University of Amsterdam
Economie & Bedrijfskunde
Financiering & Organisatie
Bachelor Thesis
“The Short-‐Term Effect of Acquisition Announcements on
Shareholder Value for Acquiring and Target Firms in the US Banking
Sector”
Author:
Thomas Terstegen
10098321
Supervisor:
Razvan Vlahu
January 2016
Abstract
In this paper the short-‐term effect of acquisitions on shareholder value is
examined. The sample exists of 206 deals between US banks in a 10-‐year period from 2005 to 2014. An event study is applied to calculate the cumulative average abnormal returns (CAAR) from 5 days before until 5 days after the acquisition’s announcement day. This is done for the acquiring and target firms separately. The CAARs of both the acquirer and the target are proven significantly positive, with values of 1.06% and 25.64% respectively. Also the effect of the financial crisis of 2008 is measured through multiple regressions with the firms’
cumulative abnormal return (CAR) as dependent variable and a dummy variable, indicating whether the acquisition was before or after the crisis, as independent variable. From this follows that short-‐term shareholder value creation from acquisitions is significantly bigger after the crisis than before the crisis, which has not been found in earlier research. In future studies the reasons behind this must be investigated.
Statement of Originality
This document is written by Student Thomas Terstegen who declares to take full responsibility for the contents of this document. I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it. The
Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
Table of Contents
1. Introduction………4
2. Literature Review………..……….………..7
2.1. Acquirer Effect……….………7
2.2. Target Effect………..………8
2.3. Abnormal Return Explaining Variables……….…...9
3. Data………..…10
3.1. M&A Selection Criteria………..………...10
3.2. M&A Data Characteristics………..………11
3.3. Firms’ Stock Return Data……….………..11
4. Methodology and Hypotheses…...………...12
4.1. Event Study……….………...……….…………...12
4.1.1. Estimation Window………...………13
4.1.2. Event Window………..14
4.1.3. Event Study Hypotheses……….……14
4.2. Regression Analysis and Hypotheses...………..15
4.2.1. Crisis Variable……….…..15
4.2.2. Size Variable……….……….16
4.2.3. Listing Variable………16
5. Results………...………17
5.1. Event Study Results………...………17
5.1.1. CAAR Acquirer Firms………..……….17
5.1.2. CAAR Target Firms………...……….18
5.2. Regression Results……….19 5.2.1. Acquirer Results………...………...19 5.2.2. Target Results………...21 6. Conclusion……….………23 References………..……….25 Appendix………...……27
1. Introduction
Since the early 1990s the amount and volume of merger and acquisition (M&A) activities worldwide has grown substantially. Apart from the temporary drop caused by the early 2000s recession, by 2008 the annual number of M&As had quadrupled. The value of M&A transactions even increased tenfold in these 17 years. Then in 2008 the growth stagnated due to the financial crisis. The
worldwide number of M&As dropped with more than 10%, although being stable since then. The total value of all annual M&As even decreased from 5.000 in 2007 to 2.200 billion dollars in 2009. From 2014 this value is back at 4.000 per year and in spite of the major fluctuations it is certain that M&As have become of great importance in today’s economy (numbers from IMAA).
An ongoing question in M&As has been the motivation of management to take part in it. The theory of the principal-‐agent problem teaches us that the agent (management) may act in its own interest, rather than in that of the principal (shareholders). The decision of management to engage in M&As can also be subject to this problem. The shareholders appoint management in order to maximize shareholder value, while that is not the main concern of management itself. The shareholders want management to engage in M&As only if it will improve efficiency and thereby enhance stockholder wealth. However,
management can be willing to engage in M&As with numerous other incentives, like maximizing own remuneration (bonuses for increased size), reducing competition and thereby effort (building an empire via M&As) or in financial industries growing-‐by-‐acquisition (in order to become too big to fail, possibly resulting in governmental guarantees). These reasons are neither in the interest of the shareholders nor in the welfare of the company and thus should M&As not be based upon them.
DeYoung et al. (2009) confirm suspicions above by stating that the effect of M&As not always seems to be significant in terms of shareholder value
maximization. A common way to observe this is measuring the abnormal stock returns that come with the announcement of M&As. Their post-‐2000 research on North-‐American M&As finds that M&A activities can be efficiency improving, but
they find no significant stockholder value enhancement. However, studies from outside North America sometimes do find a stockholder value enhancement effect of M&A.
