P RIVATE VS . P UBLIC A CQUISITIONS :

(1)

P

RIVATE VS

.

P

UBLIC

A

CQUISITIONS

:

T

HE DIFFERENCE IN SHAREHOLDER RETURNS OF BIDDING FIRMS BETWEEN PRIVATE AND PUBLIC ACQUISITIONS IN

E

UROPE

Author: B.G. Serto

University of Groningen

Faculty of Economics & Management and Organization Business Administration, MSc Finance

Supervisor: dr. ing. N. Brunia

(2)

PRIVATE VS.PUBLIC ACQUISITIONS

P

RIVATE

VS.

P

UBLIC TAKEOVERS

:

T

HE DIFFERENCE IN SHAREHOLDER RETURNS OF BIDDING FIRMS BETWEEN PRIVATE AND PUBLIC ACQUISITIONS IN

E

UROPE

Abstract

Although a large part of acquisitions is on private targets, the difference in shareholder returns of bidding firms between acquisitions of privately and publically held firms, or the ‘listing effect’, has received little attention in M&A research. This study examines the question: ‘Why is there a listing effect’. We test this question by means of event study methodology (market and risk adjusted model) and ordinary least squares regression. The sample consists of 1485 acquisitions from 2001 through 2007 in Europe, of which 1014 acquisitions involve private targets and 471 public targets. We show that the listing effect is present: on the three day interval surrounding the acquisition announcement, shareholders of the bidding firm earn higher returns for acquisition involving private targets than for acquisitions involving public targets. Ordinary least squares regression shows results are robust to inclusion of a variety of variables. We accept the hypothesis that bidding firm shareholders returns for private targets is higher for acquisitions paid for by means of shares than for acquisitions paid for by means of cash. This indicates that positive wealth effect of shareholders of bidding firms concerning private targets is related to monitoring activities by the target shareholders and reduced information asymmetries. The creation of a new monitor leads to higher returns to shareholders of the bidding firm for acquisitions involving private targets when payment is in shares relative to a payment in cash.

B.G. Serto

bartserto@gmail.com

Student number: 1323393

(3)

PRIVATE VS.PUBLIC ACQUISITIONS PREFACE

This thesis is the final work of my MSc Finance, at the University of Groningen. This thesis is the result of research that started in April 2007 in Berlin, on the Humboldt University. After collecting the data abroad, I finalized this thesis in Amsterdam. The aim of this thesis is to highlight the differences in shareholder returns of bidding firms when looking at public and private targets.

I would like to thank my supervisor, Ig. Dr. N. Brunia. He has been of great support to me. Furthermore I would like to thank mr. A. Hey for letting me use the sophisticated database systems of the Humboldt University in Berlin. Last but not least, I would like to thank my family and friends for their support.

Amsterdam, October 2007

(4)

TABLE OF CONTENTS

1 INTRODUCTION ... 6

2 LITERATURE REVIEW ... 9

2.2 Empirical evidence ... 9

2.2 Reasons for the listing effect ... 11

2.2.1 The listing effect ... 11

2.2.2 Difference in information availability ... 11

2.3 Acquisition characteristics influencing the listing effect ... 14

2.3.1 Method of payment ... 14

2.3.2 Other acquisition characteristics influencing the listing effect ... 15

3 DATA ... 18

3.1 Sample construction and data sources ... 18

3.2 Sample characteristics ... 19

4 METHODOLOGY ... 21

4.1 Event study methodology ... 21

4.2 Market and risk adjusted model ... 22

4.3 Tests ... 24

4.3.1 Test for normality ... 24

4.3.2 Parametric test ... 26

4.3.3 Non-Parametric test ... 27

4.4 Ordinary Least Squares regression ... 29

5 RESULTS AND ANALYSIS ... 30

5.1 Results and analysis of tests for the listing effect ... 30

5.2 Results and analysis on the method of payment ... 34

5.3 Results and analysis on other acquisition characteristics ... 37

6 CONCLUSION AND RECOMMENDATIONS ... 39

(5)

6.2 Limitations and Recommendations ... 40

REFERENCES ... 41

BOOKS ... 45

WEBSITES ... 45

APPENDIX ... 47

(6)

6 1 INTRODUCTION

Many companies use acquisitions in order to gain competitive advantages, creating efficiency gains, widening the growth potential and, in the end, increase the wealth of shareholders (Kohers [2004]). In Europe, during 2006 more than 12,000 acquisitions occurred, with a value of almost $1.5 trillion, compared to $956 billion in 2005 (source: Factset Mergerstat LLC). These figures point out the need to examine the effects of acquisitions for shareholders of bidding firms. Studies provided evidence that acquisitions involving public targets, in general, add value to the combined entity. However, the majority of the wealth gain in the acquisition is transferred to shareholders of the target firm, while shareholders of the bidding firm experience insignificant wealth changes or significant losses (Conrad and Niden [1992], Cheung and Shum [1993] and Goergen and Renneboog [2004]).

Most studies focus on the share performance of bidding firms focus on acquisitions involving public (listed) targets, however figure I shows that between 60 and 80 percent of acquisitions involve private (non-listed) targets.

Figure I

The number of acquisitions involving private and public targets in Europe during the period 2001-2007

source: Thomson Financials’ Securities Data Company Platinum database (SDCP)

Theories that are developed to explain the shareholder returns of the bidding firm in acquisitions involving public targets are not necessarily valid when looking at acquisitions involving private targets. The question then arises: why would the lion’s share of studies in the field of M&A focus on acquisitions involving public targets? The focus on public targets can be explained by the greater availability of data, and the bigger average and total deal size of acquisitions, and therefore prominence of public targets, as Figure II shows.

(7)

7 Figure II.

Average and total deal value of acquisitions involving private and public targets in Europe during the period 2001-2007

source: Thomson Financials’ Securities Data Company Platinum database (SDCP)

vidence on the effect of acquisition announcements involving private firms on shareholders returns

period o E

of the bidding firm is sparse. Available evidence for the US market is limited to Chang [1998], Ang and Kohers [2001], and Fuller, Netter and Stegemoeller [2002] and for the European market to Faccio, McConnell and Stolin [2006]. Chang [1998] shows, in line with others, shareholder returns of bidding firms are higher for acquisitions involving private targets than for acquisitions involving public targets. This phenomenon is since then also known as the ‘listing effect’. Factors that give rise to the listing effect are still largely unexplored. However differences between public and private targets in liquidity, bargaining power, information asymmetry and publicity show the need for further analysis.

This thesis examines the listing effect by examining abnormal returns in the announcement f an acquisition to acquirers of private and public targets in Europe over the interval 2001-2007. We examine the effects of an acquisition announcement on shareholder returns of bidding firms with the help of standard event study methodology (Mackinley [1997]) and regression analysis, similar to Chang [1998] and Faccio et al. [2006]. This thesis tries to explain the listing effect, and what

(8)

PRIVATE VS.PUBLIC ACQUISITIONS characteristics of the acquisition drive the listing effect. This leads to the following main research question:

‘Why are shareholder returns of bidding firms higher for acquisitions involving private targets, than for acquisitions involving public targets?’

