Acquisitions and CEO Power:

Evidence from French Networks

Sabrina Chikha Jean-Yves Filbienb

Abstract

CEOs of the acquiring firms are more likely to cancel acquisitions when themarket does not react favorably to the announcement (Luo, 2005). Using a sam-ple of French acquisition announcements from 2000 to 2005, this paper studiesthe conditions under which CEOs are more willing to listen to investors. Fromthe power typology of Finkelstein (1992), we find that CEOs with a high exper-tise listen to the market. Considering the strong networks of the French EliteSchools Alumni and board connections, we also find that the well-connectedCEOs are more likely to complete a deal in spite of a negative market reactionon acquisition announcement.

EFM Classification: 150, 160JEL Classification: G30, G34Keywords: Acquisitions, CEO power, Corporate Governance

a Sabrina Chikh (corresponding author), PhD Student, Univ. Lille Nord de France - LSMRC,1 Place Deliot - 59000 Lille (France), e-mail: s.chikh@esc-lille.fr, b Jean-Yves Filbien, PhD Stu-dent, Louvain School of Management & FUCaM - 151, Chaussee de Binche - 7000 Mons (Belgium)and Univ. Lille Nord de France - LSMRC - ECCCS, 1 Place Deliot - 59000 Lille (France), e-mail:jean yves.filbien@fucam.ac.be. The authors thank advisors Isabelle Platten, Eric de Bodt, and PascalGrandin for their comments. In addition, this paper benefitted from remarks by Christophe Majois,Helen Bollaert, and participants of the CIGE Conference (Bordeaux, France) 2008, Asian Pacific Con-ference (Paris, France) 2008, the CIGE Conference 2009 (Florence, Italy), the AFFI Annual Meeting2009 (Brest, France), and the MFS 2009 Conference (Crete, Greece). Remaining errors are the soleresponsibility of the authors.

1 Introduction

“Takeovers reflect individual decisions.”– Roll, 1986, p. 3.

On June 20, 2000, Jean-Marie Messier, the Chief Executive Officer (CEO) ofVivendi, declared the acquisition of Seagram. However, investors reacted negativelyand the announcement has driven Vivendi shares down by 8.7%. In despite of thediscord of financial markets, Jean-Marie Messier did not to listen to the market andthe deal has been completed later. In 2002, he was fired from Vivendi.

The above event suggests the major objective of this paper. What are the conditionsunder which CEOs are more willing to listen to investors? In this way, we examine therelationship between CEO power and the attention paid to the market assessments.CEOs are elected for representing the shareholders’ interests and thus they have aright to expect that they take into account their opinion for deciding investments.Managerial scandals seemingly have multiplied in recent years (e.g., Enron, WorldCom,Accor, Saint-Gobain, Societe Generale, Fortis, etc.), prompting calls to establish agovernance system that ensures the welfare of the shareholders. New rules of controlattempt to limit personal incentives but they are not always sufficient to prevent CEOmotivations (Combs et al., 2007). This study aims to provide a better understandingof the corporate governance malfunction sources.

Market reactions to acquisition announcement can predict whether the bidding firmwill complete it (Luo, 2005). According to Luo, firms have a higher incentive to learnfrom the market when it has more information than the companies have. Furthermore,Kau et al. (2008) also find that managers are likely to listen to the market reactionsand tend to cancel investments to which the market reacts unfavorably, especially whenagency costs are involved. Dye and Sridhar (2002) argue that financial markets can bebetter informed than managers themselves. However, managers consider their privateinformation as superior to the aggregate public information set (Roll, 1986). Indeed,Jennings and Mazzeo (1991) find that manager’s decisions following an acquisitionannouncement are not consistent with their learning from financial markets.

Mergers and acquisitions (M&A) are an event over which managers have substantialimpact. Moreover, M&As are among the most important investment decisions for anycompany, and there are various reasons to justify M&A operations.1 However, empiricalstudies find returns close to zero or even negative for firms that acquired public targetsaround the M&A announcements (Jensen and Ruback, 1983; Mulherin and Boone,2000; Andrade et al., 2001; Fuller et al., 2002). If there is no value creation fromM&As, why do acquiring firms still involve in them? According Morck et al. (1990),acquirers managerial objectives may drive poor bidding firms performance.

The need for power determines the CEOs’ motivations (Rovenpor, 1993). WhenCEOs establish an external growth policy, they want to build an empire (Rhoades, 1983;Schneider and Dunbar, 1992) from which they boost their career and extract financialor fringe benefits. In addition, they are in search of increasing their compensation(Wright et al., 2002). Besides, acquisitions provide opportunities for the managing forresults. If CEOs anticipate insufficient results, they can improve them by acquiringprofitable firms (Shleifer and Vishny, 2003). Moreover, CEOs might decrease global

1See Collis and Montgomery (1998) for the effects of synergy, Husson (1989) for the benefits ofdiversification, Myers and Majluf (1984) regarding the reduction of asymmetric information relativeto financing, Mitchell and Mulherin (1996) for the management of industrial challenges, Bradley et al.(1983) about the under-valuation of target stock market prices, and Shleifer and Vishny (2003) andRhodes-Kropf and Viswanathan (2004) for the asset evaluation resolution.

1

risk by diversifying their firm (Morck et al., 1990). They might push firm investmentstoward assets that have some relationship with their abilities (Edlin and Stiglitz, 1995).By gearing their investments toward other sectors or specific partners, CEOs maystrengthen their position in the firm. The resulting entrenchment may guarantee moresecurity and more leeway to the CEOs, because dismissing such a CEO would be moreexpensive for shareholders (Shleifer and Vishny, 1989). Finally, Hubris could leadCEOs to overestimate their own competence (Roll, 1986), such that they believe theycan manage target firm better than its current CEO (Gupta et al., 1997; Hayward andHambrick, 1997). The illusion of control (Langer, 1975) may lead CEOs to overestimatethe success of M&As, underestimate exogenous events, and become convinced that theycan extract some benefits from the new entity. According to Malmendier and Tate(2005), overconfident CEOs overestimate their capacity to manage the firm, yet themarket does not approve such overconfident CEOs acquisitions, because they involvein the purchase of firms for more than their market value. The CEOs who displaysuch hubris often suffer the winner’s curse (Roll, 1986). According to Brown andSarma (2007), CEOs motivations are meaningful only if we consider CEO dominance,which reflects CEOs ability to convince stakeholders to follow the proposed directionor impose their point of view on them. Brown and Sarma highlight that M&A activityis determined primarily by the CEO’s power over the board.

The determinants of the degree to which CEOs listen to the market derive fromFinkelstein’s (1992) classification of CEO power, which distinguishes structural, own-ership, expertise, and prestige powers. We expect that well-connected CEOs, i.e., witha high prestige power, are more willing to pursue a decision even if the market did notapprove it. We consider the French context where CEOs are strongly well-networkedas well through education, directorships, and non profit boards. We distinguish twoformal types of French networks, the elite education and the directorship. Concerningthe elite education, the major high schools to get to CEO function are Ecole Nationaled’Administration (ENA) and Polytechnique (X). For the directorship, CEOs of Frenchblue chips usually sit on each other boards. For example, Michel Pebereau, the Chair-man of BNP-Paribas and the former CEO, holds simultaneously several directorships inbig French companies’ board (e.g., Axa, EADS, Lafarge, Galerie Lafayette, and Total).Considering the conditions under which CEOs are more willing to listen to markets,we focus on acquisition decisions.

Using a sample of French acquisition announcements from 2000-2005, our results areconsistent with Luo’s (2005) result that the likelihood to complete a deal is determinedby the market reaction. Secondly, we test the impact of the four types of power definedby Finkelstein (1992) on probability to complete an acquisition according to the marketreaction. Structural power is negatively associated to the probability to complete anacquisition that reveals a low degree of CEO listening to the stock market. Ownershippower increases the CEO’s attention paid to the market. The more the CEO ownershiplevel is, the more the CEO relies on the market signal. However, this result is notsignificant. Our results also suggest that expertise power leads the leader to listento the market. Finally, considering the strong networks of the French Elite SchoolsAlumni, we find that the well-connected CEOs are more likely to complete a dealin spite of a negative market reaction on acquisition announcement. It seems that aprivate information access leads the CEO to discount the market signal when makingacquisition bids.

Our results corroborates the dominance of social connections among managers(Burt, 2000) The well-connected CEOs access to private information and they are

2

less affected by performance because of the protection obtained from their network(Nguyen-Dang, 2008). They are also more likely to associate with people to whomthey are similar (McDonald et al., 2008) and take these similar acquaintances’ adviceto strengthen themselves in their choice. Thus, managers increase their self-confidenceby confirming their beliefs according to their friends.

Through our knowledge, few studies consider CEO power through acquisition ac-tivity. In recent research, Oler et al. (2007) and Bauguess and Stegemoller (2008)underline the CEO’s impact on the firm’s acquisition policy. These results do not al-low us to conclude on the real CEO impact. Indeed, the activity acquisition reflectsmore the firm’s expectations. Precisely, CEOs can be selected for the high capacityto lead M&As. We propose an original empirical study concerning the characteristicsof the French context, which provides a setting marked by the predominant role ofCEOs and strong networks, both of which pertain to prestige power. From a practicalperspective, our investigation also provides a better understanding of the role playedby CEOs in acquisition decisions and offers some answers about information gatheringbehavior.

