For this purpose, I divide the sample into a high, neutral, and low valuation periods, based on the P/E ratio of the market in the month the acquisition was announced

(1)

MASTER THESIS MSC BA:FINANCE

ACQUISITION TIMING AND SHORT TERM SUCCESS IN EUROPE The influence of market valuation

JEL Code: G34

University of Groningen

Faculty of Economics and Business

Groningen, August 18, 2009

Author: Nico Schaftenaar

Student Number: 1357646

First Supervisor: dr. W. Westerman Second Supervisor: dr. ing. N. Brunia

(2)

MASTER THESIS MSC BA:FINANCE

ACQUISITION TIMING AND SHORT TERM SUCCESS IN EUROPE The influence of market valuation

Abstract

In a sample of 316 acquisition announcements of countries that belong to the original twelve Euro zone countries, I study the influence of the market valuation on short term acquirer share returns. For this purpose, I divide the sample into a high, neutral, and low valuation periods, based on the P/E ratio of the market in the month the acquisition was announced. I find that acquisitions announced in high market valuation months yield a higher short term acquirer share return than those in neutral market valuation months. In low market valuation months the lowest share returns are yielded. This effect holds when including control variables and when dividing between cash and share paid offers, public and private targets and domestic versus cross-border deals. The results imply that acquisition timing matters. Acquisitions announced during high market valuations yield higher short term share returns to the acquirer.

(3)

TABLE OF CONTENTS

ACQUISITION TIMING AND SHORT TERM SUCCESS IN EUROPE 3

TABLE OF CONTENTS 1. INTRODUCTION ... 4

1.1 ACQUISITION PROFITABILITY ... 4

1.2 MERGER WAVES ... 5

1.3 ACQUISITION TIMING ... 6

1.4 RESEARCH QUESTION ... 7

1.5 STRUCTURE OF THIS THESIS ... 8

2. LITERATURE ... 9

2.1 MARKET VALUATION AND ACQUISITION VOLUME ... 9

2.2 MARKET VALUATION AND ACQUISITION PROFITABILITY ... 10

2.3 COMPANY- AND DEAL CHARACTERISTICS ... 11

2.3.1 Cash versus share paid deals ... 12

2.3.2 Public versus private targets ... 12

2.3.3 Domestic versus cross-border acquisition announcements ... 13

3. DATA AND METHODOLOGY ... 16

3.1 DATA COLLECTION PROCESS ... 16

3.2 MEASUREMENT OF VALUATION PERIOD ... 17

3.3 SAMPLE CHARACTERISTICS ... 18

3.4 CONTROL VARIABLES ... 21

3.5 METHODOLOGY ANNOUNCEMENT EFFECT ... 21

4. RESULTS ... 23

4.1 MARKET VALUATION ... 23

4.2 CASH VERSUS SHARE PAID ACQUISITION ANNOUNCEMENTS ... 24

4.3 PUBLIC VERSUS PRIVATE ACQUISITION ANNOUNCEMENTS ... 25

4.4 DOMESTIC VERSUS CROSS-BORDER ACQUISITION ANNOUNCEMENTS ... 25

4.5 OLS REGRESSION WITH CONTROL VARIABLES ... 25

4.6 SUMMARY ... 27

5. CONCLUSION AND RECOMMENDATIONS ... 28

5.1 CONCLUSIONS ... 28

5.2 RECOMMENDATIONS ... 29

REFERENCES ... 30

APPENDIX A: P/E RATIO PER PERIOD ... 31

(4)

INTRODUCTION

1. INTRODUCTION

Companies that want to grow and expand their business have only two alternatives.

They can either grow internally by developing new products or new markets, or they can acquire competing or complementing businesses. Due to lack of appropriate competences, knowledge or resources, lack of time or fierce competition, it may be hard for companies to expand business by growing internally. Therefore, engaging in acquiring other businesses is a very interesting and fast method for managers to fulfil their growth strategy. Therefore, it is conducted on a large scale around the world.

Of course, the success of all these acquisitions is not self-evident. Some acquisitions are very successful, while others fail. This makes it an interesting phenomenon to study for researchers in corporate finance. Not surprisingly, mergers and acquisitions (M&A) activity is one of the most thoroughly studied fields in corporate finance.

1.1 Acquisition profitability

Traditionally, the research in this field focused on the deal and company characteristics that predict the success of deals. Studies in this field can roughly be divided in long-term and short-term orientated studies.

