Mergers and Acquisitions
Do target firms get abnormal return from mergers in the US?
Date: 26.06.2017
Student: Alexei Arcea
Student number: 10779671
Faculty of Economics and Business
Code: BSc ECB
2 This document is written by student Alexei Arcea, who declares to take the 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.
Introduction/relevance
Business expansion opens doors to new markets and diversifies risks. Nevertheless, it comes at cost and requires large amount of additional capital to invest. Takeovers happen when an acquirer can provide the capital necessary for the expansion and starts several individual steps of merging process before becoming a single entity with the target. On the one hand, merging doesn’t come at no risk. Uncertainty can arise from the insufficient information on the target companies. On the other hand, profits can be made from unexploited synergies.
The focus is on the target side to compare the cost of merging with the expectations of future synergies. The unclear relation between risk and return around the date of the merger announcement has been analyzed by many researchers in the past. Nevertheless, most of them applied CAPM, when estimating the normal returns. The aim of this thesis is to obtain the insight on the behavior of abnormal returns during the merger process including risk factors that are omitted by CAPM. In this thesis merger cases are assessed also with Fama and French 4-factor model (FF4), which adjusts the results to target’s size, value and momentum.
An event study on 2505 merger cases between 1995 through 2009 is taken to obtain more significant results than in the previous literature. The research question is formulated as follows: Do target firms get abnormal return from mergers in the US? To answer this question and relate it to the existing literature, the zero hypothesis tests cumulative abnormal returns against zero for different time-frames and for both CAPM and FF4. The results bring us the step closer in explaining the abnormal returns arising from takeovers and are relevant for all participants in takeover negotiations. This research can serve as a benchmark for estimating break-even in premium paid in a merger transaction.
This thesis is structured and divided in several chapters according to the following table of contents:
3
Contents
Introduction/relevance ... 2
1 Literature review ... 3
1.1 Economic theory of M&A ... 3
1.2 Empirical evidence and reasoning for the difference ... 5
2 Methodology ... 6
2.1 Hypothesis ... 6
2.2 Description of financial models applied ... 6
2.3 Estimating Abnormal Return ... 8
2.4 Data and sampling ... 10
3 Results ... 11
4 Conclusion ... 14
5 References ... 16
1 Literature review
1.1 Economic theory of M&A
Merger and acquisition (M&A) deals are now commonly used by companies all over the world to pursue their goals and objectives related to strategic growth (Gaughan, 2005). All U.S. industries have been impacted by M&A deals, with some of most large firms in the U.S. economy being to some extent products of past M&A (Mueller, 1997).
According to Gersdorff & Bacon (2009), an acquisition or a merger can be defined as a combination of two firms where the bidder usually pays a premium. Depending upon the synergies involved, abnormal returns should arise. Nevertheless, the difference between a merger and an acquisition may not actually matter, since both transactions conclude to the same result: two (or more) companies, which previously had a separate ownership, start to operate as one legal entity after the M&A deal takes place. This happens with the aim to attain some financial or strategic objectives (Sherman and Hart, 2006).
The abnormal return differs among acquisitions, which can be categorized as friendly or hostile takeovers. When acquiring firm makes an offer to target’s shareholders without consulting with the management boards takeover is hostile. In contrast to that in a friendly takeover both shareholders and current management try to get to a mutual agreement. In addition, hostile takeovers are more likely to occur when the acquirer is uncertain about the information provided about the financial situation of the target firm (Schnitzer, 1996).
Mergers happen for the reason that acquiring companies are willing to pay acquisition premium, substantially greater amount than the target firm’s market value. According to Nielsen and Melicher (1973), those exist for the reason that management believes that
4 expected synergy obtained by acquisition is bigger than the premium. Therefore, increase in premium increases the acquirer’s break-even for entire transaction and this creates a conflict of interest between target’s and acquirer’s shareholders (Schoenberg, 2003). The aim of this thesis is to quantify and generalize the amount of premium paid (significance of abnormal return) for the large amount of takeovers in the past in order to conclude the winner of the conflict described by Schoenberg (2003).
Since abnormal return is influenced by expected synergies, these should be
categorized. Most common synergy profits are economies of scale benefits or diminishing of administrative and overhead costs (Johnson et al., 2008). In addition, economies of scope are relevant, because merged entities could make an advantage of cross selling and collaborate in creation of new products (Sevenius, 2003). Further advantages are financial and market synergies. Financial synergies are obtained by reducing cost of capital with the decrease in the risk. Market synergies increase the market power by better negotiations with suppliers and customers.