The remainder of this paper is organized as follows: In Section 2, we provide thehypotheses development. We present the data and method in Section 3, followed byour analysis of the empirical results in Section 4. Section 5 summarizes and offers someconcluding remarks.

2 Hypotheses Development

Acquiring firms usually take into account investors’ behavior (Luo, 2005) and partic-ularly, CEOs listen to the market (Kau et al., 2008). We focus on the determinantsof the degree to which CEOs listen to the market. Our aim is to determine whatare the mechanisms which lead CEOs to be more receptive to the signals sent by theinvestors. We define these determinants from the typology of Finkelstein (1992) whodistinguishes structural, ownership, expertise, and prestige powers. We position ourhypotheses on the basis of this classification. Concerning ownership and structuralpower, our expectations are unequivocal. However, concerning structural and exper-tise powers, literature results are mixed. Therefore, predictions are ambiguous thus wedevelop the different arguments. We summarize our hypotheses in figure 1.

2.1 Structural Power

Structural power refers to CEO duality, i.e, the CEO is as well the CEO of the firmas the chairperson of the board of directors. Therefore, structural power providesdecisions-making power. Thus CEOs may cope with criticisms and take responsibilityif their decisions are not successful (Oler et al., 2007). We expect that structural powerinfluences negatively, ceteris paribus, the likelihood to consummate the deal when itcaused a negative market reaction.

However, Baliga et al. (1996) suggest that CEO duality may avoid the dilution of theCEO leadership by reducing the probability of overlapping between the decisions andthe objectives of the management and the board. Furthermore, Baliga et al. observethat CEO duality do not create a competition between the chairman of the boardand the CEO. CEO duality reduces the governance control of the board. Therefore,CEOs who are also chairman of the board rely more on their beliefs and may be more

3

Figure 1: Summary of Hypotheses

overconfident. A high structural power leads CEO not to listen to the signals sent bythe stock market.

Hypothesis 1: CEO structural power affects the probability of completing a deal thatcauses a negative market reaction at its announcement date.

Hypothesis 1’:Under the decision making responsibility argument, we expect thatCEO structural power decreases the probability of completing a deal that causes a neg-ative market reaction at its announcement date.

Hypothesis 1”: Under the overconfidence argument, we expect that CEO structuralpower increases the probability of completing a deal that causes a negative market re-action at its announcement date.

2.2 Ownership Power

The CEO’s own financial investment in the firm’s capital implies that the managersmay be directly involved in the corporate investment decisions. Managerial ownershipcontributes to internal mechanism to reduce agency costs, i.e., managers with a lowownership power maximize their own utility at the expense of the value creation ofshareholders (Jensen and Meckling, 1976; Fama and Jensen, 1983). The relationshipbetween the CEO behavior towards a negative market reaction at an investment an-nouncement date and his ownership concerns is relatively intuitive. CEOs that areconcerned by high ownership power have less diversified situation. Thus the moreCEOs have assets in the firm; the more they will avoid risky strategy. Ownershippower determines positively the degree to which CEOs listen to the market.

4

Hypothesis 2: CEO ownership power decreases the probability of completing a deal thatcauses a negative market reaction at its announcement date.

2.3 Expertise Power

Expertise power refers to the capacity to foresee, better and quicker than competitors,the market behavior. Experienced CEOs are more able to interpret market signals.Previous literature often considers tenure (Walters et al., 2007) and CEO prior acqui-sition experience (Oler et al., 2007). At the beginning of their careers, CEOs oftenhope to grow the firm, which implies adopting more risky strategies, without focusingnecessarily on market reactions. Longer CEO tenure instead corresponds to a decreaseof the CEO’s propensity to acquire information about the process of acquisition, devel-opment, and growth (Hambrick and Fukotomi, 1991; Finkelstein and Hambrick, 1996;Audia et al., 2000; Kroll et al., 2000). Those CEOs who are about to retire, have aneven higher risk aversion (Bertrand and Schoar, 2003). Moreover, a longer tenure mayresult in stronger control of the board. These elements all refer to the status quo ofCEOs (Hambrick and Fukotomi, 1991; Finkelstein and Hambrick, 1996) and suggestthat CEOs with a long tenure may not complete a deal to which the market reactionhas been negative.

“There is little reason to expect that a particular individual bidder will refrain frombidding because he has learned from his own past errors. Although some firms engagein many acquisitions, the average individual bidder/manager has the opportunity tomake only a few takeover offers during his career. They may convince themselves thatthe valuation is right and that the market does not reflect the full economic value ofthe combined firm.” (Roll, 1986, p199). We consider that the number of acquisitionsdone by the CEO could be more reliable for estimating the expertise power in theacquisition policy context but that this number is not enough for learning from hispast acquisitions. CEOs will base themselves more on their intuition.

Hypothesis 3: CEO expertise power affects the probability of completing a deal thatcauses a negative market reaction at its announcement date.

Hypothesis 3’: Under the learning argument, CEO expertise power decreases theprobability of completing a deal that causes a negative market reaction at its announce-ment date.

Hypothesis 3”: Under the know-how argument, CEO expertise power increases theprobability of completing a deal that causes a negative market reaction at its announce-ment date.

2.4 Prestige Power

Prestige power relates on the advantages that CEOs obtain from their social networksand connections, which play a considerable role in the boards’ composition.

The French CEO market is characterized by two managerial elites, namely, CEOsfrom the Ecole Nationale d’Administration (ENA) and the Polytechnique (X). Theseschools provide the only access to ministerial secretary positions, which are very con-nected.2 These civil servants, regardless of the length of their careers, often get ap-

2The best students from Polytechnique often enter Les Mines or Pont et Chaussees to becomehigh-ranking civil servants and those from ENA often get jobs with Inspection des Finances, La Cour

5

pointed to corporate boards or even take CEO posts in notable firms (Bauer andBertin-Mourot, 1995). For example, a high-ranking civil servant of the French Statehas been recently appointed CEO of BPCE, one of the biggest European banks. Theappointment of a high-ranking civil servant in the head of big French firms is fre-quent. In 2008, 50% of the CAC40 CEOs came from the ENA or X (Dameron, 2008).Education-based networks exist not only because of the comradeship of the membersof experience during their studies but also as a result of the different meetings inwhich they participate for their title promotion or the entries in their address books.Moreover, by participating on other firms’ boards, CEOs can extend their networks.Seats on these boards enable them to access better quality information and establishpartnerships.

The consequences of this power are multiple. First, prestige power likely increasesthe CEOs’ ego and grants them more confidence on their judgments. According toBarber et al. (1995), CEO network ties among managerial elites increase the likelihoodof an acquisition event, because CEOs who belong to favored networks overestimatethe accuracy of their own judgment, even in the face of negative market reactions. Sec-ond, Beckman and Haunschild (1998) distinguish two types of acquisition information:private information (e.g., network, friendship) and public (e.g., reports, communica-tion). Private information has greater impact on CEOs’ decisions, because it suggests acompetitive advantage. Well-connected CEOs have access to private information thatmay encourage them to pursue decisions disapproved of by the market. Third, prestigepower grants CEO more legitimacy and endows them with an important persuasivecapacity during negotiations. For example, CEOs might develop friendly relations withthe board of directors to divert their attention away from disagreeable elements andto obtain their consent (Lauenstein, 1977). Fourth, their networks protect connectedCEO from lay offs and help them find new jobs more easily (Nguyen-Dang, 2008). As aconsequence, they take more risks and they complete an acquisition even if the marketdoes not approve it. Figure 2 shows the double informative relationship between the ac-quirer and the market during an acquisition announcement period. The characteristicsof acquisition (the bidding firms, the target firms, the mean of payment, the number ofcompetitors, ...) are interpreted by the market. The market sends its opinion throughits reaction. Other types of information are considered during this investment periodand more particularly when networks are important. On one hand, networks can pro-vide an informational advantage to the bidder comparing to the market. On the otherhand, research on decision making conclude that managers assign greater weight topersonal sources than to impersonal sources (McDonald and Westphal, 2003). Peo-ple often persevere in their beliefs and persist despite negative performance outcomes.They seek information from sources that are more likely to affirm those beliefs and toavoid sources that are more likely to disconfirm them (confirmation bias). Thus net-works can help CEOs to confirm their initial judgment and he will not revise correctlyhis decision.

Hypothesis 4: CEO prestige power increases the probability of completing a deal whichthat causes a negative market reaction at its announcement date.

des Comptes, or Le Conseil d’Etat.

