Abnormal returns to U.S. bidding firms of M&A transactions from 2010 to 2014 : the impact of border, industry and relative firm size

(1)

Abnormal Returns to U.S. Bidding Firms of M&A

Transactions from 2010 to 2014

The impact of border, industry and relative firm size

UNIVERSITY OF AMSTERDAM

FACULTY OF ECONOMICS AND BUSINESS

BSc Economics & Business

Specialisation in Finance and Organization

Author:

C. LI

Student number:

10621768

Thesis supervisor: Dr. Jan Lemmen

(2)

PREFACE AND ACKNOWLEDGEMENTS

The inspiration for me to write this thesis was a report indicating that Alibaba experienced declines in its stock price recently every time it announced a merger or acquisition. Even though many scholars have conducted studies regarding the impact of mergers and acquisitions, it is still interesting to start from literature reading and model conduction. As a Bachelor student, it is extremely hard for me to establish a new model or come up with transformative results. However, the learning process of writing a Bachelor’s thesis has proven to be valuable for me. I would like to thank both the university as well as my supervisor who taught me how to properly conduct academic research and think critically. There is never a fixed answer, and there is always an argument to question perceived truths. As a result, I should always be ready to question and analyze every possible conclusion. The same applies for life outside of academics, I will be prepared to always be open and never stop thinking critically.

Thank you all for your guidance and assistance over the past three years.

STATEMENT OF ORIGINALITY

This document is written by Chenhan Li 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.

COPYRIGHT STATEMENT

The author has copyright of this thesis, but also acknowledges the intellectual copyright of contributions made by the thesis supervisor, which may include important research ideas and data. Author and thesis supervisor will have made clear agreements about issues such as confidentiality.

(3)

ABSTRACT

This study examines whether in recent years, from January 1, 2010 to December 31, 2014, announcements of domestic and cross-border mergers and acquisitions have similar effects on publicly traded U.S. companies as in previously conducted research. It is interesting to have a look into the market’s reaction to these transactions after the world economic crisis. The study’s main findings are that there are significantly higher cumulative abnormal returns for domestic mergers and acquisitions compared to cross-border deals. Also, cumulative returns for M&A transactions involving two firms operating in the same industry are observed to be significantly higher. But the relation to the relative firm size is not significant enough to reject the hypothesis.

Keywords: Event study; Acquisition; Abnormal return; U.S. stock market; cross-border; domestic JEL Classification: G14

(4)

LIST OF TABLES

Number Table name Page

Table 1 Data description of independent variables p. 11

Table 2 Daily returns summary p. 11

Table 3 Data description of dependent variables p. 12

Table 4 Standardized cross-sectional t-statistic of cumulative abnormal returns p. 13

Table 5 Regression Model of three days’ event window p. 14

(6)

CHAPTER 1: Introduction

A large volume of mergers and acquisitions (“M&A”) transactions have been witnessed across the world since last century, in line with increased globalization. Even though M&A transactions happen every day throughout the world, the majority of firms transact with firms operating in their own countries. Based on Hitt, Harrison and Ireland’s (2001) research on M&As between 1999 and 2000, only 40% of M&A transactions involved companies from two different countries. Nevertheless, it has become harder to tell which country the company belongs to. A corporation might be merged into a foreign company which later would be acquired by a domestic group. The borders of business are disappearing and the whole world has become flatter.

If a company is willing to enter a new industry or expand their business in a new region, there are plenty of methods to take. Within all of those efforts they could make, M&A represents a shortcut to integrate existing business activities. With the booming of the mergers and acquisitions, academic research on such transactions is one of the most popular areas in financial studies (Agrawal, Jaffe & Mandelker, 1992，p. 1605). Agrawal, Jaffe and Mandelker (1992) also pointed out that even though mergers and acquisitions are continuously studied, there is never an absolute or final conclusion of the effects on the bidders’ market value. Most of scholars who have conducted studies surrounding the announcement of mergers and acquisitions regard the abnormal return as a measurement of effects of the announcement. With regard to the announcement of M&A transactions, investors might react differently to different types of deals based on information available in the market. Therefore, looking into peoples’ expectations, as reflected in the share prices of the free float, towards specific deal fundamentals might help find the disciplinary movement of the valuation of M&A deals. Since both markets and investor preferences are changing, the research of the announcement of M&A transactions in the current business environment will always have added value when compared to prior studies.