This thesis differs from existing literature, because it examines a different geographical coverage and a different period of study. Chang [1998], Ang and Kohers [2001], Fuller et al. [2002] focus on the US market, whereas this thesis focuses on the European market, which is outperforming the US market in terms of the total value of acquisitions (source: Factset Mergerstat LLC). Chang [1998], Ang and Kohers [2001], Fuller et al. [2002] and Faccio et al.[2006] used data from the 1980’s through 2001, whereas this study focuses on the period 2001-2007. This has the advantage of the availability of a more complete dataset on acquisitions in the main data source of this thesis: Thomson Financials SDC Platinum. Furthermore, this thesis uses a more complete set of acquisition characteristics, which could explain the listing effect.

The remainder of this thesis is structured as follows. Section 2 gives a brief overview of existing literature and theories. Section 3 describes the data sample used in this study. Section 4 describes the methodology of the event study and regression analysis. Section 5 presents the test results and interpretation of these results. Section 6 concludes and gives recommendations for future research.

(9)

2 LITERATURE REVIEW

This section discusses the differences in shareholder returns of bidding firms between acquisitions involving private and public targets. Section 2.1 discusses empirical results of studies on acquisitions involving public and private targets. Section 2.2 discusses the theoretical background of the differences in shareholder returns of bidding firms between acquisitions of private and public targets. Section 2.3 discusses the method of payment and the effects on shareholder returns of bidding firms. Section 2.4 discusses other possible characteristics of the acquisition that can explain shareholder returns of bidding firms.

2.2 Empirical evidence

Almost all research following the seminar work of Jensen and Ruback [1983] shows that an announcement of a public acquisition generates positive returns to shareholders of the target firm, relative to the pre-announcement share price (Jarrell and Poulsen [1989], Servaes [1991]). There is little consensus about the wealth effect of the shareholders of the bidding firm: no or a slightly positive abnormal returns (Schwert [2000], Eckbo and Thorburn [2000]), or slightly negative abnormal returns (Walker [2000], Mitchel and Stefford [2000]) to shareholders of the bidding firm. There are several explanations why firms make acquisitions, even if abnormal returns of the bidding firm are not positive. Weston, Siu and Johnson [2001] explain that zero abnormal returns are consistent with a competitive corporate control market in which firms earn normal returns in their operations. Bruner [2001] calculated that “sixty to seventy percent of all M&A transactions are associated with financial

performance that at least compensates investors for their opportunity cost.” To test whether an

acquisition announcement leads to abnormal returns to shareholders of the bidding firm, we propose:

H1: “All else being equal, acquisitions lead to abnormal performance around the

acquisition announcement day”

Announcements of acquisitions involving private firms react in positive abnormal returns for the bidding firm. As can be seen in Table I, Chang [1998], Draper and Paudyal [2006], Faccio, McConnel and Stolin [2006] and Conn et al. [2005] find that private firms are purchased at a discount, relative to public targets. Therefore, in contrast to negative abnormal returns to shareholders of bidders for public targets, positive abnormal returns shareholders of bidders for private targets are found. Ang and Kohers [2001] find positive abnormal returns of bidders for private targets and public targets, but bidders for private firms outperformed bidders for public firms. Table I shows that common significant characteristics of an acquisition are the method of payment, the creation of a new block holder and the relative acquisition size. These acquisition characteristics are explained in section 2.3. Section 2.2 discusses the main reasons for the listing effect.

(10)

PRIVATE VS.PUBLIC ACQUISITIONS Table I

Overview of previous research

Table I summarizes five studies that examined the difference between private and public targets. The last column only shows characteristics that were of significance in explaining the abnormal returns. Chang [1998] shows only CARs depending on the method of payment. Results are therefore shown relative to the method of payment. * significance at 5 percent level

Author(s)

[year] Sample Period of study Event window Coverage CAR Significant characteristics

Chang [1998] 536

acquisitions 1981-1992 -1, 0 US method of payment, creation

of new block holder, relative size

CAR

Stock CAR Cash

Public 255 -0,4061* -0,02 Private 281 2,64* 0,09 Kohers [2004] 2886 acquisitions 1984-1997 -1, 0 US method of payment,

relative size, same industry Public 1811 0,532* Private 1075 1,297* Draper and Paudyal [2006] 8597

acquisitions 1981-2001 -1,+1 UK 0,6562 method of payment, relative size

Public 1098 -0,4061* Private 7499 0,8114* Faccio, McConnel, Stolin [2006] 4429

acquisitions 1996-2001 -2,+2 Europe method of payment, acquirer size, country

(11)

PRIVATE VS.PUBLIC ACQUISITIONS 2.2 Reasons for the listing effect

This section discusses the reasons for the listing effect. Section 2.2.1 describes the background of the listing effect and the prominent reason for the existence of the listing effect. Section 2.2.2 argues that the difference in information availability is the primary reason for the listing effect.

2.2.1 Background of the listing effect

Why would the shareholder returns of bidding firms acquiring private targets outperform the returns of bidding firms acquiring public targets? In acquisitions involving private targets, the bidding firm is able to capture more of the wealth gain that is created in the acquisition relative to the target firm, because the bidding firm can acquire private firms at a discount relative to the acquisition of public firms. Koeplin et al. [2000] compare book multiples and earnings multiples of private and public targets. Results show that private firms are purchased at a discount of about 20 percent compared to public firms. However, Capron and Shen [2007] report methodological concerns: ‘private firms have

no observable price to serve as an objective measure of market value from which to calculate a private firm discount’ (Capron and Shen [2007], pp. 3). Moreover, Koeplin et al. [2000] do not explain

the causes of the listing effect.

Chang [1998] and Fuller et al. [2002] discuss the lack of market liquidity as a reason for the existence of the listing effect. Liquidity refers to how quickly an asset can be converted to cash without the owner incurring substantial transaction costs or price concessions. Here, price concessions refer to the phenomenon whereby the sale of an asset takes place at a discount relative to the current market price of the asset (Bajaj, Denis, Ferris, Sarin [2001]). Fuller et al. [2002] explain that the existence of a market for shares of public firms provides a public seller with an alternative to cash-out its shares, rather than to sell them to an acquirer. On the other hand, the lower liquidity of shares in private targets limits the opportunity to cash-out its shares. Therefore the seller of a share in a private firm is in a weaker position to appropriate the wealth gain that is created in the acquisition, relative to the seller of a share in a public firm. However, Faccio et al. [2006] find no support for the lack of liquidity as reason for the existence of the listing effect. Faccio et al. [2006] conclude that the listing effect cannot be explained by just the liquidity effect. Therefore, in section 2.2.2, we try to explain the causes of the listing effect by the difference in the availability of information between private and public targets. Furthermore, in section 2.3.1, we argue that the creation of outside monitors in acquisitions involving private targets is a cause of the listing effect.