The event of M&A is a special case in a life-time of some companies and directly affects target’s and bidder’s prices of common stock with a different magnitude. Creating and sustaining mergers involves several challenges due to the complexity of synergy valuation, which influences the price paid for the target. Sometimes the value is overestimated, which creates losses for acquirer’s shareholders (Sirower, 2004).
Rankine (2001) described merger-failure as a situation when shareholders value does not increase or pre-specified goals at the date of buying the firm are not achieved. Again often failures happen due to overestimating synergies. Johnson et al (2004) describes three reasons for it: optimistic managers, poor financial advice and lack of experience. Another reason why disappointing results may arise is the information asymmetry: when acquirer has incomplete information, he will tend to overpay. This transfers the wealth to shareholders of the target firm and creates abnormal returns.
Another important issue that Hendrikse (2003) mentioned is the principal-agent problem, which is “disastrous” in the merger making process. The remuneration of the managing board should be in line with endowing a merger. Negative abnormal returns may arise with the conflict of interest between CEO (agent) and shareholders (principal).
Therefore, takeovers often lead to golden parachutes – large compensation payments to managing board triggered upon a change in control, which causes agents focus on their own gains instead of their responsibilities for shareholders.
More empirical literature and economic theory explanations are challenged and discussed. Followed by the explanation of the data gathering and its corresponding methodology. Lastly, the results are presented, described and concluded.
5
1.2 Empirical evidence and reasoning for the difference
Past research indicates that around the M&A announcement, target firms earn significant
positive abnormal returns (Dodd & Ruback 1977, Jensen & Ruback 1983, Franks, Harris &
Titman 1991, Campa & Hernando 2004, Swanstrom 2006, Chakraborty 2010), as compared to the bidding firms, which abnormal returns are not significantly different from zero (Jensen & Ruback 1983, Campa & Hernando 2004). From here arises the acquirer’s risk of
overestimating synergies and the need for further research on the required amount of premium paid in merger transactions.
Jarrell & Poulsen (1989) studied US companies between 1963 and 1986 and found significant positive abnormal returns for the target firms within the period of -20 to +10 days around the date of the announcement.
The work of Jensen and Ruback (1983) describes the results of 13 empirical studies from 1956 to 1981. They conclude that targets’ shareholders receive abnormal return varying between 20-30% around the date of the announcement, which is consistent with the results of Jarrell and Poulsen (1989).
These studies were followed by Mulherin and Boone (2000) that analyzed the sample of 376 targets from 1990-1998 and found that the median abnormal return in 3 days around the announcement date is 18.4%.
Returns are influenced by several factors that include the method of payment (cash, stock or mix), the kind of acquisition (domestic or cross border), bidder’s asset base and the type of merger (horizontal, vertical or conglomerate). Some past papers conclude that cash acquisitions generate higher positive abnormal returns than stock offers, due to the tax-exemption for capital gains (Wansley, Lane & Yang 1983, Huang & Walking 1987).
In addition, Chen, Chou, & Lee (2011) distinguished between the daytime and overnight transactions and found that an overnight acquisition announcement with a cash transaction involved tends to earn significantly positive abnormal return, which is insignificant for daytime announcements.
A study by Rosen (2006) indicated that when there is a large number of mergers, target firms earn high returns around the announcement date, whereas in the long run their returns are reversed. These long run results were estimated to be even worse than the short run results in the cold merger markets. However, supporters of short-term effects research argue that the market’s initial reaction is a good predictor of the actual long-term
performance of a deal (McWilliams and Siegel, 1997).
There are many papers with similar conclusions. Nevertheless, the impact of M&A announcement on target’s returns depends on the chosen event-window, market benchmark and the sample selection (Andrade, Mitchell, & Stafford 2001, Scholtens & Wit 2004).
6
2 Methodology
2.1 Hypothesis
The null hypothesis to the question whether target firms get abnormal return from mergers in
the US is:
H0: Target firms do not incur abnormal returns H1: Target firms do incur positive abnormal returns
Six one-sample t-tests are executed for Raw, CAPM and FF4 testing if CARs 10 and 20 days after the announcement are significantly different from zero.
The results are assessed with one-sided significance level, because the past research clearly specifies increase in the demand for target shares and its increase of the price.