Concentrating on geographic regions, Europe and the US share the same
development in M&A activities. However, the number of M&As in Europe slightly outnumbers that of the US, while for the total value of M&A transactions it is just the opposite. When M&A developments are compared among different
industries, the banking sector remarkably stands out in one aspect. The reactions in both amount and volume of M&As on the recession of the early 2000s and the financial crisis of 2008 are considerably more severe in banking than in the other big industries. The amount of M&As in banking industries worldwide decreased by 30% and 40% respectively due to the above mentioned events, while the value of the M&As even decreased with over 80% and 75% respectively (numbers from IMAA).
This thesis will focus on the US banking sector. Very recent studies in the
European banking sector have shown M&As can result in significant shareholder value enhancement. However, no research yet has found similar consistent results in the US. DeYoung et al. (2009) argue that this could well be because European banks involved in M&A deals learned best-‐practices (and worst-‐ practices) from observing the earlier US deals. In addition they find that the reason for this can partly be assigned to the period being studied. The researches taken into account in their literature-‐reviewing article are all from before the financial crisis of 2008 though. The industry has changed substantially since then and considering the above argument of DeYoung et al. (2009) new research might well result into different findings. The first question to be answered in this thesis is the following: What is the short-‐term effect of M&As on the shareholder value of both the acquirer and the target? A supplementary regression will be run in order to answer the second question: To what extent does the short-‐term effect of M&As depend on the time period in which the M&As is carried out? Here the aim is to examine the effect of the financial crisis of 2008 that took place midway the studied time interval (2005-‐2014).
This thesis will proceed as follows. In section 2 existing literature is reviewed. In section 3 the data is described, whereas in section 4 the methodology is
explained and the hypotheses are stated. In section 5 the results are presented and finally section 6 concludes.
2. Literature Review
Prior to the actual research the existing literature is reviewed. Extensive research has been done on the effect of M&As on the shareholder value of the participating firms. However, previous studies do not agree with each other consistently. The majority of the researches that try to measure the effect of M&As are event-‐studies in which abnormal stock returns by means of M&A announcements are measured for the shareholders of both the acquirer and the target firms separately. The event-‐studies reviewed in thesis are from the US, Europe and Asia. Also they cover different time periods, ranging from 1979 until 2009. In this section a representative overview is given of the existing literature, distinguishing acquirer and target effects beforehand.
2.1. Acquirer Effect
Considering M&A banking literature from the early 1990s a certain pattern can be observed throughout time. The oldest researches studied in this thesis on the M&A effect on the value of the US acquiring banks’ stock do not seem to result in significant fluctuations of the stock at all. In addition, the insignificant observed effect even tends to be negative more often than positive (Houston and Ryngaert, 1994; Hudgins and Seifert, 1996; Pilloff, 1996). In his study of US bank mergers from 1980 until 1997 Becher (2000) also does not find any significant results. On their turn Houston at al. (2001) failed to find conclusive evidence that mergers create value for large bank deals between 1985 and 1996. However, they observe that mergers after 1990 were more likely to be accompanied by higher abnormal returns than previously.
Then as worldwide M&A activities increased towards the 2000s, the effect of M&As on shareholder value slightly changes. In a sample of 3135 M&As, Fuller et al. (2001) find a significant positive effect for the acquirer. A difference is that this research from 1990 until 2000 is focused on US banking firms that make multiple acquisitions. Moeller et al. (2003) also find a positive effect for the acquirer in their research on 12,023 M&As in the US between 1980 and 2001.
When the horizon is broadened outside the US to Europe and Asia comparable findings are observed. In West-‐Europe, however, the effect of M&As on the stocks of 4429 acquirers seem to depend on whether the target is listed or not. Acquirers of unlisted targets experience significant positive effects, while acquirers of listed targets encounter insignificant negative effects (Faccio et al., 2006). In Asia the recession of the early 2000s did not obstruct the growth of M&A activity (IMAA). Wong (2009) investigates the effect of 658 M&As in China, Japan, Taiwan, Korea, Singapore and Hong Kong that took place from 2000 until 2007, finding significant positive effects.
The slight pattern that can be observed as time progresses is that M&As increasingly create value for the acquiring firm’s shareholders. However, DeYoung et al. (2009) argue that the effect strongly depends on the time period being studied.
2.2. Target Effect
Most existing literature regarding the effect of M&As studies the effect on the acquiring and the target firm simultaneously. Therefore mostly the same studies are referred to in this subsection as in the previous one. The effect on targets logically tends to be more positive due to premiums being paid by acquirers or efficiency improvements as a result of M&As.