6

Figure 2: The Information Flows by Considering the Networks Influence

3 Empirical Method

3.1 Sample Selection

We collected a sample of M&A announcements from the Securities Data Company’s(SDC) database. The announcements we gather all satisfy the following criteria: (1)They occurred between January 1, 2000, and December 31, 2005; (2) both the acquiringand target firms are publicly traded; (3) the acquiring firms are French; (4) the dealvalue is e1 million or higher; (5) the expected percentage owned by the acquirer afterthe transaction is at least 50%, and (6) the acquiring firms appear in the SBF250 indexbetween 2000 and 2005.3

The stock price data come from Datastream. We hand-collected the CEOs’ charac-teristics using Who’s Who and the Factiva database. We obtained information aboutthe characteristics of the corporate governance of acquiring firms from Dafsalien, andwe hand-collected the corporate governance characteristics of the acquiring firms toconstruct a corporate governance index. Financial statements come from ThomsonFinancial. Data unavailability in some cases required us to delete some acquisitionannouncements. Our final sample also does not include multiple acquisition announce-ments by an acquiring firm on the same day, because the information sent to the marketmay overlap. Thus, we obtain a final sample of 205 acquisition announcements.

Table 1 and figure 3 reveal the frequencies of completed and uncompleted acqui-sitions for our sample of 205 acquisitions announcements between January 1, 2000,and December 31, 2005. Approximately 76% of the sample, or 156 announcements,achieved success during the six-year period, which is generally consistent with the over-all M&A market (Bauguess and Stegemoller, 2008). At end of 2000, the financial crisiscorresponds to a decline in acquisitions (Kau et al., 2008). Moreover, we only identify16 acquisition announcements in 2003. We also observe that our sample contains 134

3The SBF250 (Societe des Bourses Francaises 250 Index) is a French stock market index represent-ing all sectors of the French economy.

7

Table 1: Acquisition Announcements FrequencyThis figure presents the year-by-year acquisition announcements frequencies, for both uncompletedand completed acquisition announcements, and acquiring firms and CEO frequencies. Our samplecontains 260 acquisitions announced between January 1, 2000, and December 31, 2005. The acquiringand targets firms are publicly traded. The deal value is e1 million or more.

Year Acquisitions Completed Uncompleted FirmAnnouncements Acquisitions Acquisitions Frequency

2000 65 54 11 302001 46 34 12 302002 27 20 7 212003 16 12 4 142004 26 16 10 232005 25 20 5 16Total 205 156 49 134

Figure 3: Acquisition Announcements FrequencyThis figure presents the year-by-year acquisition announcements frequencies. The minimum dealvalue is e1 million, the minimum percentage of shares that acquiring firms sought to own after thetransaction is 50%, and all deals were announced between January 1, 2000, and December 31, 2005.

0

10

20

30

40

50

60

70

2000 2001 2002 2003 2004 2005

Year

Acq

uisi

tio

n A

nno

unc

emen

ts

0.0010.0020.0030.0040.0050.0060.0070.0080.0090.00

Per

cent

of

Com

plet

ed

Acq

uisi

tion

Acquisition Announcements % of Completed Acquisition

acquiring firms. It means that each bidding firm announces, on average, about 1.5acquisitions.

3.2 Method

To examine the relationship between CEO power and the probability of M&A’s success,we estimate the following probabilistic model:

prob(ACQ = 1) =exp(µ + γX)

1 + exp(µ + γX), (1)

where ACQ is a dummy variable that takes the value of 1 if the acquisition is completed,and 0 otherwise; and X is the vector of independent variables, including those relatedto the market reaction to the acquisition announcements, CEO power, and the controlvariables pertaining to corporate governance, the deal, and the financial statements ofacquiring firm (See the Appendix 1 for the variables description).

8

Market Reaction to M&A Announcements Luo (2005) suggests that the mar-ket reaction to M&A announcements provides a means to predict if the deal will becompleted. To measure the CEO’s ability to complete an acquisition, despite negativemarket reactions, we include a market reaction variable to the acquisition announce-ments (CARi). We calculate the returns from the event study methodology and use themarket-adjusted returns model to estimate the abnormal component of returns (AR)of stock i on day t:

ARi,t = Ri,t − (αi + βiRm,t), (2)

where Ri,t indicates the returns of stock i at time t, Rm,t is the market return at time t,αi reveals the abnormal return of stock i, and t = 0 is the announcement date.4 We usethe returns of the SBF250 index as the measure of market returns, though the resultsare robust against various return measures. Following Brown and Warner (1985), weestimate the coefficients α and β of stock i over the period from 300 days to 15 daysbefore the acquisition announcement. Then we estimate the cumulative abnormalreturns for acquiring firms over the 7 trading days that surround the announcementdates [-3 days; +3 days]:

CARi =3∑

t=−3

ARi,t. (3)

This window of 7 trading days helps us control for possible news leaks and allowsinvestors time to gather additional information.

CEO Power We use proxies of structural, ownership, expertise, and prestige powers.Structural power uses a dummy variable that takes a value of 1 if the CEO of theacquiring firms also is the chairman of the board (CEO Duality). To test whetherCEO has ownership power, we include the percentage of equity owned by CEO (CEOOwnership) and a dummy variable that equals one if CEO is the founder or the heirof founder family (Family Firm). We use CEO tenure (Tenure) and prior acquisitionexperience (Acquisition Experience) as proxies of the CEO’s expertise power. Finally,we consider two proxies to measure the prestige power of the CEOs. First, we notethe networks resulting the CEOs’ education in Grandes Ecoles. We include a dummyvariable that equals to 1 if the CEO is a graduate of Ecole Nationale d’Administrationor Polytechnique (Elite Education). Second, we take into account the networks thatstem from the CEOs’ presence on the other firms’ boards. The corporate directorshipsvariable equals the number of CEO directorship (Outside Boards). Our considerationis not exhaustive and excludes some networks in France (e.g., La Franc-Maconnerie,Le Siecle) the members of these latter networks are difficult to identify.

Control Variables We also include independent variables that control corporategovernance, financial statements, and deals, drawing mostly from finance literature.Following Gompers et al. (2003)’s work, Bollaert et al. (2009) developed a French cor-porate governance index (CG Index) whose items reflect the specificities of corporategovernance in the French context. The items used are merged into four categories:ownership structure, board composition, board mechanisms, and transparency. The cat-egory ownership structure includes three items: ultimate owners, family control, andpresence of a blocking minority. The category board composition refers to the former

4We calculate the returns as follows: Ri,t = ln( Pi,t

Pi,t−1), where Pi,t is the price of stock i on day t.

9

Figure 4: Networks in CAC40 Boards

of “ENA” or “Polytechnique” of CEO. The category board mechanisms contains itemson the existence of governance, remuneration, audit or nomination committees. Thecategory transparency involves items negatively perceived for the firm’s transparency:absence of details concerning the board evaluation, charter unavailable, and poor in-formation about the directors’ backgrounds. Following the study of Gompers et al.,we add, for each firm, one point (+1) for every characteristic that reduces shareholderrights. The higher score, the weaker is governance. We also use the size of board ofdirectors (Board Size) to control monitoring mechanisms that supervise the acquisitiondecisions of the CEOs. We expect an inverted U-shaped relationship between boardsize and the degree of board monitoring, so we also adopt a quadratic form of the boardsize variable in our analysis. Firms with a large board benefit from protective gover-nance (Bauguess and Stegemoller, 2008). However, the oversized boards are less likelyto control effectively CEOs (Jensen, 1993; Yermack, 1996). Considering the leverageof the acquiring firm (Leverage) (Harford, 1999), we expect an inverse relationshipbetween this level and the probability of an acquisition’s success. The size of a firmmay also have an impact on acquisitions. Bigger firms are more likely to complete anacquisition, therefore we use market value (Size) as a measure of firm size. Greaterperformance increases the probability of completing an acquisition (Morck et al., 1990),prompting us to use return on equity (ROE) to measure the performance of acquir-ing firms. The method of payment to pay the transaction could exert an importantimpact. A cash deal may signal to the market that the acquiring firm’s stock pricesare undervalued (Myers and Majluf, 1984). Cash acquisitions often get consummatedmore quickly (Datta et al., 1992). Therefore, we expect that the effect of a dummyvariable - equal to 1 if the acquisition uses cash (Cash) - is positive on the probability ofacquisition success. Finally, we include dummy variables to control for acquisitions dueto economic and technological shocks (Mitchell and Mulherin, 1996; Harford, 2005).5

4 Empirical Results

4.1 CEOs’ and Acquiring Firms’ Characteristics

We present the descriptive statistics regarding the CEOs’ and acquiring firms’ charac-teristics in table 2 (full sample) and table 3 (two subsamples).

5Although the industry and year dummy variables are included in all our models, for the sake ofaesthetics, we do not present them in our results.

