basis points

Problem 3: liquidity and monetary policy

1. Eonia lower than policy rate implies that effective stance is more accommodative than what revealed by MRO

2. Difficult to separate between liquidity and monetary policy objectives [recent: eg euribor/eonia are creeping up now – what is the macroeconomic effect?]

Communication challenge

B. Banks’ funding – some puzzling facts

• In the crisis collapse of saving deposit beyond what explained by business cycle conditions

• This reflects the ``unusually’’ steep yield curve(M3 and yield curve not stable unlike euribor and

M1)• This must have implications for banks’ funding


-5

0

5

10

15

20


-5

0

5

10

15

20

Simulated

Actual

Exercise 2 – Results annual growth rates, sa; 68% confidence interval2006-08 surge of M3 and its collapse in the crisis is not ``business as usual’’ (difference with M1)

Page 47

This is explained by saving deposits (M2-M1 component of M3)

(M3-M1) versus slope of yield curve: stable relation

M3 - M1annual growth rate,

percent, lhs

10yr – 2yrpercentage points,

rhs

Actual (solid line) and Counterfactual (dashed line) interest rates: 3-month and 10-year interest rate

Jan-96 Jan-97 Jan-98 Jan-99 Jan-00 Jan-01 Jan-02 Jan-03 Jan-04 Jan-05 Jan-06 Jan-07 Jan-08 Jan-09 Jan-10-1

0

1

2

3

4

5

6

7

8

9

3-month euribor

Ten-year rate

Comments

• This is different issue of transmission mechanism – long term interest rate stickiness not well understood

• This has implications for banks’ funding over the cycle

• More research needs to be done here

Outline


Wrapping up

• We have seen an experiment of monetary policy in presence of financial frictions

• Main friction is in interbank market• If frictions more than one interest rate matters• CB has more than one interest as tool – therefore can cope

with some financial stability issues but not all• When main friction in money market and financial system is

banked based QE ECB style makes sense and some evidence that it has worked

• However, beyond the short run, need to understand incentives in presence of banks’ heterogeneity and bank funding strategies in relation to term spread behaviour

Extra slides

Jan-07 Jul-07 Jan-08 Jul-08 Jan-09 Jul-09 Jan-10 Jul-10-600

-400

-200

0

200

400

600

800

1000

-600

-400

-200

0

200

400

600

800

1000

Long term

Eurosystem operations

Short term (excl. Eurosystem operations)

Main liabilities

Normalization and recent tensions

Documents

ECB policy during the crisis: “food for thoughts” for monetary policy research