Long term studies are characterized using by accounting measures of success. The long term measure fits the long term focus of companies well. After all, acquisitions are mostly done to realize long term profits. Downsides of this way of measuring acquisition success include the fact that accounting returns are in most cases only reported on a quarterly or yearly basis and the biased measurement of returns due to differences in accounting methods and window dressing. It is impossible to completely segregate the profits from a particular acquisition from other influences on the profits like other acquisitions and changes in the profitability of the original business of the acquiring company.

Short term orientated studies mainly depend on share market returns to measure acquisition success. Short term studies measure the reaction of the investors to the acquisition announcement and rely on the assumption that the market capitalization reflects the discounted value of the future profits of the company. Taking a narrow timeframe of mostly just a few days isolates the effects of the acquisition. This method however relies on the efficient market theory, which is often disputed but seems to hold on the long run and on

(5)

INTRODUCTION

average. Studies on the profitability of acquisitions indicated that, among others, acquirer size, means of payment and industry relatedness have predictive value for the success of an acquisition.

1.2 Merger waves

Besides the profitability of acquisitions, researchers are also much interested in patterns in the occurrence of acquisitions. This can be concluded by the massive body of research that is published on these patterns. During some time periods, much more acquisitions take place than during other time periods. Time periods when many acquisitions take place are commonly referred to by the term ‘merger waves’.

The existence of merger waves has been studied for at least a half century. Detailed historical data about US M&A’s from the end of the nineteenth century until now and from after the Second World War until now for Europe gives shows the existence of six merger waves from 1890 to present (Martynova and Renneboog (2008)). The start of these merger waves typically coincides with economic, political and regulatory changes (Harford (2005);

Martynova and Renneboog (2008)). Furthermore, the end is in most cases marked by a share market crisis. A recent example is the ending of the sixth merger wave, which coincides with the share market crash of 2008. Another characteristic of merger waves is that they coincide with economic cycles. Just like the economic cycle, merger waves show a great deal of similarity in their occurrence between the US and Europe (Martynova and Renneboog (2008)).

An explanation for the existence of merger waves is sought in two directions. The behavioural hypothesis is developed on the assumption that merger waves follow from managerial timing of market overvaluations of their firms (Harford (2005)). Shleifer and Vishny (2003) and Rhodes-Kropf and Viswanathan (2004) develop models based on this hypothesis. The neoclassical hypothesis on the other hand states that merger waves occur in response to specific industry shocks that require large scale reallocation of assets (Harford (2005)). These two hypotheses are of course not mutually exclusive.

(6)

INTRODUCTION

1.3 Acquisition timing

Literature about merger waves learns us a lot about the reasons why there are many acquisitions during one time period while there are little during another time period. It however also leads to new questions. The findings of the research on company characteristics and their relation to acquisition profitability, as described in section 1.1, leave enough options for acquisitions, also in worse economical times. It seems interesting to study if managers are right to go with the flow and to wait with their acquisitions and do it during the top of a merger wave when everyone does it. Looking for example at the economic situation we are in right now in the summer of 2009 we see that the overall sentiment is that it is not a good time for acquisitions. It can be observed that very little acquisitions take place right now, especially compared to two years ago when the market was booming. In spite of that, also in bearish markets, outside of merger waves, acquisition activity goes through. In January 2009, the American pharmaceutical company Pfizer announced plans to take over rival Wyeth for the sum of $ 67 billion and more close to home, the German energy company RWE has offered € 9.3 billion for the Dutch energy company Essent.

Why are there only some companies that still engage in acquisitions? With the low share valuations, it seems a good time to buy cheap. In spite of this observation, the number of acquisitions is low nowadays. The explanation of this phenomenon can maybe be found in more behavioural aspects of the field of finance. During hot market valuations, managers as well as investors are more optimistic about almost everything, including M&A. This leads managers to overestimate the potential of a possible acquisition, for instance by predicting high synergies and a flawless integration. The investor also reacts more positive in high valuation periods. The latter often results in positive abnormal returns on the announcement of the deal, while the former can result in disappointing long term results because the acquisition a company is involved in is not so rosily after all.

The issue of the influence of market valuation, and thus the timing of acquisitions, on individual acquisition success is only very new. Petmezas (2009) and Bouwman, Fuller and Nain (2009) address this issue in their recent articles. They use the monthly P/E ratio of the market to classify each acquisition announcement into high, neutral or low valuation period.