2.2.2 Difference in information availability

(12)

PRIVATE VS.PUBLIC ACQUISITIONS with potential bidders. The visibility of public firms to potential bidders is increased by regulatory disclosure requirements, analysts and press coverage, and the greater relationship with investment banks relative to private targets. This results in no uncertainty about the market value of public firms. The market serves as a source of information and as a device to value assets of public targets for potential bidders. In contrast, smaller private firms are, in general, more difficult to locate and to value as acquisition targets, due to less visibility and transparency to the bidding firms (Deeds et al. [1999]). Furthermore, private targets have greater difficulties to signal their value to bidders relative to public targets (McConnell and Pettit [1984]). Therefore bidding firms need to expend considerable resources in screening private targets and in assessing the quality of their assets, whereas bidders for public targets avoid redundant search efforts (Shen [2006]). We argue that these informational differences explain the different shareholder returns of bidding firms when acquiring a private target compared to acquiring a public target.

The greater information disclosure on public targets also implies that any potential bidder has identical knowledge about the target firm, i.e. bidding firms have no private information about the public target. Barney [1986] introduces the concept of strategic factor market, i.e. a market where the resources necessary to implement a strategy are acquired. If strategic factor markets are perfect, the costs of acquiring strategic resources will be similar to the economic value of those resources. Bidding firms with identical knowledge that are competing in factor markets will have equal expectations about the value of the resources. Therefore, bidding firms with identical knowledge acquiring public targets, face competition for the same resources. If we assume the opportunity set is continuous, in the end, acquired resources will not generate above normal economic performance for the bidding firm acquiring public targets. Barney [1988] concludes that it is only possible to earn positive abnormal returns when the strategic factor market is not perfectly competitive. This occurs, when there is differential knowledge about the target firm between potential bidders or when there is a unique fit between the potential bidder and the target firm. The differential knowledge about the target firm among bidders is bigger in acquisitions involving private targets relative to acquisitions involving public targets(Chang [1998]). Therefore we argue that shareholder returns of bidding firms are higher for acquisitions involving private targets, than for acquisitions involving public targets.

Bidders for private firms can extract a larger percentage of the wealth gain in the acquisition, because private targets face weaker bargaining power due to three private firm specific effects: private firm discount, lower bidder competition, and lower publicity on the acquisition of aprivate target firm.

(13)

PRIVATE VS.PUBLIC ACQUISITIONS i.e. not what it looks like. Furthermore, as explained in section 2.2.1, Fuller et al [2002] argue that private targets are discounted to reflect the lack of liquidity.

Lower competition among bidding firms in acquisitions involving private firms Bidding firms face less competition in the market for private firms relative to the market for public firms. We explained that due to the absence of market prices and less visibility and transparency to the bidding firms, buying a private firm creates frictions in the selecting process. The different way of negotiating does also affect the bargaining power of public and private targets. The sales of public targets are typically auction-like in nature. Milgrom [1987] shows that the more is known about the target firm, the more potential bidders are attracted. Therefore a public target in a relatively poor bargaining position vis-à-vis a potential acquirer can benefit from the auction-like process inherent to the share market. Kooli et al [2003] show that in contrast to public targets, private targets are generally sold through negotiations based on a voluntary exchange. Private targets lack the financial resources and relationships with investment banks to attract more bidders, in order to create a competitive auction.

Another difference between private and public targets is the motive for selling the firm. Private targets consider selling to a preferred buyer more important than creating a competitive auction process: often the decision to sell depends more on the employee welfare and cultural factors than the price (Gräbner and Eisenhardt [2004]). Managers of public targets are, due to corporate governance mechanisms, facilitating bidder competition, in order to find the best price for their shareholders. (Thomas and Thompson [2004])

Lower publicity on the acquisition of private target firms An acquisition announcement that reveals new information about the future perspective of a public target leads to a increase in the share price of a public target before the acquisition is completed (Schwert [1996]). This ‘runup’ is, according to Schwert [1996] an added costs to the bidding firm. Private information of the bidding firm on private targets is less likely to lead to this ‘runup’, because of the lower publicity on the acquisition of private firms. Even if the shareholders of the private target are aware of the private information of the bidder, shareholders have no possibilities to extract the value of it, unless they solicit rival bids.

In summary, private information is expected to be a greater source of value creation in acquisitions involving private targets than in acquisitions involving public targets. Bidders extract a larger percentage of the wealth gain in the acquisition involving private firms relative to the acquisition involving a public firm. Reason for this phenomenon is the lack of bargaining power, due to lower bidder competition, private firm discount, and a lower publicity on the acquisition process associated with private targets. Therefore we propose:

H2: “All else being equal, the return to shareholder of bidding firms for acquisition

involving private targets is higher than the return to shareholders of bidders for acquisition involving public targets”

(14)

PRIVATE VS.PUBLIC ACQUISITIONS 2.3 Acquisition characteristics influencing the listing effect

Section 2.1 discusses the empirical findings on the listing effect. We discuss the theoretical reasons for the differences in performance of bidders between private and public target firms in section 2.2. However, before we can conclude that a listing effect exists, we have to test whether the listing effect is robust for characteristics of the acquisition. This section distinguishes between the method of payment and the other characteristics, whereas the different abnormal returns of bidding firms between private and public targets depending on different methods of payment implies the existence of monitors. Furthermore it implies a smaller asymmetric information problem in acquisitions involving private targets relative to acquisitions involving public targets.

2.3.1 Method of payment

There is strong evidence that the method of payment in an acquisition has an impact on the share price of bidding firms (Goergen and Renneboog [2004]). Travlos [1987] shows negative abnormal returns of bidders financing an acquisition involving public targets with shares and no abnormal returns of bidders financing an acquisition involving public targets with cash. However Chang [1998] shows positive abnormal returns of bidding firms financing an acquisition involving private targets with shares and no abnormal returns of bidding firms financing an acquisition involving private targets with cash. The positive returns to shareholders of bidding firms involved in acquisitions of private targets is related to monitoring activities by the target shareholders and reduced information asymmetries.

Schleifer and Vishny [1986] and Ang et al. [2001] show that firm value can be increased by the creation of active monitors. Ang et al. [2001] show that monitoring is only cost effective if the monitor is a large investor, due to the high costs of monitoring. Privately owned firms are per definition owned by only a small group op shareholders, relative to public firms. Therefore if the acquisition is paid for by means of shares, shareholders of the target firm are new outside block holders in the bidding firm. As Schleifer and Vishny [1986] show, this leads to positive abnormal returns of the bidding firm to the acquisition announcement.