10

Table 2: Summary Statistics for Acquiring Firms’ and CEOs’ CharacteristicsThis table presents descriptive statistics of the variables related to CEO power, corporate governanceand financial statements for acquiring firms, and the deal. The sample consists of 205 acquisitionannouncements made between January 1, 2000, and December 31, 2005. CEO Duality is a dummyvariable that equals 1 if the CEO also holds the board chair, and 0 otherwise. Family Firm is adummy variable that takes a value of 1 if CEO is also the founder or heir of the founder family, and0 otherwise. CEO Ownership is the proportion of outstanding shares held by the CEO in the firm.Tenure is the number of years the individual has been CEO of the firm. Acquisition Experience is thenumber of M&As previously made by the CEO. Elite Education is a dummy variable that takes a valueof 1 if the CEO has a degree from an elite French school (ENA or X), and 0 otherwise. Outside Boardsis the number of CEO directorships. CARi, expressed as a percentage, is the cumulated abnormalreturn surrounding M&A announcement of the acquiring firm. CG Index is the corporate governanceindex. Board Size is the number of directors on the board. Size, expressed in million euros, is theproduct of shares outstanding and share price. Leverage, expressed as a percentage, is equal to debtdivided by assets. ROE, expressed as a percentage, is equal to the result per share divided by returns.Cash is a dummy variable equals 1 if the acquisition is paid for by cash only, and 0 otherwise.

Variable Mean St-Dev. Min. Q1 Med. Q3 Max.Panel A: CEO Power VariablesStructural Power CEO Duality 0.70 0.46 0 0 1 1 1Ownership Power Family Firm 0.18 0.38 0 0 0 0 1

CEO Ownership 0.98 5.55 0 0 0 0 52.35Expertise Power Tenure 7.11 6.52 0 2 5 11 35

Acquisition Experience 21.87 26.52 0 3 14 30 126Prestige Power Elite Education 0.51 0.50 0 0 1 1 1

Outside Boards 11.02 8.40 0 5 10 13 50Panel B: Control VariablesMarket Reaction CARi -0.19 -0.12 -20.16 -2.88 -0.19 2.80 22.26Governance CG Index 0.49 0.16 0 0.38 0.48 0.58 0.88

Board Size 17.71 5.64 5 13 18 21 31Financial Statements Size 4545 11976 65 287 930 2670 73289

Leverage 28.48 12.30 0.00 19.01 30.37 36.40 59.23ROE 12.22 10.75 -45.34 8.34 12.72 17.62 62.60

Deal Cash 0.81 0.39 0 1 1 1 1

The frequency with which CEOs serve as the chairman is high (0.70), though themean value is lower (0.67) for completed acquisitions than for uncompleted acquisi-tions (0.80). The differences in both the mean and median values of this variable aresignificant at the 5% level.

Approximately 18% of CEOs are founders or heirs of the founder’s family. Forcompleted acquisitions, this percentage is 13%, and it rises to 33% for uncompletedacquisitions. We also find, on average, that the portion of capital owned directly by theCEO equals 0.98% in our sample. This CEO ownership is 0.27% in the subsample ofcompleted acquisition but reaches 3.22% for uncompleted acquisitions. This differencealso is significant.

The average of CEO tenure is 7.11 years, and prior CEO acquisition experienceinvolves an average of 21.87 acquisitions. The mean tenure of CEOs is significativelylonger for uncompleted acquisitions (9.20 years) than for completed acquisitions (6.45years).

The frequency with which CEOs is a graduate of elite schools (ENA or X) is 51%.For completed acquisitions, 55% of CEOs come from ENA or X, whereas this percentagedrops to 39% for the subsample of uncompleted acquisitions. On average, CEOs hold

11

11.02 directorships.Almost half of the acquiring firms’ stock experience a negative CAR during the 7-

day window surrounding the announcement of an acquisition. This result is consistentwith previous studies of the wealth creation for acquiring firms’ shareholders aroundM&A announcements. This result suggests that the market reaction to acquisitionannouncements may predict whether the acquiring firms will complete the deal later.

With regard to the corporate governance variables, we find that completed an-nouncements usually involve acquiring firms with lower corporate governance scores.The mean (median) board size is 17.71 (18), 18.39 (18), and 15.76 (16) for the full,completed and uncompleted samples, respectively. That is, the board size for the sub-sample of completed acquisitions is significatively larger than that for uncompletedacquisitions. We also find that for the subsample of completed announcements, acquir-ing firms are larger and have a lower leverage, which suggests that it may be easier tocomplete a deal if firms are less financially constrained.

In table 4, we report the correlation among our variables.6 Among the correlations,the educational background tends to have large boards (0.54) although it is negativelycorrelated with the CEO founder or heir of the founding family. There is a positivecorrelation between CEO tenure and three others CEO power variables: CEO duality,CEO founder, and CEO ownership. The proxies of expertise power are also correlated(0.31). We find that firms with large boards are not associated with a CEO founder(or heir the founding family) and a high CEO ownership.

6We have verified that collinearity is not a problem for the analysis of our model, based on theconditional number and the variance inflation factors of our independent variables.

12

Table 3: Univariate Statistics of Acquiring Firms and CEOs CharacteristicsThis table presents the mean and median univariate statistics related to the variables that describe theCEO characteristics, corporate governance and financial statements for acquiring firms, and the deal.The sample consists of 205 acquisitions announcements made between January 1, 2000, and December31, 2005. CEO Duality is a dummy variable that equals 1 if the CEO also holds the board chair, and0 otherwise. Family Firm is a dummy variable that takes a value of 1 if CEO is also the founder orheir of the founder family, and 0 otherwise. CEO Ownership is the proportion of outstanding sharesheld by the CEO in the firm. Tenure is the number of years the individual has been CEO of the firm.Acquisition Experience is the number of M&As previously made by the CEO. Elite Education is adummy variable that takes a value of 1 if the CEO has a degree from an elite French school (ENAor X), and 0 otherwise. Outside Boards is the number of CEO directorships. CARi, expressed as apercentage, is the cumulated abnormal return surrounding M&A announcement of the acquiring firm.CG Index is the corporate governance index. Board Size is the number of directors on the board. Size,expressed in million euros, is the product of shares outstanding and share price. Leverage, expressedas a percentage, is equal to debt divided by assets. ROE, expressed as a percentage, is equal to theresult per share divided by returns. Cash is a dummy variable equals 1 if the acquisition is paid forby cash only, and 0 otherwise. ***, **, and * indicate statistical significance at the 1%, 5%, and 10%levels, respectively.

VariableTotal Sample ACQ = 0 ACQ = 1 DifferencesN=205 N=49 N=156Mean Median Mean Median Mean Median Mean T Stat. Median Z Stat.

Panel A: CEO Power VariablesStructural PowerCEO Duality 0.70 1.00 0.80 1.00 0.67 1.00 −0.12∗∗ -2.01 −0.00∗∗ -1.99Ownership PowerFamily Firm 0.18 -0.00 0.33 -0.00 0.13 -0.00 −0.20 0.88 −0.00 0.88CEO Ownership 0.98 -0.00 3.22 -0.00 0.27 -0.00 −2.95∗∗∗ 3.33 −0.00∗∗ 2.37Expertise PowerTenure 7.11 5.00 9.20 7.00 6.45 5.00 −2.76 -0.63 −2.00 0.10Acquisition Experience 21.87 14.00 17.90 12.00 23.12 14.00 5.22 1.10 2.00 0.53Prestige PowerElite Education 0.51 1.00 0.39 -0.00 0.55 1.00 0.16 -1.20 1.00 -0.26Outside Boards 11.02 10.00 11.55 8.00 10.86 10.00 −0.69∗∗ -2.38 2.00∗∗ -2.04Panel B: Control VariablesMarket ReactionCARi -0.19 -0.12 -0.69 -0.81 0.05 -0.08 0.75 1.55 0.73 1.54GovernanceCG Index 0.49 0.48 0.47 0.48 0.49 0.48 0.02 0.50 −0.00 -1.00Board Size 17.71 18.00 15.76 16.00 18.33 18.00 2.57∗∗∗ -2.83 2.00∗∗ -2.08Financial StatementsSize 4545 930 3816 758 4774 1028 958.00 -0.49 271.00 -0.78Leverage 28.48 30.37 30.16 30.80 27.95 30.37 −2.21∗∗∗ 3.25 −0.43∗∗∗ 3.18ROE 12.22 12.72 12.97 12.43 11.98 12.80 −0.99∗∗∗ 2.62 0.37 0.86DealCash 0.81 1.00 0.86 1.00 0.80 1.00 −0.06 0.56 −0.00 -0.34

13

Tab

le4:

Cor

rela

tion

Mat

rix

Thi

sta

ble

pres

ents

the

corr

elat

ion

mat

rix.

The

sam

ple

cons

ists

of20

5ac

quis

itio

nan

noun

cem

ents

mad

ebe

twee

nJa

nuar

y1,

2000

,an

dD

ecem

ber

31,20

05.