They find that during high market valuations, the reactions to the announcement in terms of

(7)

INTRODUCTION

share price movements are more positive than during low market valuations. This thesis is closely related to abovementioned literature.

1.4 Research question

This thesis will be grounded by the question how profitable the acquisitions are that still go through in the economic downturn between two merger waves, when acquisition volume is low. Are these acquisitions more profitable than deals made during high market valuations, maybe because of bargains that can be found in slow markets? Or is it the other way around? Are deals in higher market valuations more profitable? In times of high market valuations, markets are overall very positive and the many managers that do their deals during these times cannot all be wrong, can they?

In this thesis, I want to contribute to the knowledge in this field by researching relation between market valuation and acquirer return. I hope this research will lead to better insight about the underlying mechanisms. I will continue this paper guided by the following research question:

Does the overall market valuation at the time of an acquisition announcement influence the short term bidder returns? And, if so, how can this be explained?

This research question is interesting because there has not been very much research in this field. Other than the abovementioned articles by Petmezas (2009) and Bouwman et al.

(2009), I am not aware of publications that focus on this topic. Much research has been done covering the occurrence of clustering of acquisition activity in merger waves and the factors, like managerial behavioural finance, that explain this phenomenon. Besides that, the profitability of individual deals, based on deal- and company characteristics, has been researched thoroughly. However, the impact of the clustering of deals in merger waves – and the higher market valuations that go together with it – on the profitability of individual deals, have only recently received some attention. Petmezas (2009) and Bouwman et al. (2009) use data from the UK and the US respectively. I will use European data and will thereby add to the geographical scale of this field of research.

(8)

INTRODUCTION

1.5 Structure of this thesis

This thesis structures as follows: In the next section I will deal with the occurrence and relevant measurement of merger waves and market valuation, the main issues of this thesis.

Furthermore, I will examine the highlights of the vast amount of current literature on the determinants of deal success and the way this is measured. This will lead me to my hypotheses. In section three, the data and methodology of this thesis will be discussed, followed by the results of the tests in section four. This will lead to section five, in which I will present my conclusions and recommendations for further research.

(9)

LITERATURE

2. LITERATURE

In this section, I will discuss the relevant literature on market valuation and the profitability of individual acquisitions in and outside these merger waves. I will start by discussing literature on company and market valuation and then cover the company and deal characteristics that are important for this research. I will conclude this section by summarizing what literature has taught us on this topic thus far. This will lead to the hypotheses that will be tested and discussed in the remainder of this thesis.

2.1 Market valuation and acquisition volume

The occurrence of merger waves is a widely established fact in recent literature. Much research has been done in order to try to explain the underlying mechanisms that explain the occurrence of merger waves. Rhodes-Kropf and Viswanathan (2004) argue that the success of a share based acquisition can be explained by information asymmetries. They argue that target’s management can only fully assess the value of the own firm, while the acquirer has more information to estimate the results of the acquisition. When evaluating if a bid on their company is acceptable, they not only look at the stand alone value of their company, the also look at the synergies that the acquirer will gain. They will accept the offer if they feel they get rewarded a reasonable portion of the synergies the acquirer will benefit from. In times of market overvaluation, when the price per share is relatively high to the earnings per share, the acquisition bid seems higher to the target than during low valuation periods. So, targets are more likely to accept an offer during high valuation periods than during low valuation periods. Jovanovic and Rousseau (2002) develop a model that explains the investment rate of a company by its Q, which represents the ratio of the market value to the replacement costs of capital. They find evidence for their hypothesis that a high market value to replacement value leads to more investments and thus to more acquisitions. Because a high market valuation leads to a higher Q ratio, this is an explanation for the existence of merger waves. Under the behavioural hypothesis, managers are eager to use over valuated share to buy other companies (Harford (2005)). Sometimes this can have unwanted effects. Managers can be overconfident about the market, their own company and their own competences. Usually this happens in a high valuation period when the overall expectations about the market are very positive. This

(10)

LITERATURE

can lead managers to engage more in acquiring other firms. This managerial hubris can also explain why more acquisitions take place during high valuation periods.

2.2 Market valuation and acquisition profitability

The reasons for higher acquisition volume are quite convincing and give a credible explanation for the occurrence of merger waves. They obviously do not explain about the profitability of the deals that are done. However, in explaining the profitability of acquisitions, some of the underlying mechanisms are also interesting when evaluating the relation between merger waves and acquisition profitability.