The limitation of this methodology is that there are no control variables to check for firm specific factors that influence the abnormal returns. This thesis focusses on the
abnormal returns and not the factors that influence them.
There is no available database or source of information or to verify terms of
controlling interest for bidding firms within M&A transactions. Therefore, following Moeller et al. (2004), in this thesis a M&A transaction is defined as a deal in which an acquirer
increases its holdings from less than 50% to more than 50% of stock or assets.
Lower results than in the existing literature are expected, due to different time frames. For example, Dodd and Ruback (1983) was researching the merger activity between 1958 and 1976. In the contrast to the last century this decade can be described as century of technological and informational revolutions with improved information systems and advanced technological techniques to avoid the overestimation of synergies and decrease the premium and therefore, decrease in abnormal return of target firms.
In this investigation measures the difference between actual and normal return that ignores the announcement of the merger. The issue with the second one is that it has to be estimated. In the contrast to the most of existing literature this is done with both CAPM and Fama & French 4-factor (FF4) model and compared with the raw returns.
2.2 Description of financial models applied
Most researchers mentioned in the empirical evidence section applied CAPM when estimating the normal returns and following them CAPM is used here to get comparable results and check for differences in period chosen. Further aim of this thesis is to obtain the insight on the behavior of abnormal returns during the merger process including risk factors that are omitted by CAPM. In this thesis merger cases are assessed also with Fama and French 4-factor model (FF4), which adjusts the results to target’s size, value and momentum.
7 The Capital Asset Pricing Model (CAPM), is considered “the most important model of the relationship between risk and return” (Berk & DeMarzo, 2013, p.351). Even several decades after the introduction of the CAPM, it is still recognized as a “powerful and
intuitively pleasant” tool, which is applied for estimating risk and relating expected return to the systematic risk (Fama & French, 2004, p. 25)
There are three assumptions underlying CAPM. First one is the possibility for
investors to buy and sell all securities at competitive market prices ignoring transaction costs and incurring taxes. In addition, it should be possible to lend and borrow at the risk free interest rate. Secondly, CAPM assumes that investors hold only efficient portfolios of traded securities. The last assumption of CAPM is that investors have homogenous expectations regarding the expected return of securities, volatilities and correlations (Berk & DeMarzo, 2013, pp.379-380). In other words, CAPM regresses the excess return on a certain security (𝑟𝑟𝑖𝑖− 𝑟𝑟𝑓𝑓) on the market risk premium (𝐸𝐸[𝑟𝑟𝑚𝑚] − 𝑟𝑟𝑓𝑓) and indicates the sensitivity to systematic risk with the coefficient beta (β):
�𝑟𝑟𝑖𝑖− 𝑟𝑟𝑓𝑓� = β�𝐸𝐸[𝑟𝑟𝑚𝑚] − 𝑟𝑟𝑓𝑓� (1)
Where 𝑟𝑟𝑖𝑖 is the return on the stock 𝑖𝑖, 𝑟𝑟𝑓𝑓 is the risk-free interest rate, and 𝐸𝐸[𝑟𝑟𝑀𝑀𝑀𝑀𝑀𝑀] is the expected return on the market portfolio (Berk & DeMarzo, 2013, p.341).
However, to make the equation (1) complete, it is necessary to account for the short-term fluctuations from the market security line (denoted by α), and the unsystematic risk that can be diversified (denoted by the error term ε) (Berk & DeMarzo, 2013, p.404). Hence, after including these factors equation (1) changes to:
�𝑟𝑟𝑖𝑖− 𝑟𝑟𝑓𝑓� = 𝛼𝛼 + β�𝐸𝐸[𝑟𝑟𝑀𝑀𝑀𝑀𝑀𝑀] − 𝑟𝑟𝑓𝑓� + 𝜀𝜀 (2)
Consider equation (1) and for simplicity in explanation the excess return on a certain security (𝑟𝑟𝑖𝑖− 𝑟𝑟𝑓𝑓) is replaced by 𝑌𝑌, and the market risk premium (𝐸𝐸[𝑟𝑟𝑀𝑀𝑀𝑀𝑀𝑀] − 𝑟𝑟𝑓𝑓) by 𝑋𝑋. Then equation (1) transforms to:
𝑌𝑌 = β𝑋𝑋 (3)
Because the variables 𝑟𝑟𝑖𝑖, 𝑟𝑟𝑓𝑓, and 𝐸𝐸[𝑟𝑟𝑀𝑀𝑀𝑀𝑀𝑀] are taken from the data, the term beta (β) is estimated. It can be seen as an indicator of stock’s cyclicality. In the case when returns of an asset follow precisely the returns of the market, then the risks associated with that stock’s returns are relatable to the risks of the market and measured in units of systematic risk: beta (β) (Berk & DeMarzo, 2013, p.337).