CEO

Dua

lity

isa

dum

my

vari

able

that

equa

ls1

ifth

eC

EO

also

hold

sth

ebo

ard

chai

r,an

d0

othe

rwis

e.Fa

mily

Fir

mis

adu

mm

yva

riab

leth

atta

kes

ava

lue

of1

ifC

EO

isal

soth

efo

unde

ror

heir

ofth

efo

unde

rfa

mily

,an

d0

othe

rwis

e.C

EO

Owne

rshi

pis

the

prop

orti

onof

outs

tand

ing

shar

eshe

ldby

the

CE

Oin

the

firm

.Ten

ure

isth

enu

mbe

rof

year

sth

ein

divi

dual

has

been

CE

Oof

the

firm

.Acq

uisi

tion

Exp

erie

nce

isth

enu

mbe

rof

M&

As

prev

ious

lym

ade

byth

eC

EO

.Elit

eEdu

cation

isa

dum

my

vari

able

that

take

sa

valu

eof

1if

the

CE

Oha

sa

degr

eefr

oman

elit

eFr

ench

scho

ol(E

NA

orX

),an

d0

othe

rwis

e.O

utsi

deBoa

rds

isth

enu

mbe

rof

CE

Odi

rect

orsh

ips.

CA

Ri,

expr

esse

das

ape

rcen

tage

,is

the

cum

ulat

edab

norm

alre

turn

surr

ound

ing

M&

Aan

noun

cem

ent

ofth

eac

quir

ing

firm

.C

GIn

dex

isth

eco

rpor

ate

gove

rnan

cein

dex.

Boa

rdSi

zeis

the

num

ber

ofdi

rect

ors

onth

ebo

ard.

Size

,ex

pres

sed

inm

illio

neu

ros,

isth

epr

oduc

tof

shar

esou

tsta

ndin

gan

dsh

are

pric

e.Le

vera

ge,ex

pres

sed

asa

perc

enta

ge,is

equa

lto

debt

divi

ded

byas

sets

.RO

E,ex

pres

sed

asa

perc

enta

ge,is

equa

lto

the

resu

ltpe

rsh

are

divi

ded

byre

turn

s.C

ash

isa

dum

my

vari

able

equa

ls1

ifth

eac

quis

itio

nis

paid

for

byca

shon

ly,an

d0

othe

rwis

e.**

*,**

,an

d*

indi

cate

stat

isti

calsi

gnifi

canc

eat

the

1%,5%

,an

d10

%le

vels

,re

spec

tive

ly.

Var

iabl

es(1

)(2

)(3

)(4

)(5

)(6

)(7

)(8

)(9

)(1

0)(1

1)(1

2)(1

3)(1

4)C

EO

Dua

lity

(1)

1.00

0.10

0.07

0.34∗∗∗

0.28∗∗∗

0.11

−0.0

8−0

.05

−0.0

00.

04−0

.08

0.03

−0.0

60.

02Fa

mily

Fir

m(2

)0.

101.

000.

26∗∗∗

0.31∗∗∗−0

.02

−0.3

2∗∗∗

0.09

0.06

−0.0

9−0

.37∗∗∗

0.14∗∗

−0.0

20.

020.

09C

EO

Owne

rshi

p(3

)0.

070.

26∗∗∗

1.00

0.34∗∗∗−0

.11

−0.1

3∗0.

040.

040.

06−0

.30∗∗∗−0

.05

0.01

−0.0

00.

04Ten

ure

(4)

0.34∗∗∗

0.31∗∗∗

0.34∗∗∗

1.00

0.31∗∗∗−0

.13∗

0.25∗∗∗−0

.05

0.12∗−0

.15∗∗−0

.05

0.00

0.11

0.14∗∗

Acq

uisi

tion

Exp

erie

nce

(5)

0.28∗∗∗−0

.02

−0.1

10.

31∗∗∗

1.00

0.37∗∗∗

0.07

−0.1

10.

18∗∗

0.32∗∗∗−0

.01

0.12∗

0.08

−0.0

4Elit

eEdu

cation

(6)

0.11

−0.3

2∗∗∗−0

.13∗

−0.1

3∗0.

37∗∗∗

1.00

−0.0

6−0

.08

0.16∗∗

0.54∗∗∗−0

.01

0.13∗

−0.0

8−0

.11

Out

side

Boa

rds

(7)

−0.0

80.

090.

040.

25∗∗∗

0.07

−0.0

61.

000.

070.

16∗∗−0

.04

0.02

0.10

0.04

0.03

CA

Ri(8

)−0

.05

0.06

0.04

−0.0

5−0

.11

−0.0

80.

071.

000.

04−0

.15∗∗

0.04

−0.0

40.

000.

07C

GIn

dex

(9)

−0.0

0−0

.09

0.06

0.12∗

0.18∗∗

0.16∗∗

0.16∗∗

0.04

1.00

−0.0

7−0

.07

0.01

0.01

−0.0

1Boa

rdSi

ze(1

0)0.

04−0

.37∗∗∗−0

.30∗∗∗−0

.15∗∗

0.32∗∗∗

0.54∗∗∗−0

.04

−0.1

5∗∗−0

.07

1.00

−0.1

5∗∗−0

.01

0.15∗∗

−0.0

8Si

ze(1

1)−0

.08

0.14∗∗

−0.0

5−0

.05

−0.0

1−0

.01

0.02

0.04

−0.0

7−0

.15∗∗

1.00

−0.2

2∗∗∗

0.02

−0.1

1Le

vera

ge(1

2)0.

03−0

.02

0.01

0.00

0.12∗

0.13∗

0.10

−0.0

40.

01−0

.01

−0.2

2∗∗∗

1.00

−0.0

90.

09RO

E(1

3)−0

.06

0.02

−0.0

00.

110.

08−0

.08

0.04

0.00

0.01

0.15∗∗

0.02

−0.0

91.

000.

23∗∗∗

Cas

h(1

4)0.

020.

090.

040.

14∗∗

−0.0

4−0

.11

0.03

0.07

−0.0

1−0

.08

−0.1

10.

090.

23∗∗∗

1.00

14

4.2 Logit Estimates of the Likelihood of Completed Acquisi-tion

4.2.1 Completed Acquisition and Market Reaction

Luo (2005) finds that the market reaction around M&A announcements enables to pre-dict the success of that acquisition. Relying on Luo’s (2005) results, we test whetherthe bidding firms later consummate the deal after recognizing the market’s reactionto the acquisition announcement. Our results are consistent with Luo’s (2005) result.As we show in table 5, the probability of success of the acquisition decreases significa-tively when the market reaction to the announcement is worse. We also anticipated acurvilinear, inverted U-shaped relationship between the board size and the probabilityto complete a deal. In our model, we use the quadratic form of board size and wefind that it is indeed significantly negatively related to the probability to completean acquisition. The point of inflection in the relationship between the board size andthe probability to complete deal is approximatively 20 directors. We thus proposeextending this model to investigate the effects of the CEO’s characteristics.

Table 5: Logistic Regression of the Probability of Completing an AcquisitionThis table presents the results of logistic regression of the probability of completing an acquisition. Thedependent variable is ACQ, a dummy variable, that equals 1 if the firm has completed an acquisition,and 0 otherwise. The sample consists of 205 acquisition announcements made between January 1,2000, and December 31, 2005. CARi, expressed as a percentage, is the cumulated abnormal returnsurrounding M&A announcement of the acquiring firm. CG Index is the corporate governance index.Board Size is the number of directors on the board. Size, expressed in million euros, is the product ofshares outstanding and share price. Leverage, expressed as a percentage, is equal to debt divided byassets. ROE, expressed as a percentage, is equal to the result per share divided by returns. Cash is adummy variable equals 1 if the acquisition is paid for by cash only, and 0 otherwise. ***, **, and *indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

Explanatory ACQ = 1Variables (N = 205)Intercept −2.859CARi 5.975∗

CG Index 0.182Board Size 0.462∗∗∗

Board Size2 −0.011∗∗

logSizet−1 0.157Leverage −0.015ROE −0.014Cash −0.074MaximumLikelihood: χ2=34.62∗∗∗

PercentConcordant: 76.0

R2 Adjusted: 23.29

Although the industry and year dummy variables are

included in our model, for the sake of aesthetics, we do notpresent them in our results.

4.2.2 Completed Acquisition, Market Reaction and CEO Power

In table 6 Panel A, we estimate a logistic model in which the dependent variable isthe probability that the acquiring firms consummate an acquisition; we include CEO

15

characteristics among the independent variables. To take into account the marketreaction, we also use a set of interactive terms. Because we define CEO power as hisor her ability to complete a deal even when the market reaction is negative, we focushere on the results involving the interactive variables. We find negative and significantrelationship between structural power and the probability of completing a deal. Asa result, we accept Hypothesis 1 and Hypothesis 1”. We argue that CEOs who arealso chairman of the board rely more on their beliefs and may be more overconfident.Moreover, the likelihood of completed acquisitions increases with one of our proxiesfor ownership power; that is, the percentage of capital owned by the CEO has not asignificant positive impact on the completion likelihood of an acquisition, not in supportof Hypothesis 2. This result suggest that the ownership power determines if managers’actions subsequent to an acquisition announcement do not learn from market reaction.CEO tenure also has a positive effect, in support of Hypothesis 3 and Hypothesis 3’.Expertise power affects positively the degree to which CEOs listen to the market.Finally, our main result pertains to the strong relationship between prestige power andthe likelihood of completing an acquisition. Thus, Hypothesis 4 is supported. CEOswith elite education and a board connections privilege their acquaintances’ advice overmarket signals.