The ongoing debate on the explanation of high merger activity during high valuation periods leads Bouwman et al. (2009) and Petmezas (2009) to study the profitability of these deals. This is something that is largely neglected in the literature on merger waves. In doing so, the abovementioned authors contribute to the ongoing research about the profitability of M&A deals.

Bouwman et al. (2009) make a distinction between high, neutral and low valuation periods on the basis of monthly P/E data on the S&P 500. Just as the P/E ratio of a company says something about the valuation of a company, the P/E ratio of the market tells about the valuation of the market. The market P/E is therefore taken as a proxy for the market valuation.

In their sample of 1,973 acquisitions announced between 1979 and 1998 in the US, they find that acquisition announcements are perceived much more positively in times of high valuations than in times of low valuations. For robustness reasons, they control for whether the acquisition is paid for in cash or in shares and they control for mergers versus tender offers. They also find that on the long run, acquisitions on low valuation periods earn positive abnormal returns and acquisitions made during low market valuation periods earn negative returns. The authors assume that the during high valuation periods – booming markets – managers are influenced by the overvaluation of the market and not only by the valuation of their own firm. The positive sentiment can lead to acquisitions that lack sufficient synergies:

managers suffer more from managerial hubris during high valuation periods.

Petmezas (2009) does a study similar to the abovementioned one of Bouwman et al.

(2009). He also classifies the relevant months into high, neutral and low valuation periods to determine market valuation. The sample consists of 2,937 successful domestic UK

(11)

LITERATURE

announcements between 1984 and 2003. He also uses more control variables than Bouwman et al. (2009), namely cash versus share paid deals, public versus private targets, industry relatedness, book to market, size, relative size, acquirer return, market return and former merger activity. Yet he draws the same conclusions. He also finds positive abnormal announcement returns on high valuated share markets that reverse to lower relative performance on the long run. Petmezas (2009) argues that it could be that during high valuation periods, managers feel the ‘urge to merge’ – as he calls it – and do not evaluate acquisitions as thoroughly as in low valuation periods, herby arguing in favour of the behavioural hypothesis. This leads me to formulate the following hypotheses:

Hypothesis 1a

Short term acquisition announcement returns will be higher for acquisitions announced in high valuation periods than in neutral valuation periods.

Hypothesis 1b

Short term acquisition announcement returns will be higher for acquisitions announced in neutral valuation periods than in low valuation periods.

2.3 Company- and deal characteristics

Much research has been published that examines the influence of all kinds of deal and company characteristics on the profitability of M&A deals. Studies have been done all over the world and through a long period of time. In this thesis, these characteristics are not the main focus. The main focus is the profitability of deals through periods of different market valuation. Some company characteristics and deal characteristics however can help explain the before mentioned. That is why these characteristics receive attention in this section. For a review of literature on the profitability of M&A deals see e.g. Bruner (2002), Tuch and O'Sullivan (2007), or Martynova and Renneboog (2008). Because it is not possible to examine all possible influences on deal profitability in this section, I have chosen to limit the discussion of company and deal characteristics to the three that are described below. These are the main characteristics to be taken into account when looking at profitability of deals in

(12)

LITERATURE

valuation periods because data on these variables is available and these variables have proven to lead to interesting results in earlier research.

2.3.1 Cash versus share paid deals

The method of payment has been found to have a significant influence in explaining bidding firms’ share returns. Travlos (1987) was the first to examine the effects of the method of payment on bidding firms’ share returns. He developed the hypothesis that the method of payment has a signalling effect to the market about overvaluation of the bidding firm, that is based on theory by Myers and Majluf (1984). When the firm pays with common share, this signals that the shares are overvalued. This signal about the overvaluation, that is given when a company bids on another one, results in the market not only pricing in the result of the (possible) acquisition, but also pricing in the overvaluation of the bidding firm. This results in negative bidding firm share price change on announcement (Travlos (1987)).

In addition, Rhodes-Kropf and Viswanathan (2004) find that the valuation period influences whether the deal is paid in cash or in shares. They find that share paid deals occur more in times of overvaluation and cash deals occur more when the market is under valuated.

The reason for this could be that managers are resilient to pay in shares in times of low valuations because they think their share is valuated too low, so they would make a bad deal.