From 1979 until 1985, well before the rise of the M&A activities, Neely (1987) finds positive effects for US target banks. Almost a decade later Hudgins and Seifert (1996) show the exact same results. After observing M&As from 1980 until 1997 Becher (2000) also concludes that target shareholders earn significant positive value. Fuller et al. (2001) focuses on frequently acquiring firms and measures the effect on the target firms. Taking into account 2135 M&As the authors distinguish public from private targets. The result is that the M&A effect on both type of targets is positive and public targets seem to benefit more. Scholtens and de Wit (2004) expand the area of research to Europe and the US. They observe 78 bank mergers from 1990 until 2000 and the found effect is also positive. The only research having findings contrary to all aforementioned
results is from Wong (2009). From 2000 until 2007 he finds a negative effect for Asian target firms.
Summarizing, existing literature shows that throughout time and certainly in the US target firms participating in M&As seem to earn positive value.
2.3. Abnormal Return Explaining Variables
In explaining the acquirer and target effects, there are three variables that have to be taken into account. These variables are featured in this subsection and will be the independent variables in the regression.
The most important variable to be observed is the time period in which the M&A takes place. DeYoung et al. (2009) argue that results of multiple studies for a large extent can be assigned to this variable. With the financial crisis of 2008 positioned in the timeframe of the research, it will be investigated whether effects are different before and after it. The second variable is the size of the acquiring firm. Large acquiring firms tend to overpay the target, which is expressed in a significant higher positive effect for the small acquirer than for the big ones, which are defined as those with a market capitalization above the 25th percentile of NYSE firms in the year in which the acquisition is announced (Moeller et al., 2003). A variable that also seems to be significant for the acquirer is whether the target is listed or not. Faccio et al. (2006) find that the acquirer of an unlisted target benefits more. This so called listing effect persist in time, across countries and independent of size or method of payment.
3. Data
In this section the composition of the dataset is described. The M&A events studied in this thesis must satisfy a number of criteria, which will be listed in part 3.1. The characteristics of the selected M&As will then be exhibited in 3.2. And in the last part is described from where the M&A participating companies’ stock return data is collected.
3.1. M&A Selection Criteria
At the University of Amsterdam, the library gives access to multiple online databases, one of them being Zephyr. This database contains comprehensive M&A data with integrated detailed company information. It also allows applying a broad range of specifications in search of M&As. To compose a useable dataset a number of these specifications were selected in the Zephyr search engine, leaving a sample in which the merger or acquisition has to satisfy the following criteria:
a) The transaction is completed.
b) The transaction (completion) was made between 01-‐01-‐2005 and 31-‐12-‐ 2014.
c) The acquirer and the target are from the United States of America. d) The acquirer and the target are from the banking sector.
e) The acquirer must be a listed at the NYSE or the NASDAQ. f) The target must be either unlisted or delisted.
This search yielded 298 results, of which 293 are acquisitions and only 5 are mergers. In order for results to become more specific and to avoid possible biases it is decided that mergers will be excluded from the sample. Also deals with an unknown deal value or with a deal value less than 10 million dollars will be excluded from the sample, due to the risk of possible biases. The conclusive criteria are the following:
g) The type of deal is an acquisition.
h) The minimum deal value is 10 million dollars.
3.2. M&A Data Characteristics
Table 1 shows the characteristics of the 206 M&As that are studied in thesis.
Table 1: Descriptive Statistics of the Sample of Acquisitions. All numbers regarding deal value are in US dollars.
Acquisition Sample for CAR's Descriptive Statistics Acquirer Target Number Listed 206 89 Unlisted 0 117
Deal Value Mean 221,737,048.54
Std. Dev. 785,081,469.07 Median 64,336,500.00 Min 11,000,000.00 Max 9,000,000,000.00
3.3. Firms’ Stock Return Data
For obtaining the stock return data of the 295 listed companies (206 acquirers plus 89 targets), another database is used. The Library of the University of
Amsterdam also has access to the Thomson Reuters Datastream-‐database, which is a global financial and macroeconomic database covering equities, stock market indices, currencies, company fundamentals, fixed income securities and key economic indicators for 175 countries and 60 markets. In this thesis the ISIN-‐ numbers of the companies and stock indices are used to get their returns and calculate the companies’ cumulative abnormal returns.
4. Methodology and Hypotheses
Statistical research is needed in order to test the hypotheses. Hypothesis 1 and 2 are subject to the question whether there are any significant abnormal returns expected after M&A announcements. An event study is performed to test this. To test hypothesis 3, 4 and 5 multiple regressions are run. In this section both statistical procedures are described in detail and the hypotheses are presented. Each hypothesis is based upon previous literature and is supported in the subsections.
4.1. Event Study
The event study methodology is used to measure the short-‐term economic impact of events, such as M&A announcements, through using daily stock prices (Brown and Warner, 1985). The impact is measured by the abnormal returns in the event window. This is the period of interest and often consists of several days. By all means the announcement day is included in the event window, but periods prior to and after the event may also be interesting to observe
(MacKinlay, 1997).