Table 1 shows that the full sample of 205 acquisition announcements is relativelyunbalanced. Thus, we present a confusion matrix in table 6 Panel B. The matrix ofconfusion suggests the accuracy of our model. It correctly predicts 170 announcements,with an average rate of 82.9%.

5 Robustness Tests

5.1 Heteroscedasticity

To take into account the potentiel heteroscedasticity issue of abnormal returns, westandardize abnormal returns by their standard deviation on estimation period (Patell,1976; Boehmer et al., 1991). We calculate the standardized cumulated abnormal re-turns:

SCARi,t =ARi,t

σARi,t

√N

, (4)

where σARi,tindicates the standard deviation of abnormal returns calculated on esti-

mation period of stock i and N is the number of days contained in the event period.We provide the results of our regression in table 7. Thus, before capturing the relativestrength of the signal, we find that our results are consistent with our previous results.The market signals enable CEOs to gain experience. We also observe that structuraland prestige powers increase the probability of an acquisition whose the announcementcauses a negative market reaction although expertise power increases this likelihood.

5.2 Sample Splitting

To test the robustness of our results, we also split our sample into two groups accordingthe market signals. We provide the results of these regressions in table (8). We onlykeep the significant control variables in table 6. Our results are also consistent withthe previous conclusions. For the subsample of negative market reaction, we observethat prestige power increases the probability to complete a deal. For the subsample

16

Table 6: Logistic Regression of the Probability of a Completed AcquisitionThis tables presents the results of the logistic regression of the probability of completing an acquisi-tion. The dependent variable is ACQ, a dummy variable that equals 1 if the firm has completed anacquisition, and 0 otherwise. The sample consists of 205 acquisition announcements made betweenJanuary 1, 2000, and December 31, 2005. CEO Duality is a dummy variable that equals 1 if the CEOalso holds the board chair, and 0 otherwise. Family Firm is a dummy variable that takes a valueof 1 if CEO is also the founder or heir of the founder family, and 0 otherwise. CEO Ownership isthe proportion of outstanding shares held by the CEO in the firm. Tenure is the number of yearsthe individual has been CEO of the firm. Acquisition Experience is the number of M&As previouslymade by the CEO. Elite Education is a dummy variable that takes a value of 1 if the CEO has adegree from an elite French school (ENA or X), and 0 otherwise. Outside Boards is the number ofCEO directorships. CARi, expressed as a percentage, is the cumulated abnormal return surroundingM&A announcement of the acquiring firm. CG Index is the corporate governance index. Board Sizeis the number of directors on the board. Size, expressed in million euros, is the product of sharesoutstanding and share price. Leverage, expressed as a percentage, is equal to debt divided by assets.ROE, expressed as a percentage, is equal to the result per share divided by returns. Cash is a dummyvariable equals 1 if the acquisition is paid for by cash only, and 0 otherwise. ***, **, and * indicatestatistical significance at the 1%, 5%, and 10% levels, respectively.

Panel A: Logit Regressions Modeling the Likelihood of Completing an AcquisitionExplanatory ACQ=1Variables (N=205)Intercept −1.877

Structural Power CEO Duality −0.870CEO Duality*CARi −44.811∗∗∗

Ownership Power Family Firm 0.194CEO Ownership −0.187∗

Family Firm*CARi −1.863CEO Ownership *CARi 2.004

Expertise Power Tenure 0.019logAcquisition Experience −0.044Tenure*CARi 2.608∗∗

logAcquisition Experience*CARi 0.790Prestige Power Elite Education 0.234

Outside Boards −0.022Elite Education*CARi −19.195∗

Outside Boards*CARi −1.315∗

Control Variables CARi 51.713∗∗∗

CG Index −0.966Board Size 0.620∗∗

Board Size2 −0.015∗∗

logSizet−1 0.141Leverage −0.024ROE −0.011Cash −0.343

Although the industry and year dummy variables are included in our model, for the sake of

aesthetics, we do not present them in our results.

Likelihood Maximum: χ2=60.80∗∗∗

Percent Concordant: 83.1R2 Adjusted: 38.47

Panel B: Confusion MatrixPredicted

Uncompleted Completed Total

Actual

Uncompleted 21 28 4942.86% 57.14% 100%

Completed 7 149 1564.49% 95.51% 100%

Total 30 175 205

17

Table 7: Logistic Regression of the Probability of a Completed AcquisitionThis tables presents the results of the logistic regression of the probability of completing an acquisi-tion. The dependent variable is ACQ, a dummy variable that equals 1 if the firm has completed anacquisition, and 0 otherwise. The sample consists of 205 acquisition announcements made betweenJanuary 1, 2000, and December 31, 2005. CEO Duality is a dummy variable that equals 1 if the CEOalso holds the board chair, and 0 otherwise. Family Firm is a dummy variable that takes a value of1 if CEO is also the founder or heir of the founder family, and 0 otherwise. CEO Ownership is theproportion of outstanding shares held by the CEO in the firm. Tenure is the number of years theindividual has been CEO of the firm. Acquisition Experience is the number of M&As previously madeby the CEO. Elite Education is a dummy variable that takes a value of 1 if the CEO has a degreefrom an elite French school (ENA or X), and 0 otherwise. Outside Boards is the number of CEOdirectorships. SCARi is the standardized cumulated abnormal return surrounding M&A announce-ment of the acquiring firm. CG Index is the corporate governance index. Board Size is the number ofdirectors on the board. Size, expressed in million euros, is the product of shares outstanding and shareprice. Leverage, expressed as a percentage, is equal to debt divided by assets. ROE, expressed as apercentage, is equal to the result per share divided by returns. Cash is a dummy variable equals 1 ifthe acquisition is paid for by cash only, and 0 otherwise. ***, **, and * indicate statistical significanceat the 1%, 5%, and 10% levels, respectively.

Explanatory ACQ=1 ACQ=1Variables (N=205) (N=205)Intercept −2.940 −1.307

Structural Power CEO Duality . −0.846CEO Duality*SCARi . −1.978∗∗∗

Ownership Power Family Firm . 0.131CEO Ownership . −0.193∗

Family Firm*SCARi . −0.015CEO Ownership*SCARi . 0.134

Expertise Power Tenure . 0.020logAcquisition Experience . −0.036Tenure*SCARi . 0.136∗∗

logAcquisition Experience*SCARi . −0.022Prestige Power Elite Education . 0.196

Outside Boards . −0.021Elite Education*SCARi . −1.000∗

Outside Boards*SCARi . −0.074∗∗

Control Variables SCARi 0.409∗∗ 2.557∗∗∗

CG Index 0.201 −0.756Board Size 0.478∗∗∗ 0.572∗∗

Board Size2 −0.011∗∗ −0.014∗∗

logSizet−1 0.154 0.142Leverage −0.016 −0.027ROE −0.012 −0.011Cash −0.093 −0.264

Although the industry and year dummy variables are included in our models, for the sake of

aesthetics, we do not present them in our results.

Likelihood Maximum: χ2=36.23∗∗∗ χ2=58.71∗∗∗

Percent Concordant: 76.3 82.2R2 Adjusted: 24.28 37.33

18

Table 8: Logistic Regression of the Probability of a Completed AcquisitionThis tables presents the results of the logistic regression of the probability of completing an acquisitionby subsamples of cumulative abnormal returns. The dependent variable is ACQ, a dummy variablethat equals 1 if the firm has completed an acquisition, and 0 otherwise. The sample consists of 205acquisition announcements made between January 1, 2000, and December 31, 2005. CEO Duality is adummy variable that equals 1 if the CEO also holds the board chair, and 0 otherwise. Family Firm isa dummy variable that takes a value of 1 if CEO is also the founder or heir of the founder family, and0 otherwise. CEO Ownership is the proportion of outstanding shares held by the CEO in the firm.Tenure is the number of years the individual has been CEO of the firm. Acquisition Experience isthe number of M&As previously made by the CEO. Elite Education is a dummy variable that takes avalue of 1 if the CEO has a degree from an elite French school (ENA or X), and 0 otherwise. OutsideBoards is the number of CEO directorships. Board Size is the number of directors on the board. ***,**, and * indicate statistical significance at the 1%, 5%, and 10% levels, respectively.

ExplanatoryVariables

CAR≤0 CAR>0(N=110) (N=95)ACQ=1 ACQ=1

Intercept 3.683 −6.659∗∗

Structural Power CEO Duality 0.199 −2.829∗∗

Ownership Power Family Firm −0.209 0.241CEO Ownership −0.316∗∗ −0.061

Expertise Power logAcquisition Experience −0.327 0.176Tenure −0.078 0.218∗∗

Prestige Power Elite Education 1.263∗ −1.984∗

Outside Boards 0.011 −0.076Control Variables Board Size −0.141 1.306∗∗∗

Board Size2 0.004 −0.031∗∗∗

Although the industry dummy is included in our models, for the sake of aesthetics, we do not

present them in our results.