Rhodes-Kropf and Viswanathan (2004) however find that the market reacts more positively on the news of a cash deal than news of an equity deal. The earlier research on the way the acquisition is announced to be paid for leads me to the following hypothesis.

Hypothesis 2

Short term acquisition announcement returns will be higher for cash-paid deals than for share-paid deals.

2.3.2 Public versus private targets

Conn, Cosh, Guest and Hughes (2005) studied the influence of the target being a public or private firm on the acquirer’s short term CAR. They find that bidder CARs are better for private targets than for public targets. For domestic public targets they find

(13)

LITERATURE

significantly negative bidder returns, whilst cross-border public targets result in zero announcement returns. In contrast, for both domestic and cross-border acquisition announcements of private companies, they find significantly positive announcement returns.

They argue that three factors explain the better results for private companies. The first one is that acquisitions of private targets can be negotiated behind closed doors. Bidders can end negotiations without the loss of face that can occur in a public bid. It is therefore less likely that poor acquisition outcomes occur that are induced by managerial hubris. Secondly, it is more likely that shareholders in a private firm play an active monitoring role in the acquisition. They can be more effective in this, because shareholders in private firms more commonly exist of large block holders. The third and last explanation is the fact that in private acquisitions it is less likely that competing bids occur. Moreover, speculation by shareholders – as frequently can be observed in acquisition announcements of public firms – usually not happens in acquisitions of private firms. The acquirer is therefore likely to receive positive announcement returns because of this discount on the price paid. Therefore, I formulated the following hypothesis.

Hypothesis 3

Short term acquisition announcement returns will be higher for private deals than for public deals.

2.3.3 Domestic versus cross-border acquisition announcements

Both Petmezas (2009) and Bouwman et al. (2009) use a sample that only contains domestic acquisition announcements. By using data from the European Union, it is clear that a number of the acquisition announcements that are used in this research, involve cross-border acquisitions, since the European Union member states still are separate countries. Moeller and Schlingemann (2005) compare the results of domestic and cross-border acquisitions for US acquirers. They find that acquisitions of cross-border targets lead to a significant 1% lower short term announcement return than domestic acquisitions do. They argue that the reason for this could be the relatively high costs that are involved with doing business abroad. They find that country specific factors can explain some of the differences. They for example state that

(14)

LITERATURE

the UK takeover market is more competitive than the takeover market in the US, so that it could be that relatively expensive acquisitions are done there, which has a negative impact on the announcement returns. Goergen and Renneboog (2004) find evidence that is significant on the 10% level, for European domestic acquisitions also resulting in a 1% higher return than cross-border acquisitions. This leads me to the following hypothesis.

Hypothesis 4

Short term acquisition announcement returns will be higher for domestic deals than for cross-border deals.

Concluding this section, I present a summary of the literature that I used to develop the abovementioned hypotheses in table 1. It shows the origin of the arguments I used and the data and conclusions of the papers that are used. This table can be found on the next page.

(15)

LITERATURE

Author (year) Variables Data Methodology Conclusions

Bouwman et al.

(2009)

Market valuation and acquirer return

2944 domestic acquisitions in the US, announced between 1979 and 2002

Three day acquirer

abnormal return

Acquirers buying during high market valuations have significantly higher announcement returns than those buying during low valuation markets Petmezas (2009) Market

valuation and acquirer return

2973 domestic acquisitions in the UK, announced between 1984 and 2003

Five day acquirer abnormal return

Managers that undertake mergers during bullish markets are rewarded by the generalized upward trend of the market on the short run

Travlos (1987) Cash versus share paid acquisitions

107 US acquisitions between 1972 and 1981

21, 31 and 153 day acquirer abnormal return

Cash paid announcements yield better acquirer returns than share paid announcements because of the signalling effect of the method of payment Conn et al.