The abnormal returns are measured by taking the actual return of the security minus the normal return, during the event window. The normal return is defined as the expected return without conditioning on the event taking place. For
modeling this normal return a market model is used. The market model is a statistical model that assumes a stable linear relation between the market return, based on a broad based stock index, and the security’s return. A period before the event window, an estimation window is taken in which the intercept and slope of this linear relation is calculated. The event window itself is logically excluded from the estimation window to avoid possible biases. (MacKinlay, 1997). With the linear estimates for the normal return model, the abnormal return can be calculated as follows. First the expected return on a security is calculated:
𝑅
!,!= 𝛼
!+ 𝛽
!𝑅
!,!+ 𝜀
!,!Where the expected value of the error term is assumed to be zero and its variance to be constant. Then the equation for the abnormal return is the following:
𝐴𝑅
!,!= 𝑅
!,!− 𝑅
!,!The abnormal return observations must be aggregated in order to draw overall inferences for the event on interest. The aggregation is along two dimensions, through time and across securities. We first aggregate the abnormal returns across securities by taking the average of them all to reach the average abnormal return (MacKinlay, 1997).
𝐴𝐴𝑅
!=
1
𝑁
𝐴𝑅
!,! ! !!!The average abnormal returns can then be aggregated over time (the t!, t!
event window) to get the cumulative average abnormal return (CAAR), which is
the main value of interest.
𝐶𝐴𝐴𝑅 𝑡
!, 𝑡
!=
𝐴𝐴𝑅
! !! !!!!4.1.1. Estimation Window
The estimation window is the period in which the security’s relation to the market index is measured. This window should end one day before the start of the event window, but different starting days are used in existing literature. However both MacKinlay (1997) and Brown and Warner (1985) argue in their articles focused on event study methodology that it is best to use an estimation window up to 250 days. In this thesis an estimation window of 252 days will be applied, as this is the length of an entire year of stock trading days nowadays. The NASDAQ composite and the NYSE composite are the stock indices used as
benchmark in this research. All securities in the sample of acquisitions that have missing values in the estimation window will be excluded from the CAAR
analysis. This holds for 5 acquiring companies and for 7 target companies.
4.1.2. Event Window
The event window is the period of days around the event day, being the announcement day in this case. The length of it varies considerably among previous literature. In some researches relatively long event windows are used, ranging from (-‐70,210) used by Neely (1987) to (-‐50,50) used by Wong (2009). Other researches focused on the days prior or after the event, looking at Becher (2000) using (-‐30,5) and Scholtens and de Wit (2004) using (-‐3,31) respectively. However, most researches tend to use rather small event windows. Faccio et al. (2006) and Fuller et al. (2001) both use an event window of (-‐2,2) to search for abnormal returns. Moeller et al. (2003) even uses (-‐1,1) in their study on US acquisitions. MacKinlay (1997) and Brown and Warner (1985) also use different windows here. The first uses a 41-‐day-‐event window (20,20) in his example of an event study, while in the second research it is argued that an 11-‐day-‐event window is best to use. In this thesis an 11-‐day-‐event window is used as well, balanced against all aforementioned studies.
Figure 1: A timeline of events.
An overview of the estimation window and the event window in days. The estimation window ranges from 258 to 6 days before the announcement day and the event window is set 11 days around the announcement day ranging from 5 days before until 5 days after the event day.
4.1.3. Event Study Hypotheses
This thesis’ first hypotheses are on the effect of M&As on the shareholder value of the acquirer. Taking research of different periods into account the hypotheses are presented as follows.
H𝟎: The CAAR for the acquirer is equal to 0. H𝟏: The CAAR for the acquirer is different than 0.
Throughout time and certainly in the US the effect of M&As on the shareholder value of the target seems to be positive. Based upon that this thesis’ second hypotheses are presented as follows.
H𝟎: The CAAR for the target is positive. H𝟏: The CAAR for the target is non-‐positive.
4.2. Regression Analysis and Hypotheses
When the abnormal returns of each individual security is calculated, they will be added over the event window to generate the securities’ cumulative abnormal returns (CAR).
𝐶𝐴𝑅(−5,5)
!=
𝐴𝑅
!,! !! !!!!With these cumulative abnormal returns a number of tests are done. For every test a regression will be run with the firms’ CAR (-‐5,5) as dependent variable. In that way it is measured what factors may explain the CAR (-‐5,5). The regressions are run for acquiring firms and target firms separately. In the next subsections the hypotheses on the abnormal return explaining variables are presented.