Nevertheless, Fama and French criticize the empirical record of CAPM, because of the simplifying assumptions mentioned earlier. Researchers describe the importance of other variables like size, various price ratios and momentum that add to the average returns
8 In 1992 Fama and French expanded the CAPM by adding the size and value factors to the market factor and later in 1997 they introduced the momentum factor, which all together leads to the FF4: Fama & French 4-factor model (Carhart, 1997):
�𝑟𝑟𝑖𝑖− 𝑟𝑟𝑓𝑓� = 𝛼𝛼 + β1�𝐸𝐸[𝑟𝑟𝑀𝑀𝑀𝑀𝑀𝑀] − 𝑟𝑟𝑓𝑓� + β2(SMB) + β4(HML) + β5(UMD) + 𝜀𝜀 (4)
The equation 4 is a multifactor model, where (SMB) stands for “small minus big” and adjusts normal stock’s excess return to the difference between the return on the portfolio of small stocks and the return on the portfolio of large stacks. Next factor (HML: “high minus low”) represents the difference between the return on the portfolio of high-book-to-market stocks and the return on the portfolio of low-book-to-market stocks.
The last factor momentum (UMD) is described as the tendency for the stock price to continue rising if it is going up and to continue declining if it is going down. Momentum can be calculated by substracting the equal weighted average of the worst preforming firms from the equal weighted average of the highest performing firms (Carhart, 1997).
Already in 1992 Fama and French proved that the 3-factor model (excluding UMD) reduces the average absolute value of alpha from 25 to 30 basis points per month predicted by CAPM to 5-10 basis points per month. Researchers explain the deviation in the results arguing that their multifactor model captures more variation in the cross-section of average stock returns and absorbs the anomalies that CAPM cannot explain.
2.3 Estimating Abnormal Return
The methodology assumes a rational capital markets, where the merger event directly affects stock prices of target companies. Traditional event study uses the calculation of the
Cumulative Average Abnormal Return (CAAR) that starts with the following regressions (Fama et al, 1969):
CAPM: 𝑅𝑅𝑖𝑖 = 𝛼𝛼𝑖𝑖+ 𝛽𝛽𝑅𝑅𝑅𝑅 + 𝜀𝜀𝑖𝑖 (5)
FF4: 𝑅𝑅𝑖𝑖 = 𝛼𝛼𝑖𝑖+ 𝛽𝛽1𝑅𝑅𝑅𝑅 + 𝛽𝛽2[𝑆𝑆𝑅𝑅𝑆𝑆] + 𝛽𝛽3[𝐻𝐻𝑅𝑅𝐻𝐻] + 𝛽𝛽4[𝑈𝑈𝑅𝑅𝑈𝑈] + 𝜀𝜀𝑖𝑖 (6) The regression 5 is a single-index model CAPM, where 𝑅𝑅𝑅𝑅 represents the excess return on CRSP U.S. Total Market Index adjusted to the daily return on the One Month Treasury Bill, which is assumed to be the most appropriate choice for the risk free rate. In this implication the market factor of the 4-factor model is the same as the one in CAPM, in other words same choice for market portfolio and risk free rate.
First, CAPM and Fama & French 4-factor model betas are computed from the estimation window by regressing the excess target return either only on market factor in regression 5 and then on [Mktrf, Smb, Hml, Umd] in the regression 6.
9
Timeline 7: Takeover window for estimation
(7)
The estimation window (see timeline 7) is chosen from t= -280 and t= -30 in order to have a long frame for computing betas and exclude a possible information leak in the month of the merger.