Myers and Majluf [1984] show that when bidding firms pay by means of shares for acquisitions involving public targets they suffer from the asymmetric information problem. Myers and Majluf [1984] show that issuing equity to the public leads to a negative abnormal returns when managers have superior information. The authors demonstrate that bidding firms only offer shares when they believe the share price is overvalued. Therefore, the market reaction to an acquisition announcement paid for by means of shares will be negative. However, when the bidding firm offer shares for an acquisition involving a private target with a small number of shareholders, the bidding firm can solve the asymmetric information problem by supplying the shareholders of the private target with private information of the bidding firm. Furthermore, the shareholders of the private target have an incentive to examine the prospects of the bidding firm carefully, because they will end up as block

(15)

PRIVATE VS.PUBLIC ACQUISITIONS holder in the bidding firm after the acquisition. Shareholders of a private target will only accept shares as a means of payment, when they judge the information as favorable. Therefore this leads to positive abnormal returns of the bidding firm to the acquisition announcement involving a private target.

In summary, positive returns to shareholders of bidding firms concerning private targets are related to monitoring activities by the target shareholders and reduced information asymmetries. The creation of a new monitor leads to higher abnormal returns of bidders for private targets when payment is in shares relative to a payment in cash. Furthermore the reduced information asymmetries lead to higher abnormal return of the bidding firm for a private target than for a public target when payment is in cash. Therefore we propose:

H3: “All else being equal, the return to shareholders of bidders for acquisition involving

private targets is higher than the return to shareholders of bidders for acquisition involving public targets when payment is in shares”

2.3.2 Other acquisition characteristics influencing the listing effect

Differences in abnormal returns of bidding firms in acquisitions of private versus public firms can also be attributed to other characteristics of the acquisition. This section discusses the most common characteristics used by Chang [1998], Koeplin [2000], Kohers [2004], Facchio et al [2006] and Capron and Shen [2007].

Relative deal size

Loughran and Vijh [1997] show that takeovers involving large public targets, relative to the market capitalization of the bidding firm, result in lower abnormal returns in the post-merger period. Using relative deal size as a proxy for expected efficiency gains, we can examine the different abnormal returns of bidding firms in acquisitions of private and public targets. If we assume that agency problems are smaller in acquisitions of private firms (Tirole [2006]), a positive relationship between the relative transaction size and the returns to shareholders of the bidding firm is expected.

Cross border

(16)

PRIVATE VS.PUBLIC ACQUISITIONS firms is positive. Faccio et al. [2006] argues that a necessary condition for the Hansen and Lott [1996] argument is that shareholders of the bidding firm are equal to the shareholders of the public target firm. Lewis [1999] however shows this condition does not hold in international acquisitions, because it is unlikely that shareholders of bidders acquirers and shareholders of targets acquirers are the same. Therefore, the abnormal returns of bidding firms for cross-border acquisitions of public targets should be equal to the abnormal returns of bidding firms for private targets.

Size of the target firm

Kohers [2004] uses the size of the target firm as a proxy for the health and stability of the firm: the author argues that some private firms may be so small and unstable that the are no longer able to operate independently. The decision to sell to a bidder, may be based on the lack of resources of the target firm to survive on its own. Kohers [2004] argues that such private targets would have weaker bargaining power and this results in lower premiums paid by the bidder. Therefore, abnormal returns for small private targets are expected to be higher than for bigger private targets.

Industry related acquisitions

Comment and Jarell [1995] show that acquisitions within the own industry (industry related) are associated with higher shareholder returns of the bidding firm than unrelated acquisitions. Compared to bidders that acquire targets in an unrelated industry, bidders that acquire targets in a related industry can more easily identify private targets because of the knowledge of the own industry. Furthermore, related industry bidders can rely on their knowledge of the industry to asses the value of the targets assets and growth prospects (Singh et al [1987]).

The state of the economy

Ang and Kohers [2001] use the state of the economy, since it affects acquisition activity. Ang and Kohers [2001] state that during an expansionary period, the bidder is better equipped to pay higher prices. Also, the bargaining position of the target may be stronger, as compared to contractionary periods. This implies a higher price paid in an acquisition during expansionary periods. The timing option of private firms, with per definition concentrated ownership supply private firms with strong bargaining power not just during these expansionary periods. An insignificant relationship between expansionary periods and high prices paid, and therefore lower abnormal return for the bidder, gives extra strength to the hypothesis that bargaining power can explain the differences between bidders return and private and public targets.

(17)

Tobin’s Q

Tobin’s Q is defined as the market value of the bidders assets divided by the bidders book value and is used as a measure of growth opportunities or a measure of performance. Servaes [1991] and Lang et al. [1989] found that bidders q has a positive relationship with bidder returns in public takeovers.

Country

Capron and Shen [2007] made a distinction between US and European acquisitions. We check for differences between bidder return of UK and other European countries. Reasons for this are that in the UK, relative to the rest of Europe, more information is available on public targets and more sophisticated capital markets exist. The smaller information discrepancy between public and private firms in the rest of Europe may reduce the difference in bidder firm return of acquisitions of private and public firms.

In summary, differences in bidder abnormal returns of acquisitions of private versus public firms can also be attributed to other characteristics than the method of payment. However, Chang [1998] and others find no characteristics of the acquisition that are able to explain the listing effect. Therefore we propose:

H4: “All else being equal, characteristics of the deal, acquirer or target cannot explain the

listing effect”

(18)

PRIVATE VS.PUBLIC ACQUISITIONS 3 DATA

This section discusses the data and methodology. Section 3.1 describes the way the sample is constructed. Section 3.2 discusses the characteristics of the sample.

3.1 Sample construction and data sources

From Thomson Financials SDC Platinum, from January 2002 through April 2007, a sample of firms that successfully acquired private and public firms is collected. Announcement dates and other deal characteristics are extracted from Thomson Financials SDC Platinum. Daily stock returns (RI) of bidding firms are extracted from Thomson Financials DataStream.

Table II shows the selection criteria of the events. Selection criteria are similar to Chang [1998] and Facchio et al. [2006]. We required complete control acquisitions. We define complete control acquisitions as acquisitions in which the bidder owned less than 10 percent of the target shares prior to the acquisition announcement and the bidder increased his ownership to at least 50 percent of the target shares. Kooli et al [2003] shows that in contrast to public targets, private targets are generally sold through negotiations based on a voluntary exchange. Therefore, in order to make public targets and private targets more comparable, only friendly bids are included. After controlling for the stated criteria the dataset contains 1485 deal announcements of acquisitions involving public and private targets.

Table II Selection criteria

Criteria Number of events left Difference

European bidders and targets 103776

Public acquirers 37904 65872

Private, public and subsidiaries as target 36609 1295

Deals bigger than $5 million 12858 23751

Completed deals 9770 3088

Complete control acquisitions 7274 2496

Friendly bids 7081 193

Deals with known acquirer market value in SDC platinum 2307 4774

Bidding firms are listed and stock return data are available in

Datastream announcement 1571 736

One acquisition in the event-window 1485 86

As market portfolio we use the FSTE Europe index, a broad-based stock index with daily returns of Europe. This broad-based stock index used is similar to Faccio et al. [2006].