4.2.1. Crisis Variable
The main variable of interest in the regressions is this crisis variable. This is a dummy variable, which equals 1 if the acquisition has taken place after June 2009 and 0 if it has taken place before December 2007. This period is regarded as the financial crisis of 2008 according to the U.S. National Bureau of Economic Research. Firms participating in an acquisition in between abovementioned dates were excluded from the regression sample. The hypotheses apply to both acquirer and target and are presented as follows.
H𝟎: The time period in which the M&A takes place has no effect on the CAR. H𝟏: The time period in which the M&A takes place does have an effect on the CAR
4.2.2. Size Variable
This variable is included in the regression for two reasons. Mainly to test whether the size of the acquirer has an effect on its CAR (-‐5,5). Moeller et al. (2003) find higher positive returns for small acquirers than for big ones. Besides, size and its natural logarithm also function as control variables. The hypotheses regarding this variable are presented as follows.
H𝟎: Small acquirers benefit more from M&As than large acquirers.
H𝟏: Small acquirers do not benefit more from M&As than large acquirers.
4.2.3. Listing Variable
The listing variable tests whether the fact that the target is listed has influence on the CAR (-‐5,5) of the acquiring firms. Faccio et al. (2006) find that an acquirer of an unlisted target benefits more from M&A than an acquirer of a listed target. The hypotheses regarding this variable are presented as follows.
H𝟎: Acquirers of unlisted targets have higher CARs than acquirers of listed
targets.
H𝟏: The listing effect does not hold.
5. Results
In this section the empirical results and their significance are presented and the hypotheses are answered. First the results and hypotheses of the event study are discussed and then the regressions are examined.
5.1. Event Study Results
The results obtained by the CAAR-‐analysis are interpreted for acquiring and target firms separately and the answers to the hypotheses as well.
5.1.1. CAAR Acquirer Firms
As shown in table 2, for the acquirer a CAAR of 1.06% is found over the 11-‐day-‐ event window. A shorter event window of 3 days is studied as well and has a CAAR of 0.60%. Both CAARs are significant and it can be observed that most impact is captured in price in the few days around the announcement day. The average abnormal return (AAR) on event day 1 is 0.54% and strongly significant with a p-‐value of only 0.2%. However, apart from weaker, but still positively significant AAR at day 2 of, all the other AARs on the other event days seem to be insignificant.
In the existing literature a slight pattern was documented regarding the short-‐ term value creation for the shareholders of the acquirer. This pattern is that through M&As increasingly often value is created for the acquirer as well. However, the null hypothesis states that the CAAR of the acquiring firm is equal to zero. And with a t-‐value of 2.44 and a matching p-‐value of 1.6% the null hypothesis is rejected. The CAAR is significantly bigger than zero, meaning the alternative hypothesis is accepted.
TABLE 2: Cumulative Average Abnormal Returns for the Acquiring firm. This table shows the CAAR for acquiring firms over an 11-‐day-‐event window around the announcement day. Abnormal returns are calculated by subtracting estimated returns from the actual returns. A t-‐test is used to show the significance of the results.
Event day AAR t-‐value p-‐value
-‐5 0.23% 1.27 0.207 -‐4 -‐0.08% -‐0.91 0.362 -‐3 -‐0.01% -‐0.10 0.921 -‐2 0.15% 1.35 0.179 -‐1 0.01% 0.08 0.938 0 0.05% 0.27 0.788 1 0.54% 3.09*** 0.002 2 0.20% 1.66* 0.099 3 0.07% 0.60 0.546 4 -‐0.09% -‐0.63 0.529 5 0.01% 0.06 0.950 CAAR (-‐5,5) 1.06% 2.44** 0.016 CAAR (-‐1,1) 0.60% 2.03** 0.043 *p<0.10 **p<0.05 ***p<0.01
5.1.2. CAAR Target Firms
For the target firm the CAAR over the 11-‐day-‐event window is 25.64%, as is shown in table 3. This value is extremely significant, given the p-‐value close to zero. The CAAR over the shorter 3-‐day-‐event window is also strongly positive being 24.92% with a p-‐value close to zero as well. Just as with the acquirer the weight of the impact is close around the announcement day. Where for the acquirer day 1 and 2 have most impact, for the target firm the announcement day itself together with event day 1 captures most of the effect. On the
announcement day the AAR is 13.43% and on event day 1 the AAR is 11.33%, both significant with p-‐values close to zero.
Following the existing literature, the null hypothesis stated that the effect of M&As on the shareholder value of target firms was expected to be positive. Regarding the found value for the CAAR of 25.64%, tested with a t-‐value of 10.10 and a matching p-‐value close to zero, the null hypothesis is maintained here.