CAPM: 𝐴𝐴𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑖𝑖− 𝛼𝛼𝑖𝑖− 𝛽𝛽𝑅𝑅𝑅𝑅𝑅𝑅 (8) FF4: 𝐴𝐴𝑅𝑅𝑖𝑖 = 𝑅𝑅𝑖𝑖− 𝛼𝛼𝑖𝑖− 𝛽𝛽1𝑅𝑅𝑅𝑅𝑅𝑅 − 𝛽𝛽2[𝑆𝑆𝑅𝑅𝑆𝑆] − 𝛽𝛽3[𝐻𝐻𝑅𝑅𝐻𝐻] − 𝛽𝛽4[𝑈𝑈𝑅𝑅𝑈𝑈] (9) The 𝐴𝐴𝑅𝑅𝑖𝑖 (Abnormal return) in the equations 8 and 9 is calculated for the event-window by subtracting risk factors from the pre-event window from the actual return (𝑅𝑅𝑖𝑖). This is done for every firm for each day starting 20 days before and ending 20 days after the event. 21 day 𝐶𝐶𝐴𝐴𝑅𝑅𝑖𝑖= � 𝐴𝐴𝑅𝑅𝑖𝑖 10 −10 (10) 41 day 𝐶𝐶𝐴𝐴𝑅𝑅𝑖𝑖= � 𝐴𝐴𝑅𝑅𝑖𝑖 20 −20 (11)
Because this thesis exams not only the performance at the event date, but the performance over a longer period surrounding the merger, cumulative abnormal returns (CAR) for both CAPM and FF4 are computed for event-windows [-10, 10] and [-20, 20], where abnormal returns are aggregated from 10 (20) days before the announcement date, up to 10 (20) days after (see equations 10 and 11).
𝐶𝐶𝐴𝐴𝐴𝐴𝑅𝑅 = 1
𝑁𝑁 � 𝐶𝐶𝐴𝐴𝑅𝑅𝑖𝑖
𝑁𝑁
𝑖𝑖=1
(12)
It is not very informative to analyze each firm’s return data separately, because of stock price movements unrelated to the event under study (p. 7, Jong, 2007). Therefore, the internal validity is improved by averaging the information over total number of companies. Basically the CARs are aggregated over the cross-section of mergers to obtain cumulative average abnormal returns: CAAR (see equation 12).
𝐺𝐺 = √𝑁𝑁𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝐶𝑠𝑠 ≈ 𝑁𝑁(0,1), (13) where 𝑠𝑠 = � 1
10 Given the null hypothesis to be tested H0: E(CAR)=0, large sample size and
assuming that CARs are mutually uncorrelated, the t-test given in the equation 13 is approximately normally distributed.
2.4 Data and sampling
In order to investigate only stocks that are traded with sufficient liquidity and eliminate non-significant takeovers, this empirical research focusses only on mergers, where both target and acquirer are publicly listed US companies, target has more than $10Mln market
capitalization and acquirer obtains more than 51% of shares outstanding. The announcement dates of mergers are gathered from Thomson One database for the period between 1995-2009, which sum to 2505 observations.
After gathering the announcement days, daily stock returns and risk factors [Mktrf, Smb, Hml, rf, Umd] are gathered from Wharton CRSP and assigned to each case by constructing a variable “tday” that is equal to zero for the return at the announcement date and that counts trading days between 365 calendar days before and 180 after the
announcement to obtain long frame for computing risk factors. This is done with duplicating each observation 545 times and lead to 2505*(1+545) = 1,367,730 daily observations.
The daily observations are assigned to each merger case by constructing a variable “tday” that is equal to zero for the return at the announcement date and that counts trading days before
According to Chen, Chou, & Lee (2011) timing of acquisition announcement could lead to the overestimation of the abnormal returns. Therefore, 79 cases, where date of announcement is not a trading day, are excluded from this investigation (see appendix 1a). Excluding the non-trading days, 806,047 daily factors are assigned to 2426 merger cases left.
Estimation of abnormal returns requires either complex statistical techniques or assumptions that are questionable. In this thesis approximation to the normal distribution is done by replacing top 5% and lowest 5% price outliers with 95th and 5th percentiles, which leads to 80,488 daily factors being changed.
Furthermore, for obtaining only significant risk factors other 94 merger cases (4857 daily observations) were excluded from the investigation (see appendix 1b). The condition is specified as: at least 100 trading days in the pre-event window [-280,-30].