(2005)

Public versus private targets

4344 UK acquisitions between 1984 and 1998

Three day acquirer

abnormal return

Bidder results for private companies are better than for public companies Goergen and

Renneboog (2004)

Domestic versus cross- border acquisitions

158 big European cross-border takeovers between 1993 and 2000

Two, five, 41 and 121 day acquirer

abnormal return

In continental Europe, cross-border acquisitions lead to higher returns while in the UK domestic acquisitions lead to higher returns

Moeller and Schlingemann (2005)

Domestic versus cross- border acquisitions

4430 acquisitions with a US acquirer, between 1985 and 1995

Three day acquirer

abnormal return

Cross-border acquisitions yield to significantly lower announcement returns to the acquirer

Table 1: Literature review summary

(16)

DATA AND METHODOLOGY

3. DATA AND METHODOLOGY

The very recent articles of Petmezas (2009) and Bouwman et al. (2009) – that are extensively discussed in the preceding section – have given us insight on the influence of market valuation on acquisition profitability. By means of this thesis, I intend to follow up on that research. Therefore, I will largely use the same methodology. The added value of this thesis lies in the data I use. Petmezas (2009) and Bouwman et al. (2009) use respectively UK and US data, whereas I will use European data. This broadens the geographical scope of the research. Although most principles normally hold al well for the US, the UK as for Europe, it cannot be ruled out that the findings of abovementioned papers do only hold for Anglo-Saxon countries. Furthermore, I use more recent data so that can be tested if the results also hold for more recent years

I will continue as follows. In the next section I will explain the methodology that will be used to test the above hypotheses. The section after that will hold the empirical results to these hypotheses, followed by the conclusions and finally the recommendations in the last section.

3.1 Data collection process

The data sample used in this paper is collected from two sources. The data on the deals is collected from the Zephyr M&A database by Bureau Van Dijk. This is an extensive database that holds information on worldwide M&A activity. This database also holds some deal- and company characteristics. Data on the share price quotes is gathered from Thomson Datastream. This database holds daily share quotes for all traded firms in the world. Other data on the acquiring companies is also gathered from this database. I examine 316 acquisition announcements by companies listed in the 12 countries that originally formed the Euro zone. Targets as well have to be located in one of these 12 countries. The countries that are in the sample and the number of observations per country can be found in table 2. The included countries are the most relevant economies in Europe. Furthermore, the fact that they all use the same currency, rules out the possibility of currency effects in the results.

As far as I know, no study like this one has been done in the past on European data.

This research will thus broaden the geographical scope of this field of research. Acquisition

(17)

announcements from July 1^st 2001 up to and including June 30^th 2009 are included in the data for this research.

In line with the objective of this paper, the acquisition announcements that have been included in this study have to comply with a number of requirements that are summed up below. Most of these requirements are also used in Petmezas (2009) and Bouwman et al.

(2009).

1. Both acquirer and target cannot be a financial firm. Financial firms are excluded from this research because has been found that these firms contaminate the results.

2. The value of the deal has to be at least 5% of the market value of the acquirer.

Announcements of very small acquisitions generally do not lead to observable share price reactions. The maximum relative value is 100%. Acquirers that take over firms that are bigger than themselves, distort the data.

3. The bid has to be on at least 50% of target’s shares.

4. The deal value is at least $ 1 million.¹

3.2 Measurement of valuation period

Each calendar month will be classified into a low, neutral or high valuation month on the basis of the monthly P/E ratio of the relevant share market, the Dow Jones Eurostoxx index. This index contains approximately 300 small, medium and large sized companies from the 12 original euro zone countries. It is designed to be a broad yet liquid representation of the euro zone market and is thus a relevant index to use. These data are also gathered from Thomson Datastream. Of the total sample of 96 months, the 25% (n=24) of the months that have the highest P/E ratio are qualified as a high valuation months, 50% (n=48) of the months are qualified as a neutral valuation period and 25% (n=24) of the months are qualified as a low valuation month. Figure 1 shows how the P/E ratio of the months evolved during the sample period and in which valuation period the month is. Appendix A shows a table that holds all months that are included, the P/E ratio of that month and whether it is a high, neutral or low valuation period.

1 Bouwman et al. (2009) use a minimum of $ 50 million which leads to a big decline in the number of observations. Therefore, I follow Petmezas (2009) in the minimum of € 1 million to keep as many as possible observations for analysis.

(18)

0 5 10 15 20 25 30

07- 2001

01- 2002

07- 2002

01- 2003

07- 2003

01- 2004

07- 2004

01- 2005

07- 2005

01- 2006

07- 2006

01- 2007

07- 2007

01- 2008

07- 2008

01- 2009

Market PE Ratio Lower border high valuation period Upper border low valuation period

Figure 1: Market P/E ratio per month

3.3 Sample characteristics

Before going to the results on the hypotheses that I formulated in the literature section, I will first discuss some general results and sample statistics. A total number of 316 acquisition announcements are included in this study. In total, deal value is just under a hundred billion Euros. Table 2 shows the distribution of the acquisition announcements for the country of origin of the acquirer. As can be observed below, the number of acquisitions per country is distributed as could be expected beforehand. Acquisitions made by companies in smaller countries like Austria, Greece and Ireland represent a smaller portion of the total number of acquisitions than the three largest countries, France, Germany and Italy do.