TABLE 3: Cumulative Average Abnormal Returns for the Target firm. This table shows the CAAR for target firms over an 11-‐day-‐event window around the
announcement day. Abnormal returns are calculated by subtracting estimated returns from the actual returns. A t-‐test is used to show the significance of the results.
Event day AAR t-‐value p-‐value
-‐5 0.05% 0.18 0.858 -‐4 0.10% 0.56 0.578 -‐3 -‐0.22% -‐1.05 0.298 -‐2 -‐0.04% -‐0.19 0.848 -‐1 0.15% 0.31 0.759 0 13.43% 6.39*** 0.000 1 11.33% 4.90*** 0.000 2 0.24% 1.39 0.169 3 0.05% 0.32 0.746 4 -‐0.04% -‐0.37 0.715 5 0.60% 1.14 0.258 CAAR (-‐5,5) 25.64% 10.10*** 0.000 CAAR (-‐1,1) 24.92% 9.95*** 0.000 *p<0.10 **p<0.05 ***p<0.01 5.2. Regression Results
The results of the regressions are presented in an overview for the acquiring and target firms separately. First descriptive statistics are shown of all variables.
5.2.1. Acquirer Results
TABLE 4: Descriptive Statistics of the Acquirer Regression.
This table shows the amount of observations, the mean, the standard deviation, the minimum and the maximum of all variables. The CAR variable is the measured CAR over an 11-‐day-‐event window. Size is in million US dollars. Ln Size is the natural logarithm of Size. Target Listed and After Crisis are dummy variables taking on either 1 or 0.
Acquirer Regression
Descriptive Statistics
Obs Mean Std. Dev. Min Max
CAR 192 0.0133 0.059 -‐0.173 0.362 Size 192 3614.87 10465.49 45.826 80432.92 ln Size 192 7.038 1.368 3.825 11.295 Target Listed 192 0.448 0.499 0 1 After Crisis 192 0.495 0.501 0 1
Running this regression tests three hypotheses. The cumulative abnormal return (CAR) of the acquiring firms is the dependent variable in each model. To test the significance of the variables, in total 6 models are tested. Model 1 and 2 contain all variables and as expected have the highest R-‐squared, meaning that in these models the largest share of the dependent variable is explained.
The main variable of interest in this thesis is the “After Crisis”-‐variable. And in model 1 and model 6 its coefficient seems to be strongly significant. In model 1 its coefficient is 0.022 and in model 6 it is 0.020. This means that when the variable “After Crisis’ takes on value 1, the CAR increases with 2,2% and 2,0% respectively. The hypothesis, stating that time does not affect the CAR of the acquirer, is rejected.
The second hypothesis tested in this regression states that small acquirers tend to benefit more from M&As than large acquirers. The CAR is the measure of benefit here and it is observed to what extent the variables Size and ln(Size) increase this benefit. The hypothesis in other words is stated: The bigger the size of the acquirer, the lower its cumulative abnormal return. So in order to
maintain the hypothesis, the variables must be significantly negative. However, they do not appear to be so and therefore the hypothesis is rejected.
To test the last hypothesis in this regression of the acquirer’s CAR, the listing effect is examined. The hypothesis states that acquirers of unlisted targets benefit more from M&As than acquirers of listed targets. This would imply a negative coefficient on the variable “Target Listed”, which takes on 1 if listed and 0 if not. In all three models in which this variable is included, the coefficient indeed is negative. However, in none of these cases the coefficient reaches a level of significance. Therefore, also this hypothesis must be rejected concluding that the listing effect does not hold.
TABLE 5: Regression Explaining the Acquirer’s CAR.
In all models tested in this overview robust standard errors are used, since it was incorrect to assume homoscedasticity.
Dependent Variable: Acquirer CARs
OLS (1) (2) (3) (4) (5) (6) Constant 0.006 0.009 0.014*** 0.026 0.015*** 0.003 (0.182) (0.672) (0.003) (0.238) (0.004) (0.424) Acquirer Size 0.000 0.000 (0.817) (0.298) ln Acquirer Size 0.000 -‐0.002 (0.901) (0.521) Target Listed -‐0.008 -‐0.009 -‐0.004 (0.355) (0.350) (0.673) After Crisis 0.022** 0.022 0.020** (0.013) (0.110) (0.019) Observations 192 192 192 192 192 192 R-‐squared 0.034 0.034 0.001 0.002 0.001 0.029 *p<0.10 **p<0.05 ***p<0.01 5.2.2. Target Results
TABLE 6: Descriptive Statistics of the Target Regression.