(19)

PRIVATE VS.PUBLIC ACQUISITIONS 3.2 Sample characteristics

Table III shows the number of events per year. There is some variation in the number of acquisitions across the years, but even in 2002 and 2007, the smallest sample years, the sample numbers are respectively 92 and 86 announcements of acquisitions. As Table IV shows, UK acquirers are dominating the sample, with more than half of the acquirers coming from the UK. Other big acquiring countries are France, Italy and Germany.

Table III

Number of events per year

Acquisitions are listed by year of announcement according to SDC Platinum. Numbers reflect full calendar

years, except for 2007, where the sample runs to 1st_{of April.}

Target type 2002 2003 2004 2005 2006 2007 Total

Private 50 137 215 274 265 73 1014

Public 42 111 89 100 116 13 471

Total 92 248 304 374 381 86 1485

Table IV

Number of events per country

* remaining countries include Hungary, Iceland, Luxembourg, Czech Republik, Croatia, Bulgaria, Armenia, Malta, Latvia, Moldova, Monaco, Romania, Serbia Montenegro, Slovak Republik, Slovenia, Turkey, Ukraine, Yugoslavia

Acquirers Targets

Country No. of deals Percent No. of deals Percent

(20)

PRIVATE VS.PUBLIC ACQUISITIONS Table V shows characteristics of the events. Panel A shows the method of payment in respectively public and private target acquisitions. Panel A shows that a payment in shares in acquisitions involving public targets occurred more than twice as much as in acquisitions involving private targets . Furthermore, panel A shows slightly more acquisitions involving public targets are cross-border, as opposed to acquisitions involving private targets. Panel B shows that relative to the acquisition size, the market value of the bidding firm is bigger for acquisition involving public targets than for acquisitions involving private targets. Furthermore panel B shows that Tobin’s Q of acquirers is bigger for acquisition involving public targets than for acquisitions involving private targets. At last, panel B shows that acquirers of public targets tend to be insignificantly larger than acquirers of private targets, when looking at the acquirers market capitalization.

Table V

Characteristics of acquisitions involving public and private targets

An acquisition is all-cash if only cash is used as payment for the target. An acquisition is all-shares if only shares are used as payment for the target. Other offers are acquisitions paid for with cash, shares, future revenues, debt and others, or a combination of these. An acquisition is cross-border if the bidding firm and target firm are incorporated in different countries. An within-industry acquisition is defined as an acquisition where the bidder and the target firm have the same four-digit SIC code. The acquirers market capitalization is the market value of the acquirers common stock as of four weeks before the acquisitions announcement. Tobin’s Q ratio is the market value of the acquirer divided by the acquirers total assets. Acquirers relative deal size is the deal value divided by the bidders market value 4 weeks prior to the merger announcement. * shows significance on the 1 percent level.

Panel A. Binary variables

Target type acquisitions All-Share

(%) All-Cash acquisitions (%) Cross-border acquisitions (%) Within-Industry acquisitions (%) Public (n = 1014) 18,64 39,41 44,70 35,59 Private (n = 471) 7,78 42,27 36,26 35,07

Panel B. Other variables

Target type Acquirers relative _{deal size} Acquirers Tobin's Q _{capitalization ($ million)}Acquirers market

(21)

21 4 METHODOLOGY

This section discusses the methodology of this thesis. Section 4.1discusses the structure of an event study. The market and risk adjusted model is explained in section 4.2. Section 4.3 describes the test procedures. Section 4.4 describes the procedure and variables of an ordinary least squares regression.

4.1 Event study methodology

To measure the effects of an event on the value of firms, an event study can be used. Using the financial market data, an event study measures the impact of a specific event on the value of the firm. The specific event in this study is the announcement of an acquisition. Given the rationality in the marketplace, the effects of a takeover announcement will be immediately reflected in the share prices. Therefore, event studies measure the impact of the event on the share price observed over a relatively short period of time (Mackinley [1997])

The initial task of conducting an event study is defining the event of interest and selecting the event window, the period over which the share prices of the bidding firms will be examined. The event of interest in this thesis is an acquisition announcement, defined by SDC Platinum as: “the date when the bidder has made a formal offer to the target, or the companies involved in the deal confirm that the deal is to go ahead”. In order to correct for information leakage (James and Houston [2001]) we define the event window as the announcement date and enlarge this window to (-20,+20), narrowing the window to (0,+1). This is recommended by Mackinley [1997] and corresponds to Kohers [2004] and Faccio et al. [2006].

The estimation window is used to estimate the normal return (see ‘market and risk adjusted model in section 4.2). We use an estimation period of 120 days, prior to the event window. This estimation window follows recommendations by Mackinley [1997] and is in line with Chang [1998] and Kohers [2004]. Figure III shows the time window of the event study in this research graphically.

Figure III Time window

T 0 = -141 T 1 = -20 T 2= + 20

Estimation window Event window

(22)

22 The next step of examining the impact of an acquisition requires a measure of the abnormal return. The abnormal return is the actual ex-post return of the bidding firms share over the event window, minus the normal return of the bidding firms share over the event window. The normal return is defined as the expected return without the occurrence of the event. This results for firm i and event date t in:

1

period t. Returns are calculated using the following formula

Where , and are respectively the abnormal, actual and normal returns for the time

2 is equal to the total return of the share of firm I on time t. It and

.2 Market and risk adjusted model

y two models, the constant mean return model and the m

constant through time, whereas the market and risk adjusted model a linear relationship between the

3

price of firm I on day t

= the rate of return on a market portfolio of shares (FTSE Europe)

Where It -1 are equal to the total

return index of the share, on respectively t = 1 and t = -1.

4

The normal return ( ) can be specified b

market and risk adjusted odel. The constant model assumes that the mean return of the share is market return and the share return assumes. Mackinley [1997] shows that the market and risk adjusted model has a potential improvement over the mean return model, because the market and risk adjusted model removes the portion of the return that is related to the variation in the markets return. Therefore the market and risk adjusted model is increasingly able to detect event effects. We use the market and risk adjusted model in order to make the results in line with the literature (Chang [1998], Faccio et al. [2004], Kohers [2004]).

The market and risk adjusted model is specified by:

Where

(23)

23

= the error term, with E( ) = 0

ated values of true parameters through ordinary least squares (OLS),

ob e over an estimation period.

We use the values of the model parameters and to forecast the expected return E( ), after lo form

4

E α E _, 5

ormal returns of the individual shares can be aggregated by using for each event window. iven N events, the sample of aggregated abnormal return for period t is:

= the intercept term

= the systematic risk of stock i and are estim

tain d from the regression of on

running the market and risk adjusted model regression for each firm in the sample for the estimation period (-140,-20). We use the fol wing ula to examine the abnormal returns:

where

Average abnormal return and cumulative average abnormal return

The abn G

1

and its variance, assuming a large estimation period of 120 days (Mackinley [1997]) 6

1

7 he

T shows the average abnormal return on a specific day in the event window. The market reacted on the news of a acquisition announcement if the differs significantly from zero.