(19)

Acquirer country N %

Austria 10 3%

Belgium 12 4%

Finland 21 7%

France 78 25%

Germany 38 12%

Greece 19 6%

Ireland 6 2%

Italy 66 21%

Luxembourg 0 0%

The Netherlands 19 6%

Portugal 7 2%

Spain 40 13%

Total 316 100%

Table 2: Observations in sample per country

Table 3 shows the distribution of acquisition announcements for the most important variables that are studied in this thesis. When looking at de distribution of acquisition announcements between market valuation periods, we see that 27% of acquisition announcements are done during high valuation periods, while these periods only represent 25% of the months. This is consistent with the notion we made about the higher number of acquisitions that are made during high valuation periods. Furthermore, we notice that only 20% of acquisitions is announced in low valuation periods, where 25% of the months is classified as low valuation month. However, when we look at the deal value in the high-, neutral- and low valuation periods, we see that the neutral valuation period represents 53% of the total value, which is more than 50%, the number we would expect from 48 randomly selected months. It can be observed that while the number of observations for high valuation period may be high, they represent the lowest mean value, that is 198 million Euros, where the mean for the neutral valuation period is 353 million Euros and for the low valuation period it is 362 million Euros.

From the preliminary analysis of the sample characteristics, we can draw that the hypothesis about the clustering of acquisition activity in merger waves seems to hold for this sample.

More acquisition announcements take place during high valuation periods. We however see

(20)

that these acquisitions on average have a smaller deal size than the average deal size of the complete sample. In the results section we discuss the results on the other hypotheses.

Number of observations

% of

total Mean Median Total

% of Total

All acquisitions 316 100% 313 31 99,033 100%

High-market acquisitions 84 27% 198 45 16,602 17%

Neutral-market acquisitions 169 53% 353 29 59,651 60%

Low-market acquisitions 63 20% 362 21 22,780 23%

Cash acquisitions 140 44% 254 38 35,557 36%

Share acquisitions 63 20% 217 13 13,688 14%

Public target acquisitions 26 8% 305 123 7,934 8%

Private target acquisitions 290 92% 314 29 91,099 92%

Domestic acquisitions 230 73% 213 22 49,086 50%

Cross-border acquisitions 86 27% 581 90 49,947 50%

High-market Cash acquisitions 43 51% 172 83 7,395 45%

High-market Share acquisitions 22 26% 202 11 4,446 27%

High-market Public target acquisitions 6 7% 661 253 3,968 24%

High-market Private target acquisitions 78 93% 162 38 12,633 76%

High-market Domestic acquisitions 55 65% 153 30 8,409 51%

High-market Cross-border acquisitions 29 35% 283 92 8,193 49%

Neutral-market Cash acquisitions 84 50% 221 29 18,528 31%

Neutral-market Share acquisitions 29 17% 300 13 8,711 15%

Neutral-market Public target acquisitions 17 10% 227 132 3,857 6%

Neutral-market Private target acquisitions 152 90% 367 28 55,794 94%

Neutral-market Domestic acquisitions 120 71% 239 23 28,715 48%

Neutral-market Cross-border acquisitions 49 29% 361 88 30,936 52%

Low-market Cash acquisitions 13 21% 741 11 9,634 42%

Low-market Share acquisitions 12 19% 44 17 530 2%

Low-market Public target acquisitions 3 5% 36 41 108 0%

Low-market Private target acquisitions 60 95% 378 21 22,672 100%

Low-market Domestic acquisitions 55 87% 218 19 11,963 53%

Low-market Cross-border acquisitions 8 13% 1,352 129 10,818 47%

Transaction value (In millions of Euros)

Table 3: Summary statistics of observations in sample per tested variable

(21)

3.4 Control variables

Several company- and deal characteristics have already found to be explanatory for acquisition success in historic research determining the success of an announcement. I largely follow Petmezas (2009) and Bouwman et al. (2009) in the choice of control variables. The chosen control variables are also generally used in other studies. The control variables that I used in the tests are to ensure the robustness of the results. All control variables have been chosen because of their proven association with returns to shareholders. Below I sum up the variables I use and how they are measured.