Likelihood Maximum: χ2=28.99∗∗∗ χ2=36.77∗∗∗

Percent Concordant: 78.6 90.3R2 Adjusted: 33.31 51.67

of positive market reaction, our results suggest that structural and prestige powersdecrease the probability to consummate a deal later and expertise power decreases it.

5.3 Collinearity Issues

To test the potential collinearity issues, we examine the associations between the CEOcharacteristics and the market reaction variables. We test the relevance of our modelby considering if CEO power characteristics determine the market reaction around theacquisition announcements. We measure the frequency of variables relative to CEOpower according to the level of the market reaction of acquiring firms around theannouncement. To determine the existence of a contingency, we calculate statisticsthat test the null hypothesis of no association. From table 9, we conclude that there isno statistically significant difference between the level of CEO power and the marketreaction. These results are consistent with the absence of correlation between themarket reaction variable and the CEO power characteristics suggested by table 4.

19

Table 9: Frequency Tables Between Market Reaction and CEO PowersThis table presents the frequency matrix between market reaction and CEO power. CEO Duality is adummy variable that equals 1 if the CEO also holds the board chair, and 0 otherwise. Family Firm isa dummy variable that takes a value of 1 if CEO is also the founder or heir of the founder family, and0 otherwise. CEO Ownership is the proportion of outstanding shares held by the CEO in the firm.Tenure is the number of years the individual has been CEO of the firm. Acquisition Experience isthe number of M&As previously made by the CEO. Elite Education is a dummy variable that takes avalue of 1 if the CEO has a degree from an elite French school (ENA or X), and 0 otherwise. OutsideBoards is the number of CEO directorships. CAR is the cumulated abnormal return surroundingM&A announcement of the acquiring firm. CEOs are characterized as having a low power if CEODuality, Family Firm, or Elite Education is equal to 1 or if CEO Ownership, Tenure, AcquisitionExperience, and Outside Boards is strictly above average. CEOs are characterized as having a lowpower if CEO Duality, Family Firm, or Elite Education is equal to 0 or if CEO Ownership, Tenure,Acquisition Experience, and Outside Boards is below average. The reported coefficients of frequencyis expressed in percents.

ExplanatoryVariables

CEO with a Low Power CEO with a High Power Cramer’s VCARi ≤ 0 CARi > 0 CARi ≤ 0 CARi > 0Structural Power

CEO Duality 16.10 13.66 37.56 32.68 -0.006Ownership Power

Family Firm 44.88 37.56 8.78 8.78 -0.034CEO Ownership 51.22 42.93 2.44 3.41 -0.060Expertise Power

Tenure 31.71 29.27 21.95 17.07 0.042Acquisition Experience 35.12 29.76 18.54 16.59 -0.013

Prestige PowerElite Education 24.39 24.39 29.27 21.95 0.072Outside Boards 36.59 27.80 17.07 18.54 -0.085

20

6 Conclusion

Since the role of CEO power remains contested position within firms, this study bringsup the CEO power question in the acquisition activity. We studied the CEO listeningdegree to market reactions to acquisition announcements.

Empirical studies conclusions about this question are mixed. According to Luo(2005), firms learn from the market and are more willing to do it when the marketis better informed than itself. Kau et al. (2008) support that managers listen to themarket especially when they are likely to be concerned by agency costs. Jenningsand Mazzeo (1991) results suggest that CEOs rely more on private information thanon the consensus. We considered the different conclusions and analyzed the determi-nants of the CEOs’ attention paid to the to the market’s assessment at the acquisitionannouncement.

The listening determinants we used are based on the CEO power classificationoffered by Finkelstein (1992), which includes structural, ownership, expertise, andprestige powers. We studied the French context because it is characterized by strongnetworks through education or directorships which enable us to obtain an original pres-tige power measure. We based our empirical analysis on a sample of French acquisitionannouncements from 2000 to 2005.

We assume that structural power can increase as well as decrease the probability tocomplete a deal when the investors has been negatively reacted. The CEOs being thechairman of the board tend to be more overconfident because of their position in thefirm. Ownership power leads the CEO to share the market’s opinion. CEOs concernedby high financial implications are more risk-averse. This result is consistent with theagency costs theory. According to CEO tenure, we expect that CEOs take less riskand learn from their past errors with a high tenure. In another way, we assume thattheir acquisitions experience determines the significant learning process. Our resultsconfirm our hypothesis and they suggest that prestige power is associated with a highdegree of listening to stock market. Finally, we support our major hypothesis thatnetworks play a considerable role in the decision process. In fact, our results show thatCEO prestige power increases the likelihood that a deal will be completed althoughnot approved by the market investors.

The financial implications of the CEO behavior towards the market reaction duringthe acquisition announcement period were beyond the scope of our paper. Are well-connected managers a benefit or a cost to the companies they manage? This remainsto be discussed. From the one hand, networks improve the flow of information thusreducing the asymmetry information among managers. From the other hand, socialconnections can lead to lower monitoring and more freedom for the managers which canbe used for self-serving, and the detriment of shareholders’ wealth. We will extend thiswork by considering the financial implications of acquisitions made by well-connectedCEOs.

References

Andrade, G., M. Mitchell, and E. Stafford (2001). New Evidence and Perspectives onMergers. Journal of Economic Perspectives 15 (2), 103–120.

Audia, P., E. Locke, and K. Smith (2000). The Paradox of Success: An Archival and a

21

Laboratory Study of Strategic Persistence Following Radical Environmental Change.Academy of Management Journal 43, 837–853.

Baliga, B., R. Moyer, and R. Rao (1996). CEO Duality and Firm Performance: What’sThe Fuss? Strategic Management Journal 17, 41–53.

Barber, B., D. Palmer, and J. Wallace (1995). Determinants of Conglomerate andPredatory Acquisitions: Evidence from the 1960s. Journal of Corporate Finance 1,283–318.

Bauer, M. and B. Bertin-Mourot (1995). La Tyrannie du Diplome Initial et la Circu-lation des Elites: La Stabilite du Modele Francais. In Le Recrutement des Elites enEurope, pp. 48–63. La Decouverte.

Bauguess, S. and M. Stegemoller (2008). Protective Governance Choices and the Valueof Acquisition Activity. Journal of Corporate Finance 14, 550–566.

Beckman, C. and P. Haunschild (1998). When Do Interlocks Matter?: AlternateSources of Information and Interlock Influence. Administrative Science Quar-terly 43 (4), 815–844.

Bertrand, M. and A. Schoar (2003). Managing with Style: The Effect of Managers onFirm Policies. Quarterly Journal of Economics 118 (4), 1169–1208.

Boehmer, E., J. Musumeci, and A. Poulsen (1991). Event–Study Methodology UnderConditions of Event Induced–variance. Journal of Financial Economics 30, 253–272.

Bollaert, H., H. Daher, M. Declerk, and A. Deroo (2009). Corporate Governance andPerformance of French Listed Companies. Working Paper .

Bradley, M., A. Desai, and E. Kim (1983). The Rational Behind Interfirm TenderOffers: Information or Synergy? Journal of Financial Economics 11, 183–206.

Brown, R. and N. Sarma (2007). CEO Overconfidence, CEO Dominance and CorporateAcquisitions. Journal of Economics and Business 59 (5), 358–379.

Brown, S. and J. Warner (1985). Usual Daily Stock of Event Studies. Journal ofFinancial Economics 14, 3–31.

Burt, R. (2000). The Network Structure of Social Capital. In Research in Organiza-tional Behavior, Volume 22, pp. 345–431. Greenwich, CT: JAI Press.

Collis, D. and C. Montgomery (1998). Creating Corporate Advantage. Harvard Busi-ness Review 76 (3), 70–83.

Combs, J., D. Ketchen, A. Perryman, and M. Donahue (2007). The Moderating Effectof CEO Power on the Board Composition Firm Performance Relationship. Journalof Management Studies 44 (8), 1299–1323.

Dameron, S. (2008). La Genealogie des Nouveaux Dirigeants. In Le Nouveau Visagedes Dirigeants du CAC40, Chapter 1, pp. 7–16. Paris: Pearson Edition.

Datta, D., G. Pinches, and V. Narayanan (1992). Factors Influencing Wealth Creationfrom Mergers and Acquisitions: A Meta–Analysis. Strategic Management Journal 13,67–84.

22

Dye, R. and S. Sridhar (2002). Ressource Allocation Effects of Price REactions toDisclosures. Comtemporary Accounting Research 19 (3), 385–410.

Edlin, A. and J. Stiglitz (1995). Discouraging Rivals: Managerial Rent–Seeking andEconomic Inefficiencies. American Economic Review 85 (5), 1301–1312.

Fama, E. and M. Jensen (1983). Separation of Ownership and Control. Journal of Lawand Economics 26, 301–325.

Finkelstein, S. (1992). Power in Top Management Teams: Dimensions, Measurement,and Validation. Academy of Management Journal 35, 505–538.

Finkelstein, S. and D. Hambrick (1996). Strategic Leadership: Top Executives and theirEffects on Organizations. New–York: West Publishing.