Cumulative average abnormal returns are defined by:

(24)

24 and, ming that event windows of the N shares do not overlap in order to set the covariance to zero, the variance as:

assu

, 9

In efficient markets, the will fluctuate randomly, except when information is received. The shows whether there is a reaction on the market on a specific day in the event window, whereas the

, shows a possible reaction on the market on an interval. The , are tested in different

4.3 Tests

time frames to gain an insight in the timing of the value creation, just like Kohers [2004] and Faccio et al. [2006].

To measure the effect of an event on the value of firms the observed must be statistically tested. Before we test whether the are statistically different from zero, we have to examine whether the results would be asymptotic and the presence of non-normally distributed variables can decrea

the abnormal returns are normally distributed. According to Brooks [2002] without assuming normality,

se the valid

udy methodologies: “Although daily excess returns are also highly non-normal, there is

vidence that the mean excess return in a cross-section of securities converges to normality as the number of samples securities increases. Standard parametric tests for significance of the mean excess specified. In samples of only 5 securities, and even when event days are clustered, the

ity of a parametric test, i.e. t-test.

However Brown and Warner [1985] show that non-normality of daily returns has no obvious impact on event st

e

return are well

test typically have the appropriate probability of Type 1 error” (Brown and Warner [1985], pp 25).

To test for robustness of the results and in line with Chang [1998] we conduct a non-parametric rank test for abnormal performance in event studies (Corrado [1989]).

4.3.1 Test for normality

The distribution of the average abnormal returns ( ) in the estimation window is important, because if the ‘s are not normally distributed, parametric tests’ validity can decrease (Brooks [2002]). The distribution of the data is tested according to the Jarque-Bera test.

The Jarque-Bera (JB) test is based on the kurtosis and skewness of the sample ‘s. The null hypothesis of this test assumes normality. JB is constructed as:

3

(25)

25

ber of observations, S is the skewness, which measures whether the ean value and K is the kurtosis, which measures the fatness of the

ions involving private depending on the method of payment. As can be seen, the null hypothesis cannot be rejected for all

hares, due to skewness.

Table VI

Test for normality of acquisitions involving private and public targets rgets correspond to the average abnormal return from bidders for pr

where N is equal to the num distribution is symmetric about its m

tails from the distribution, compared to a normal distribution. The outcome of the test is tested on a 5 percent significance level.

Table VI and Table VII show, respectively, the descriptive statistics of acquisit

and public targets and descriptive statistics of acquisitions involving private and public targets, datasets, except for the public target acquisitions paid for with s

Private ta ivate targets. Public targets

correspond to the average abnormal return from bidders for public targets

Private targets Public targets

Skewness -0,0006 0,1427 Kurtosis 2,6798 3,4815 Jarque-Bera Probability 0,5170 1,5793 0,7722 0,4540 Observations 121 121 Table VII

Test for normality of acquisitions involving private and public targets, depending on the method o

An acquisition if only shares

f payment

is all-cash if only cash is used as payment for the target. An acquisition is all-shares

are used as payment for the target. Other offers are acquisitions paid for with cash, shares, future revenues, debt and others, or a combination of these.

All-Cash All-Shares Other

(26)

26 4.3.2 Parametric test

A two sided t-test is used to test the null hypothesis t e abnormal returns and the average c mean abnormal re e zero. A two si is used, because th le consensus about the sign the abnormal returns have (see literature review).

Because Mackinley [1997] states that because the vari , or the sample measure (7) i e H0 used in event studies for the

hat the averag

umulative turns ar ded test ere is litt

ance of variance

s unknown, th can be tested using:

Θ ~ 0,1 11 A

nd for the _, expressed as:

Θ ,

,

~ 0,1 12

. A two-tailed t-test is applied under

If the null hypothesis is rejected, we conclude that the announcement of a takeover based on market erceptions, created or destroyed bidders firm shareholder value.

To determine if there is a significant difference between two cumulative average abnormal returns from respectively public or private targets, we use the following

z-1983):

with N-1 degrees of freedom, depending on the sample tested

critical values of respectively 1 and 5 percent. Based on the t-test a rejection or acceptation of the null hypothesis can be done.

p

resulting test as described by Dodd

and Warner (

13 Where

(27)

27 4.3.3 Non-Parametric test

Rank Test

For the t-test to be optimal the underlying distribution must be normal (Corrado [1989]). The non-parametric Corrado rank test offers improved specification under the null hypothesis and more power under the alternative hypothesis of the performance of abnormal returns. In contrast with the signed rank and the sign test (Brown and Warner [1980] and Brown and Warner [1985]), this test does not

requirement of symmetrical abnormal returns and is correctly

the cross-sectional distribution of the abnormal returns. Furthermore the specification of the date abnormal returns variance than the parametric tests (Corrado

89]) specified by:

1

have the specified, no matter how

skewed

rank test is less affected by event [19 . The Corrado rank test is

1 2 ₁₄ where 1 1 1 2 15

where L is the event window, N is the number of events, is the rank of the abnormal returns of event I for event window time t, where t = -20,+20.

The ranking procedure transforms a distribution of abnormal returns in an uniform distribution of rank value, that is independent of the asymmetry of the original distribution (Corrado [1989]). Therefore we can assume that the distribution is normal and we can test the null hypothesis that there are no

(28)

28

Generalized Sign Test

The generalized sign test examines whether the number of shares with positive cumulative abnormal

p n

returns in the event window exceeds the number expected in the absence of abnormal performance (Cowan [1992]). The expected number is based on the fraction of positive abnormal returns in the 120-day estimation period. The generalized sign test is specified by:

1 120 1 E S E 16 S 1 if AR 0 otherwise 17

he test statistic uses the normal approximation to the binomial distribution with parameter . w is of shares in the event window for which the cumulative abnormal return is

where

T

defined as the number

positive. The generalized sing test statistic is specified by:

Z w np

(29)

29 4.4 Ordinary Least Squares regression

characteristics are expected to influence the cross-sectional differences in the wealth effects of shareholders of bidding firms in acquisitions involving

Based on the discussion in section 2.3 and Chang [1998] and Faccio et al. [2006], the following deal private and public targets. Ordinary least quares regression analysis with dummy variables is used to show the importance of the characteristics. The regression is:

s

here

w

= the 3-day cumulative average abnormal return on days -1,0 and 1 for bidder EL = deal value divided by the bidders market value 4 weeks prior to the merger

announcement

BOR = 1 if the bidder and the target are incorporated in the same country; 0 otherwise IZ = target firm size

OC = 1 if the bidder and the target have the same 4 digit SIC code; 0 otherwise RA = 1 if the takeover occurred in a period of growth as measured by Eurostat; 0

otherwise

TOB = tobin’s q, measured by dividing the market value of the bidding firm by the book value of its assets.