Then CAPM and Fama & French 4-factor model betas are computed from the pre-event window by regressing the excess target return (Ri) either only on [Rm] or [Rm, Smb, Hml, Umd]. This leads to computing 5 factors for each of 2317 merger cases left. For descriptive statistics about the factors computed see table 1a below:
11
Table 1a: Descriptive statistics about risk factors
Coefficient CAPM FF4 Rm 0.541 0.629 (0.479) (0.517) Smb - 0.516 - (0.560) Hml - 0.165 - (0.701) Umd - -0.044 - (0.531) Number of observations 2,317 2,317 Notes: figures between parenthesis are
standard deviations of mean values above them
3 Results
Figure 1: Target returns around Acquisition Announcements
Aggregating CARs over the cross-section of events leads to cumulative average abnormal returns (CAAR), which are listed in the appendix 2 and are graphically represented in the figure 1.
From the results obtained it is possible to infer that there is on average a significant information leak before the date of the announcement. For all models and time-frames, the CAAR one day before the day of the announcement makes around 50% of the total CAAR (see figure 1).
The information leak may happen because of the insider trading. It comes at the cost for acquirer’s shareholders, increases the premium and adds to the probability of merger
12 resulting in a financial failure. This result indicates that the companies should further work on the confidentiality of merger negotiations and improve the security of information systems.
Table 1b: Skewness/Kurtosis tests for Normality
Frame CARs Obs. Pr (Skewness) Pr (Kurtosis)
[-10, 10] Raw CAPM 48,593 48,593 0 0 0 0 FF4 48,593 0 0 [-20, 20] Raw 94,818 0 0 CAPM 94,818 0 0 FF4 94,818 0 0
Even after replacing the top 5% and lowest 5% price outliers with 95th and 5th percentiles, which leads to 80,488 daily factors being changed, the zero hypothesis of normality is rejected (see Table 1b and Figure 2). Therefore, following calculations are using robust standard errors.
13
Table 2: Descriptive statistics about CARs after the announcement
Frame Model CARs
Day #10 [-10,10] 2,308 Obs. Raw 0.089*** (0.108) CAPM 0.081*** (0.105) FF4 0.079*** (0.108) Day #20 [-20,20] 2,302 Obs. Raw 0.111*** (0.144) CAPM 0.097*** (0.138) FF4 0.092*** (0.142)
Notes: (i) figures between parenthesis are robust standard
deviations of mean values above them.(ii)Further, *** denotes statistical significance at 1% level
There is enough evidence to infer that CARs 10 and 20 days after the announcement are significantly different from zero, because of the large sample size. Three one-sample t-tests are executed for each time frame. In the event-window [-10, 10] CAARs for Raw, CAPM and FF4 are 8.9%, 8.13% and 7.85%. The corresponding t-values are 39.54, 37.11 and 34.82 respectively, which are significant at 1% (see table 2).
When expanding the window to [-20,20] CAARs increase to 11.1%, 9.7% and 9.17% with the decrease in t-values to 37.08, 33.73 and 30.95 (see table 2), which still remain significant at 1% level. The intuition of this result: increase in the time period over which price changes are measured increases the variance of the abnormal return, whereas the true abnormal return remains the same.
For the sample of chosen targets and the event-window of (-20,20) the significant and positive CAAR of 7.97% (CAPM) and 7.7% (FF4) on the date of the announcement (see Table 3) is consistent with the research from earlier time periods. Nevertheless, as expected, lower positive results are obtained than in the existing literature, even though a longer time frame was chosen for computing the abnormal returns.
For example, Dodd and Ruback (1983) was researching the merger activity between 1958 and 1976. In the contrast to the last century this decade can be described as century of technological and informational revolutions with improved information systems and advanced technological techniques to avoid the overestimation of synergies and decrease the premium and therefore, decrease in abnormal return of target firms.
14 In the event-window (-10,10) the CAAR on the date of the announcement is
significant and positive (see table 3): 6.85% (CAPM) and 6.72% (FF4). Here the CAAR is lower due to the fact that the ARis are positive on the average and are aggregated through a shorter time-frame, which is consistent with the economic theory.
Table 3: Descriptive statistics about CARs at the announcement
Frame Model CARs
Day #0 [-10,10] 2,317 Obs. Raw 0.072*** (0.092) CAPM 0.068*** (0.090) FF4 0.067*** (0.090) Day #0 [-20,20] 2,317 Obs. Raw 0.087*** (0.119) CAPM 0.080*** (0.114) FF4 0.077*** (0.114) Notes: (i) figures between parenthesis are robust
standard deviations of mean values above
them.(ii)Further, *** denotes statistical significance at 1% level
Lastly, the obtained CAAR applying CAPM for both time-frames are higher than the obtained CAAR applying FF4 and lower than the CAAR of raw returns. This is consistent with the economic theory as well due to the fact that FF4 captures more variation in the cross-section of average stock returns and absorbs anomalies that CAPM cannot explain.