Fuller, K., J. Netter, and M. Stegemoller (2002). What Do Returns to AcquiringFirms Tell Us? Evidence From Firms That Make Many Acquisitions. The Journalof Finance 57 (4), 1763–1794.

Gompers, P., J. Ishii, and A. Metrick (2003). Corporate Governance and Equity Prices.Quarterly Journal of Economics 118 (1), 107–155.

Gupta, A., R. LeCompte, and L. Misra (1997). Acquisitions of Solvent Thrifts: WealthEffects and Managerial Motivations. Journal of Banking and Finance 21 (10), 1431–1450.

Hambrick, D. and G. Fukotomi (1991). The Seasons of a CEO’s Tenure. The Academyof Management Review 16, 719–742.

Harford, J. (1999). Corporate Cash Reserves and Acquisitions. The Journal of Fi-nance 54 (6), 1969–1997.

Harford, J. (2005). What Drives Merger Waves? Journal of Financial Economcis 77,529–560.

Hayward, M. and D. Hambrick (1997). Explaining the Premiums Paid for Large Acqui-sitions: Evidence of CEO Hubris. Administrative Science Quarterly 42 (1), 103–129.

Husson, B. (1989). OPA: La Theorie a l’Epreuve des Faits. Revue Francaise de Ges-tion 25, 84–97.

Jennings, R. and M. Mazzeo (1991). Stock Price Movements around Acquisition An-nouncements and Management’s Response’. Journal of Business 64 (2), 139–163.

Jensen, M. (1993). The Modern Industrial Revolution, Exit, and the Failure of InternalControl Systems. The Journal of Finance 48 (23), 831–880.

Jensen, M. and W. Meckling (1976). Theory of the Firm: Managerial Behavior, AgencyCosts and Ownership Structure. Journal of Financial Economics 3, 305–360.

Jensen, M. and R. Ruback (1983). The Market for Corporate Control: The ScientificEvidence. Journal of Financial Economics 11, 5–50.

Kau, J., S. Linck, and P. Rubin (2008). Do Managers Listen to the Market? Journalof Corporate Finance 14 (4), 347–362.

23

Kroll, M., L. Toombs, and P. Wright (2000). Napoleon’s Tragic March Home fromMoscow: Lessons in Hubris. Academy Management Executive 14 (1), 117–128.

Langer, E. (1975). The Illusion of Control. Journal of Personality and Social Psychol-ogy 32 (2), 311–328.

Lauenstein, M. (1977). Preserving the Importance of the Board. Harvard BusinessReview 55 (4), 36–46.

Luo, Y. (2005). Do Insiders Learn from Outsiders? Evidence from Mergers and Ac-quisitions. The Journal of Finance 60 (4), 1951–1982.

Malmendier, U. and G. Tate (2005). CEO Overconfidence and Corporate Investment.The Journal of Finance 60 (6), 2661–2700.

McDonald, M., P. Khanna, and J. Westphal (2008). Getting Them to Think Outsidethe Circle: Corporate Governance, CEOs’ External, Advice Networks, and FirmPerformance. Academic of Management Journal 51 (3), 453–475.

McDonald, M. and J. Westphal (2003). Getting by with the Advice of Their Friends:CEO’s Advice Networks and Firm’s Strategic Responses to Poor Performance. Ad-ministrative Science Quaterly 48 (1), 1–32.

Mitchell, M. and H. Mulherin (1996). The Impact of Industry Shocks on Takeover andRestructuring Activity. Journal of Financial Economics 41, 193–229.

Morck, R., A. Shleifer, and R. Vishny (1990). Do Managerial Motives Drive BadAcquisitions? The Journal of Finance 45 (1), 31–48.

Mulherin, J. and A. Boone (2000). Comparing Acquisitions and Divestitures. Journalof Corporate Finance 6 (2), 117–139.

Myers, S. and N. Majluf (1984). Corporate Financing and Investment Decisions WhenFirms Have Information that Investors Do Not Have. Journal of Financial Eco-nomics 13, 187–221.

Nguyen-Dang, B. (2008). Does the Rolodex Matter? Corporate Elite’s Small Worldand the Effectiveness of Boards of Directors. Working Paper.

Oler, D., B. Olson, and J. Skousen (2007). Governance, CEO Power, and Acquisitions.Working Paper .

Patell, J. (1976). Corporate Forecasts of Earnings Per Share and Stock Price Behavior:Empirical Tests. Journal of Accounting Research 14, 246–276.

Rhoades, S. (1983). Power, Empire Building, and Mergers. Lexington, MA: D.C.Health & Co.

Rhodes-Kropf, M. and S. Viswanathan (2004). Market Valuation and Merger Waves.The Journal of Finance 59 (6), 2685–2718.

Roll, R. (1986). The Hubris Hypothesis of Corporate Takeovers. Journal of Busi-ness 59 (2), 197–216.

24

Rovenpor, J. (1993). The Relationship Between Four Personnal Characteristics ofChief Executive Officers and Company Merger and Acquisition Activity. Journal ofBusiness and Psychology 8 (1), 27–55.

Schneider, S. and R. Dunbar (1992). A Psychoanalytic Reading of Hostile TakeoverEvents. The Academy of Management Review 17 (3), 537–567.

Shleifer, A. and R. Vishny (1989). Management Entrenchment: The Case of Manager–Specific Investments. Journal of Financial Economics 25, 123–139.

Shleifer, A. and R. Vishny (2003). Stock Market Driven Acquisitions. Journal ofFinancial Economics 70, 295–311.

Walters, B., M. Kroll, and P. Wright (2007). CEO Tenure, Boards of Directors, andAcquisition Performance. Journal of Business Research 60, 331–338.

Wright, P., M. Kroll, and E. Elenkov (2002). Acquisition Returns, Increase in FirmSize and Chief Executive Officer Compensation: The Moderating Role of Monitoring.Academy of Management Journal 45 (3), 599–609.

Yermack, D. (1996). Higher Market Valuation of Companies with a Small Board ofDirectors. Journal of Financial Economics 40, 185–211.

25

Appen

dix

1:V

aria

ble

sD

escr

ipti

on

Type

Var

iable

sD

escr

ipti

onStr

uct

ura

lPow

erD

ualis

t(C

EO

Dua

lity)

(Dum

my)

1if

CE

Oho

lds

posi

tion

sas

both

CE

Oan

dbo

ard

chai

r,0

othe

rwis

e

Ow

ner

ship

Pow

erFo

unde

rs(F

amily

Fir

m)

(Dum

my)

1if

the

CE

Ois

the

foun

der

orth

ehe

irof

the

foun

der

fam

ily,0

othe

rwis

eC

EO

capi

talow

ned

(CEO

Owne

rshi

p)(%

)Per

cent

age

ofou

tsta

ndin

gfir

mca

pita

lhe

ldby

the

CE

O

Exper

tise

Pow

erTen

ure

(Ten

ure)

Num

ber

ofye

ars

the

indi

vidu

alha

sbe

enfir

mC

EO

Exp

erie

nce

(Acq

uisi

tion

Exp

erie

nce)

Num

ber

ofac

quis

itio

nspr

evio

usly

mad

eby

CE

O

Pre

stig

ePow

erE

duca

tion

back

grou

nd(E

lite

Edu

cation

)(D

umm

y)1

ifC

EO

isgr

adua

teof

EN

Aor

X,0

othe

rwis

eD

irec

tors

hips

(Out

side

Boa

rds)

Num

ber

ofC

EO

dire

ctor

ship

sM

arke

tR

eact

ion

Cum

ulat

ive

abno

rmal

retu

rns

(CA

Ri)

(%)

CA

Riis

the

cum

ulat

ive

abno

rmal

retu

rns

ofac

quir

ing

firm

arou

ndM

&A

anno

unce

men

tG

over

nan

ceC

orpo

rate

gove

rnan

cein

dex

(CG

Inde

x)G

over

nanc

esc

ore

deve

lope

dby

Bol

laer

tet

al.(2

009)

Boa

rdsi

ze(B

oard

Size

)N

umbe

rof

dire

ctor

son

the

boar

d

Fin

anci

alSta

tem

ents

Cap

ital

izat

ion

(Siz

e)(e

)N

umbe

rof

shar

esou

tsta

ndin

g,m

ulti

plie

dby

stoc

kpr

ices

Lev

erag

e(L

ever

age)

(%)

Deb

ton

Ass

ets

Ret

urn

ofeq

uity

(RO

E)(

%)

Res

ult

per

shar

edi

vide

dby

equi

tyD

eal

Met

hod

ofpa

ymen

t(C

ash)

(Dum

my)

1if

the

acqu

isit

ion

ispa

idfo

rby

cash

,0

othe

rwis

e

Oth

ers

Yea

r(D

umm

y)Y

ear

ofan

noun

cem

ent

Indu

stry

1if

the

stan

dard

indu

stri

alcl

assi

ficat

ion

(SIC

)co

de(2

digi

t)of

acqu

irin

gfir

mis

the

sam

eas

that

ofth

eac

quir

edfir

ms,

0ot

herw

ise

26