RI = 1 if the takeover is for a private firm; 0 for takeovers for public firms HA = 1 if the offer is paid for by means of shares; 0 for cash or other offers means of cash; 0 for shares or other offers R S F E P S

CAS = 1 if the offer is paid for by

(30)

30 5 RESULTS AND ANALYSIS

This sec on trie o givti s t e an answer on the questions whether there is a listing effect and what reasons rent he ctions. Section 5.1 discusses the results of tests whether the f e listi effec Sectio ther characteristics of the acquisition can explain the

nc of the sting ed

c n be fo d. lts d an

able VIII shows the average abnormal returns (

drive the listing effect. This section shows results of the event study for different samples and diffe event windows, based on papers from Chang [1998], Fuller et al. [2002] and Facchio et al. [2006]. T results and analysis are divided in three se

listing effect exists. Section 5.2 discusses whether the method of payment can explain the existence o th ng t. n 5.3 discusses whether o

existe e li effect. In the appendix T-values, Corrado rank-values and generaliz values a un

5.1 Resu an alysis of tests for the listing effect

T ) for three different samples: the sample that

onsists of all acquisitions, the sample that consists of acquisitions involving public targets and the ample that consists of acquisitions involving private targets.

We discuss results of the sample of all 1485 acquisition announcements, on the twenty days urrounding the acquisition announcement. Table VIII shows positive returns (

c s

s ) to shareholders of

e bidding firm on the announcement day until two days after the announcement day. Similar to Faccio et al. [2006], the two-day cumulative return (

th

) surrounding the acquisition announcement hows a significant positive return. The non-parametric Corrado rank-test shows results that are onsistent with the results of the T-test (see appendix). We therefore accept our first hypothesis that

shareholders of bidding firms around the acquisition announcement day. Results are in line with findings of Schwert [2000], Eckbo and Thorburn [2000]

p’ costs for the bidding firm, as Schwert [1996] discusses. When we focus on the sample consisting of acquisitions two days after the announcement day. Results of the non-parametric Corrado rank-test are consistent s

c

acquisitions lead to abnormal performance of

and Moeller and Schlingemann [2004], who show slightly positive returns for shareholders of the bidding firms after the announcement of an acquisition.

To examine differences between acquisitions involving private and public targets we constructed two different samples. When we focus on the twenty days surrounding the acquisition announcement in the sample consisting of acquisitions involving public targets, Table VIII shows significant positive returns on the announcement day, the day after the announcement day, and a significant negative return on five days prior to the announcement day. This significant negative return could be due to ‘runu

twenty days surrounding the acquisition announcement in the

(31)

31 targets

s ne

(32)

32 Table VIII

he average abnormal returns in % of European Acquirers around the announcement day * and **, statistically significant at 1 percent and 5 percent level

nt day T

Eve All targets (n = 1485) Public targets (n =471) Private targets (n = 1014)

(33)

33 Table IX

The cu lative aver ormal returns in % of European Acquirers around the announcement day

* and **, statistically significant percent level lumn shows the difference the

mu age abn

at 1 percent and 5 . Last co between

of acquisitions involving public targets and the of acq involving private targets

All-ta

uisitions

rgets Public targets Private targets Private-Public

(34)

34 5.2 Results and analysis on the method of payment

Section 5.1 we show that the listing effect exists. Section 5.2 tries to explain this difference by ining the method of payment. Table X shows the average abnormal returns in percents of

ement day of acquisitions involving public and private target s for different methods of payment. When we focus on the acquisitions involving public targets, able X shows that payments by means of cash or by means of ‘other’ (future revenues, debt and thers, or a combination of shares and cash) result in significant positive returns to the shareholders of the bidding firm around the announcement day, whereas payments by means of shares result in negative returns. Therefore we argue that in acquisitions involving public targets, cash is referred above shares as method of payment due to the asymmetric information problem of Myers nd Majluf [1984]. When we focus on the private targets, Table X shows positive returns to the hareholders of the bidding firm, regardless of the method of payment. Therefore we argue that when the bidding firm offers shares for an acquisition involving a private target with a small number of hareholders, the bidding firm solves the asymmetric information problem by supplying the hareholders of the private target with private information of the bidding firm.

Table XI shows cumulative average abnormal returns in percent of European acquirers round the announcement day of acquisitions involving public and private target firms for different

ethods of payment. When we look at private targets in the two-day In

exam

European acquirers around the announc firm T o insignificant p a s s s a m surrounding the

nnouncement day, results show positive returns for acquisitions paid for by means of cash, shares and ther’. In line with Chang [1998] we find positive returns for acquisitions paid for by means of shares. In contrast with Chang [1998] we find positive average abnormal returns for acquisitions paid

r by means of cash. However we accept the hypothesis that returns of the bidding firm are higher for cquisitions paid for by means of shares than for acquisitions paid for by means of cash on the

two-sume the creation of a new monitor a

‘o fo a

day interval surrounding the announcement-day. Therefore we as

(35)

35 Table X

The average abnormal returns in % of European Acquirers around the announcement day of

e targ ion

acquisitions involving public and private target firms for different methods of payment * and **, statistically significant at 1 percent and 5 percent level. An acquisition is all-cash if only cash is used as

payment for the target. An acquisition is all-shares if only shares are used as payment for th et. Other offers

are acquisitions paid for with cash, shares, future revenues, debt and others, or a combinat .

Public targets Private targets

Event

day (n = 154) All-cash All-shares (n = 77) (n = 49) Other (n = 231) All-cash All-shares (n = 49) (n = 306) Other

(36)

36 Table XI

c ative ave abnorm eturns in % of European Acquirers around the ou ent day of acquisitions involving public and private target firms, for different

m s of paym

* and statistically signific 1 percent a ercent level cquisition is all-cash if only c sed as

paym for rget. An a on is all-sh only shares are used as pay or the target r offers

are ac sit id for with es, futu e debt thers, or a c ation.

The umul rage al r

ann ncem

ethod ent **,

ent the ta cquisitiant at nd 5 pares if . An a ment f ash is u. Othe

qui ions pa cash, shar re rev nues, and o ombin

Public targets E t W (1) sh 4) (2)All-shares (n = (3)Oth (n = 4 0 54 -1,8 3,587 ven indow All-Ca (n = 15 77) er 9) -20,2 0,90 729 -10,10 0,9886 -1 2,4919 74 -0,9 1,8024 2 43 0,1 1,8883 98 * 0, 2,2311 * 0 57 -0 2,3 1 38 * -0,0 1,8 ,395 -5,5 0,87 371 -2, 1,08 851 -1,1 1,00 0005 -1, 0,49 ,0696 26 * 0, 1,26 756 46 * Private targets

Ev t W (4)All-Cash ( 1) (5)All-shares (n (n = 30(6)Oth

(37)

37 5.3 Results and anal other acquisitio acteristics

This section tries to explain the listing effect by exa the characte tics of the acquisitions. Table XII shows results for the ordinary least squar against the cumulative average

abn over interval surro announcement-day different

samples. We use the two-day interval because it shows the highest significance. Table IX shows the correlation between the included variables in the regression.