This table shows the amount of observations, the mean, the standard deviation, the minimum and the maximum of all variables. The CAR variable is the measured CAR over an 11-‐day-‐event window. Size is in million US dollars. Ln Size is the natural logarithm of Size. After Crisis is a dummy variable taking on either 1 or 0.
Target Regression
Descriptive Statistics
Obs Mean Std. Dev. Min Max
CAR 81 0.255 0.231 -‐0.247 1.033 Size 81 4773.32 14149.26 53.340 80432.92 ln Size 81 7.108 1.403 3.977 11.295 After Crisis 81 0.617 0.489 0 1
By running these regressions, the last hypothesis is tested, stating that the time period in which the M&A takes place does not have an effect on the abnormal returns. The cumulative abnormal return (CAR) of the target firms is the
dependent variable in each model. The “After Crisis”-‐variable, the main variable of interest, is tested alone in model 5 and together with the control variables Size and ln(Size) in models 1 and 2. What shows is that the coefficient of the “After Crisis”-‐variable is highest and most significant in the model without the control variables. A value of 0.116 is found (with a p-‐value of 1.9%), meaning that after the crisis the CAR as a consequence of an acquisition is 11.6% higher than when the acquisition would have occurred before the crisis. In models 1 and 2, the coefficient is also significant with values of 0.102 and 0.113 (and p-‐values of 0.049 and 0.027 respectively). The null hypothesis states that time would not be of significant influence on CARs, so the hypothesis is rejected.
TABLE 7: Regression Explaining the Target’s CAR.
In all models tested in this overview robust standard errors are used, since it was incorrect to assume homoscedasticity.
Dependent Variable: Target CARs
OLS 1 2 3 4 5 Constant 0.200*** 0.210 0.268*** 0.368*** 0.183*** (0.000) (0.122) (0.000) (0.005) (0.000) Acquirer Size 0.000 0.000* (0.110) (0.009) ln Acquirer Size -‐0.003 -‐0.016 (0.836) (0.343) After Crisis 0.102** 0.113** 0.116** (0.049) (0.027) (0.019) Observations 81 81 81 81 81 R-‐squared 0.069 0.061 0.027 0.009 0.060 *p<0.10 **p<0.05 ***p<0.01
6. Conclusion
This thesis examines the short-‐term effect of acquisitions on the shareholder value of both the acquiring and the target firm. It does so by applying an event study to a sample of acquisitions in the US banking sector between 01-‐01-‐2005 and 31-‐12-‐2014. The study provides evidence that shareholders of acquiring firms earn significant positive abnormal returns of 1.06% over the applied 11-‐ day (-‐5,5) event window (p-‐value of 1.6%). This was not to be expected from existing literature. For the target firm the abnormal returns over the same period of time are with 25.64% strongly significant (p-‐value of 0.0%). This was to be expected from existing literature.
This study also explores the effect of the financial crisis of 2008 on the short-‐ term effect of acquisitions. It does so by regressing the 11-‐day cumulative
abnormal returns on a dummy variable indicating the timing of the acquisition. It turns out that for the acquirer the CAR increases with 2.0% (p-‐value of 1.9%) if the acquisition is made after the crisis of 2008. For the target the CAR even seems to increase with 11.6% (p-‐value of 1.9%). This means that, regardless of acquiring or being acquired, the short-‐term value creation from acquisitions is significantly bigger after the crisis than before the crisis. There is no earlier research in which this has been found.
The two remaining tests on the effect of size and listing on the cumulative abnormal returns yielded no significant results and in that way this research disagrees with existing literature.
For future studies, it would be interesting to explore possible causes of the difference between the effect of M&As before and after the crisis of 2008. An assumption could be that the market was expecting the crisis ahead of the events. This would imply that the negative sentiment outweighed the
expectation of possible synergies. If this is the case, a relatively low (or high) rise in the value of M&A involved shares may even be a more reliable indicator of future movements in the market or a particular industry than price changes in
material, working on the assumption that the difference between before and after is smaller due to less restructuring as a result of government interventions.
References
Becher, D.A., 2000, The Valuation Effects of Bank Mergers, Journal of Corporate
Finance, Vol. 6, pp. 189-‐214.
Brown, S.J., Warner, J.B., 1985, Using Daily Stock Returns: The Case of Event Studies, Journal of Financial Economics, Vol. 14, pp. 3-‐31.
DeYoung, R., Evanoff, D.D., Molyneux, P., 2009, Mergers and Acquisitions of Financial Institutions: A Review of the Post-‐2000 Literature, Journal of Financial
Services Research, Vol. 36, pp. 87-‐110.