4 Conclusion
This research aims to answer the question if the announcement of a merger has a significant impact on the stock price return of the target firm. Answering this research question, the movement of each individual company’s stock price is observed within the given sample and evaluated around the announcement date. After estimating risk coefficients for both CAPM and FF4 the CAAR was calculated in the forty-one and twenty-one event periods. The sample for this research consists of US mergers from 1995 till 2009. After eliminating and transforming the data, the corrected sample includes 2308 merger cases.
The results show a significant positive abnormal return for both models in both time-frames. The abnormal returns are higher for CAPM than for FF4 in both event-windows, which is consistent with the economic theory due to better explanatory power of FF4.
15 Furthermore, results obtained in this research are lower (as expected) than the ones in the existing literature (Dodd and Ruback, 1983). Transactions in the selected sample are more recent and this can be explained by the use of advanced information systems that overestimate less the synergies and decrease the target’s abnormal returns together with the premium.
From the financial and economic perspective the results obtained indicate a decrease in the acquirer’s risk of overestimating the synergies, because of the decrease in abnormal return. Furthermore, the acquirer should adjust expectations to target’s size, value and momentum in the calculation of the amount of premium in the tender offer.
There are several limitations of this research. Some factors that influence the returns are ignored in this investigation. Those include the method of payment (cash, stock or mix), the kind of acquisition (domestic or cross border), bidder’s asset base, the type of the merger (horizontal, vertical or conglomerate), timing of the announcement and the activity on the merger market. For the further research those important merger characteristics should be included in order to explain better the behavior of abnormal returns.
16
5 References
Andrade, G., Mitchell, M. & Stafford, E. (2001). New evidence and perspective on mergers.
Journal of Economic Perspectives, 15(2), 103-120.
Berk, J., & DeMarzo, P. (2013). Corporate Finance (3
rded.). Boston: Pearson Addison
Wesley.
Campa, M. & Hernando, I. (2004). Shareholder value creation in European M&A. European
Financial Management, 10, 47-81.
Carhart, M. M. (1997). On Persistence in Mutual Fund Performance. The Journal of Finance.
52(82). 1540-6261
Chakraborty, M. (2010). The Wealth Effects of Takeover Announcement for Firms in the
Financial Services Sector in India. Journal of Emerging Market Finance, 100-227.
Chen, S. S., Chou, R. K., & Lee, C. (2011). Bidders’ strategic timing of acquisition
announcements and the effects of payment method on target returns and competing
bids. Journal of Banking & Finance, 35(9), 2231–2244.
Dodd, P. & Ruback, R. (1983). Tender offers and stockholder returns: An empirical analysis.
Journal of Financial Economics, 5(3), 351-373.
Fama, E., Fisher, L., Jensen, L., Roll, R. (1969). The adjustment of stock prices to new
Information. International economic review, 10, 1-21.
Fama, E. F., & French, K. R. (1993). Common risk factors in the returns on stocks and bonds.
Journal of Financial Economics, 33, 3-56.
Fama, E. F., & French, K. R. (2004). The Capital Asset Pricing Model: Theory and Evidence.
Journal of Economic Perspectives, 18(3), 25-46.
Gaughan, P. A. (2005). Mergers: What Can Go Wrong and How to Prevent It. John Wiley &
Sons, Hoboken.
Gersdorff, N. V. & Bacon, D. F. (2009). U.S. Mergers and Acquisitions: A Test of Market
Efficiency. Journal of Finance and Accountancy, 1(8), 1-8.
Huang, Y. S. & Walking, R. A. (1987). Target Abnormal Returns Associated with
Acquisition Announcement. Journal of Financial Economics, 19, 329-349.
Jarrell, G. & Poulsen, A. (1989). The returns to acquiring firms in tender offers: evidence
from three decades. Financial Management, 18, 12-19.
Jensen, M. & Ruback, R. (1983). The market for corporate control: The scientific evidence.
Journal of Financial Economics, 11, 5-50.
Moeller, S., Schlingemann, F., & Stultz, R. (2004). Firm Size and the Gains from
Acquisitions. Journal of Financial Economics, 73(2), 201-228.
17