The sam le that consists of ll acquisitions shows that the method of payment is a

variable. Fur ore the s pe (see row “private firm”) is significant positive. Therefore we conclude that shareholder returns of bi s are depending o hod of payment and the public status of the target. These results are in line with Faccio et al. [2006] and Chang [1998]. Other variables do not explain the shareholders returns of bidding firms. Furthermore, the explanatory power of our regression is low, but this is not unusual for cross-sectional regressions on acquirer

ysis on n char

mining ris

es regression

ormal returns the two-day unding the for three

In the Appendix,

p a significant

therm ample shows target ty

dding firm n the met

s , according to Chang [1998] and Trav 8 . We conclude that the listing effect is robust to t clusion of independent variables in this analysis. Th with findings by Faccio et al. [20

We s in the s at consists of acquisitions involving public targets, the only significant explaining vari method of payment (see n “All-shares”). The explanatory

ower of the least squares regressions for the sample consisting of acquisitions involving public and rivate targets is low (respectively 0,0088 and 0,0310), but in line with findings of Chang [1998]. The ample that consists of acquisitions involving private targets shows significant explaining variables are e method of payment and whether the bidding and target firm are active in the same industry. imilar to what Comment and Jarell [1995] show, industry related acquisitions lead to positive bnormal returns for bidding firm shareholders for private targets. Reasons for this phenomenon could e that compared to bidders that acquire targets in an unrelated industry, bidders that acquire targets in related industry can more easily identify private targets because of the knowledge of the own

y he industry to asses the

los [19 7]

he in of variety is is in line

06]. how ample th

able is the colum

p p s th S a b a

industr . Furthermore, related industry bidders can rely on their knowledge of t

(38)

38 Table XII

Regression on acquirers two-day surrounding the announcement-day

We use the two-day interval (0,+1) just as Chang [1998] and Kohers[2004]. See section 2.3 and 4.4 for explanation about the other variables used. * and **, statistically significant at 1 percent and 5 percent level.

All targets

(n = 866) Public targets (n = 280) Private targets (n = 586)

Variable Coefficient Coefficient Coefficient

All-Cash -0,0125 ** -0,0131 -0,0156 *

All-Shares -0,0036 -0,0225 ** 0,0219 **

Relative deal size 0,0007 0,0008 0,0009

Cross border -0,0058 0,0039 -0,0139

Target size 0 0 0

Industry related 0,0055 -0,0054 0,0111 **

High growth period -0,0003 -0,0078 0,003

(39)

39 6 CONCLUSION AND RECOMMENDATIONS

This section gives the main on the hypotheses, and thereby gives an answer to the main question. Section 6.2 gives limitations of th h and recommendations for further research.

ion

his thesis examines the difference in shareholders returns of bidding firms between acquisitions volving public and private targets, i.e. the listing effect. The listing effect is explained by means of vent study methodology. We use the market and risk adjusted model and ordinary least squares gression. The sample consists of 1485 acquisitions from 2001 through 2007 in Europe, of which 1014 acquisitions involve private targets and 471 public targets. The main research question is defined

the following way:

‘Why are shareholder returns of bidding firms higher for acquisitions involving private targets, than for acquisitions involving public targets?’

e show that the listing effect is present: shareholders of bidding firms earn higher returns for cquisitions involving private target firms than for public target firms. Ordinary least squares gression shows results are robust to inclusion of a variety of variables, including size of the acquirer,

ethod of payment, the acquirer Tobin’s Q, relative deal size, whether the acquisition is a cross are active in the same industry. The method of ment explains part of the listing effect. We accept the hypothesis that shareholders returns of

d for by means of shares than for acquisitions paid for by means of cash. This indicates that positive wealth effect of shareholders of

ing firm for acquisitions involving private targets when payment is in shares payment in cash. Results are in line with Chang [1998], Fuller et al [2002] and Capron and Shen [2007]. Furthermore, we conclude that the only significant explaining characteristics of the it type, method of payment, and industry relatedness. Other deal, acquirer or

conclusions research is researc 6.1 Conclus T in e re in W a re the m

border acquisition and whether the bidder and the target pay

bidding firms for private targets is higher for acquisitions pai

bidding firms concerning private targets is related to reduced information asymmetries and monitoring activities by the target shareholders. The creation of a new monitor leads to higher abnormal returns of shareholders of the bidd

relative to a

acquis ions are target

(40)

40 6.2 Limitations and Recommendations

The main limitation of this thesis is the limited amount of information available on private targets. atabases as SDC Platinum and Zephyr sometimes show false or no figures at all, which makes it impossi

the list search is to include more in-depth

haracteristics that could explain the listing effect. Furthermore, we only show results on the short D

ble to conduct a complete overview and analysis of the acquisition characteristics which drive ing effect. Therefore a recommendation for further re

c

(41)

41 REFERENCES

Arikan, I., 2005, “In the market for firms, how should a firm be sold?”, Advances in Mergers and

cquisitions, 4, pp 181–208

market for ‘lemons’: quality uncertainty and the market mechanism”, Journal of Economics, 84 (3), pp. 488–500.

ndrade, G, Mitchell M, Stafford E., 2001, “New evidence and perspectives on mergers”, Journal of

conomic Perspectives, 15(2), pp. 103–120.

ng, J., Kohers, N., 2001, “The take-over market for private companies: the US experience”,

ambridge Journal of Economics, 25, pp. 723-748

Asquith, P., Bruner, R.F., Mullins, D.W., 2003, “The Gains to Bidding Firms from Merger”, Journal

f Financial Economics, 11, pp. 121-139.

arney, J.B., 1986, “Strategic factor markets: expectations, luck and business strategy”, Management

cience, 32(10), pp. 1231–1241.

arney, J.B., 1988, “Returns to bidding firms in mergers and acquisitions: reconsidering the latedness hypothesis”, Strategic Management Journal, Summer Special Issue 9, pp. 71–78

Becchetti, L, Trovato, G., 2002, “The determinants of growth for small and medium sized firms: the

ole of the availability of external finance”, Small Business Economics 19(4), pp. 291–306.

1980, “Measuring security price performance”, Journal of Financial

conomics, September, 8(3), pp. 205-258.

returns: the case of event studies”, Journal of

inancial Economics, March, 14(1), pp. 3-31.

rvey of Evidence for the Decision-Maker”, Journal of

pplied Finance, 12, pp. 48–68. A

Akerlof, G.A., 1970, “The

Quarterly A E A C o B S B re r

Brown, S., and Warner, B.,

E

Brown, S., and Warner, B., 1985, “Using daily stock

F

Bruner, R.F., 2002, “Does M&A pay? A su