Faccio, M., McConnell, J.J., Stolin, D., 2006, Returns to Acquirers of Listed and Unlisted Targets, Journal of Financial and Quantitative Analysis, Vol. 41, No 1.
Fuller, K., Netter, J., Stegemoller, M., 2001, What do Returns to Acquiring Firms Tell Us? Evidence from Firms that Make Many Acquisitions, Journal of Finance, Vol. 57, No. 4, pp. 1763-‐1793.
Houston, J.F., James, C.M., Ryngaert, M.D., 2001, Where do Merger Gains come from? Bank Mergers from the Perspective of Insiders and Outsiders, Journal of
Financial Economics, Vol. 60, pp. 285-‐331.
Houston, J.F., Ryngaert, M.D., 1994, The Overall Gains from Large Bank Mergers,
Journal of Banking & Finance, Vol. 18, No. 6, pp. 1155-‐1176.
Hudgins, S., Seifert, B., 1996, Stockholders and International Acquisitions of Financial Firms: an Emphasis on Banking, Journal of Financial Services Research, Vol. 10, pp. 163-‐180.
MacKinlay, A.C., 1997, Event Studies in Economics and Finance, Journal of
Moeller, S.B., Schlingemann, F.P., Stulz, R.M., 2003, Firm Size and the Gains from Acquisitions, Journal of Financial Economics, Vol. 73, pp. 201-‐228.
Neely, W.P., 1987, Bank Acquisitions: Acquirer and Target Shareholder Returns,
Financial Management, Vol. 16, No. 4, pp. 66-‐74.
Pilloff, S.J., 1996, Performance Changes and Shareholder Wealth Creation Associated with Mergers of Publicly Traded Banking Institutions, Journal of
Money, Credit and Banking, Vol. 28, pp. 294-‐310.
Scholtens, B., de Wit, R., 2004, Announcement Effects of Bank Mergers in Europe and the US, Research in International Business and Finance, Vol. 18, pp. 217-‐228.
Wong, A., 2009, The Effects of Merger and Acquisition Announcements on the Security Prices of Bidding Firms and Target Firms in Asia, International Journal
of Economics and Finance, Vol. 1, No. 2.
Appendix
Graphs provided by the Institute for Mergers, Acquisition and Alliances (IMAA) and referred to in section 1. Introduction.
Announced Mergers & Acquisitions:
Worldwide, 1985-2015e
Source: Thomson Financial, Institute of Mergers, Acquisitions and Alliances (IMAA) analysis (C) 2015 IMAA 0 1'000 2'000 3'000 4'000 5'000 6'000 0 10'000 20'000 30'000 40'000 50'000 60'000 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 YTD 20 15 e V au le of T ra n sa ct ion s (in b il. US D) Nu m b er of T ra n sa ct ion s Year
Number of Deals Value
YTD: August 18
e: expected full year based on Jan 01 - Aug 18
Announced Mergers & Acquisitions:
United States of America, 1985-2015e
Source: Thomson Financial, Institute of Mergers, Acquisitions and Alliances (IMAA) analysis
0 500 1'000 1'500 2'000 2'500 0 2'000 4'000 6'000 8'000 10'000 12'000 14'000 16'000 18'000 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 YTD 20 15 e V au le of T ra n sa ct ion s (in b il. US D) Nu m b er of T ra n sa ct ion s Year Number Value YTD: August 18
Announced Mergers & Acquisitions:
Europe, 1995-2015e
Source: Thomson Financial, Institute of Mergers, Acquisitions and Alliances (IMAA) analysis (C) 2015 IMAA 0 200 400 600 800 1'000 1'200 1'400 1'600 1'800 2'000 0 2'000 4'000 6'000 8'000 10'000 12'000 14'000 16'000 18'000 20'000 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 20 15 YTD 20 15 e V au le of Tr an sa ct ions (in bil. EU R) Number of Tr an sa ct ions Year
Number Value YTD: January 01 - July 24 e: full year, expected based on Jan 01 - Jul 24
Announced Mergers & Acquisitions:
Banks, 1985-2015e
Source: Thomson Financial, Institute of Mergers, Acquisitions and Alliances (IMAA) analysis (C) 2015 IMAA 0 100 200 300 400 500 600 0 500 1'000 1'500 2'000 2'500 19 85 19 86 19 87 19 88 19 89 19 90 19 91 19 92 19 93 19 94 19 95 19 96 19 97 19 98 19 99 20 00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 YTD 20 15 e V au le o f Tr an sa ct io n s (i n b il. US D ) N u mb er of T ran sac ti on s Year Number Value YTD: August 18