Acquirer size is the market capitalization of the acquirer measured one month before the announcement date. It is the function of the number of outstanding shares and the unadjusted share price. Relative size is the total deal value divided by the acquirer size that is measured one month before the announcement date. For the Acquirer price to earnings (P/E) ratio, the price per share is divided by the last known number of earnings per share. Thomson Datastream reports this ratio on a monthly basis. I use the ratio from the preceding month.

Acquirer price to book (P/B) ratio is the last known net book value divided by the market value of the acquirer one month before announcement. Thomson Datastream reports this ratio on a monthly basis. I use the ratio from the preceding month. Acquirer total assets represents the last reported number of total assets of the acquirer. Acquirer net profit is the last reported net profit of the acquirer.

3.5 Methodology announcement effect

The announcement effect will be measured by doing an event study using market returns. The use of market returns makes it a very relevant measure as these returns are also observed and interpreted by investors. The share quote of a company is believed to reflect the discounted value of the future cash flows of the company; this makes it a relevant, forward looking measure of firm performance. This method relies on the Efficient Market Hypothesis (EMH). According to this hypothesis, all available information is incorporated into the share price and the market is efficient. This means that it is not possible to gain from arbitrage opportunities. When there is new information that becomes public, in this case that would be the acquisition announcement, investors react on that information. Some may overreact, others may under react, but on average, the market is right (Fama (1998)). In this way, the

(22)

share price movement around the acquisition reflects the opinion of investors regarding the announcement.

Following Bouwman et al. (2009) and Petmezas (2009), I will use the CARs around the announcement of the bid. This will be done by calculating the returns of the acquirer’s share in the relevant period and subtract the returns on the Dow Jones Eurostoxx index in the same period. This index is also used in the above-mentioned calculation of the valuation period.

The CAR is calculated as follows:

 ¹^, ¹ ₁¹⁽Rit Rmt⁾

CAR   ^  (1)

1,1

CAR is the cumulated abnormal return to the acquirer measured from one day before to one day after the acquisition announcement, Rit is the firm i’s daily share return on date t and Rmt is the return for the Eurostoxx index on date t.

Like most authors, I use a relatively small time frame to measure the announcement.

This has the following reason: I only want to capture the effect of the acquisition announcement. Since shareholders react on other news and developments about the company like sales figures and forecasts, as well as macro economic developments, we need a small time frame to isolate the reaction on the news of the acquisition announcement only. Studies with a longer time frame have the possibility that the share prices are impacted by other events which leads to inconsistent results.

(23)

RESULTS

4. RESULTS

In this section, I will describe the results from the empirical tests that I have done. I will describe the results per hypothesis in the same order as I have formulated them in the literature section. All results that are discussed are statistically significant unless otherwise mentioned.

4.1 Market valuation

For market valuation, I hypothesized that short term acquisition announcement returns will be higher for acquisitions announced in high valuation periods than in neutral valuation periods (hypothesis 1a) and that short term acquisition announcement returns will be higher for acquisitions announced in neutral valuation periods than in low valuation periods (hypothesis 1b). As shown in Table 4, acquisition announcements during high valuation periods show a positive three day CAR of 2.9%. Acquisition announcements during neutral- and low valuation periods show three day CARs of 2.0% and 1.7%. These results are in favour of both above hypotheses on market valuation and three day CARs. Acquisitions announced in high valuation periods yield higher returns than acquisitions in neutral valuation periods and acquisitions announced in neutral valuation periods yield higher returns than acquisitions in low valuation periods. Also when looking at subsets of the sample, the results are in favour of both above hypotheses, but they do not in all cases (fully) support the hypothesis. The CARs for cash paid acquisitions are higher in high valuation periods (2.9%) than in neutral valuation periods (2.0%). The number for low valuation period cash paid acquisitions is not statistically significant. Domestic acquisition announcements yield to higher CARs in high valuation periods (3.4%) compared to neutral valuation periods (2.2%).

Low valuation period domestic acquisition announcements however also yield a CAR of 2.2%, so domestic acquisition announcements in this sample do support the first part of the hypothesis, but not the second part. The other subsets of the sample largely follow the same pattern as the domestic and cash paid acquisitions, but not all results are statistically significant. The results can be observed in table 4.