The motivation of mergers and acquisitions is to expand a firm’s geographical reach or lines of business. The resources of the target company are integrated into the bidding firms. Therefore, the announcement of mergers and acquisitions should enhance investors’ confidence in the future return of the firm. To support this position, there is an improvement in a firm’s market value which should be reflected by the positive return around the announcement of both domestic and cross-border M&A transactions. Morck and Yeung (1992, p. 54) stated that positive abnormal returns were observed in their research on the announcement of international acquisitions. The finding was the same as what Markides and Ittner (1994) summarized. Moreover, Markides and Ittner (1994) also pointed out that the positive abnormal returns only applied to M&A transactions between firms from two countries. For domestic M&A transactions, they were convinced that zero and even negative added value would be observed post-announcement.

(7)

However, in the examination of U.S. acquiring firms, Datta and Puia concluded that shareholders of acquirers experienced a significant negative return around the announcement of cross-border targets (1995, p. 352). Negative post-merger performance of U.S. acquirers was also confirmed in another study of U.S. takeovers regardless of whether it was a domestic or international transaction (Franks, Harris and Titman, 1991, 94).

Contradicting outcomes of prior studies show that more research is needed to understand how M&A announcements have impacted share prices of acquirers in recent years. Therefore, I would like to do the study on M&A and stock prices again to find an answer to the abnormal return around announcement of mergers and acquisitions.

In addition to this curiosity, the motivation for me to do this research is to study whether investors’ expectations towards M&A transactions have changed. The majority of studies that scholars have conducted are outdated. For instance, Franks et al. studied 399 U.S. acquisitions between 1975 and 1984 (1995, p. 81). Morck and Yeung collected data of 322 foreign acquisitions by U.S. firms between 1978 and 1988 (1992, p. 45). Similarly, Markides and Ittner (1994, p. 343) did the research of events involving U.S. firms between 1975 and 1988 while Datta and Puia (1995, p. 337) used a wider time period for their research, from 1978 to 1990. Therefore, I assume that the prior studies may not reflect the effect of announcements of M&A transactions in the current era. After experiencing the financial crisis and witnessing the post-merger performance of different types of M&A transactions in previous decades, investors could have a new perspective towards M&A transactions and their different attitudes might be reflected in the abnormal return around the announcement of M&As.

There are diverse factors investors might take into consideration when they individually evaluate the M&A transactions. Within all the factors prior studies have mentioned, I am highly interested in the nature of such transactions, for instance whether acquirers and targets are from the same country and whether they operate in the same industry. In addition, from the intrinsic market value perspective, if the deal value is really small compared to the market value of the acquirer, the influence of the target to the acquirer might be relatively limited. Therefore, I would like to also consider the relative size of transaction valuations and market values of acquirers as a third variables. I am going to study the regression relationship between abnormal returns and these variables.

In this paper, I will employ the event study method used in most of previous studies to conduct the research on the abnormal return around the announcement of mergers and acquisitions for American public acquirers in current years.

(8)

CHAPTER 2: Theoretical Background

Before running the data and doing the regression analysis, the understanding of the basic concepts and theoretical overview are needed to help establish the research model. Also, from previous studies, observed relations could be viewed as references.

2.1 The concept of M&A

Mergers and acquisitions are two different terms of an event involving two corporations. If we refer to the textbook, mergers refer to the combination of two organizations while only the acquirers exist after the mergers. With regard to acquisitions, both targets and acquirers remain and there is no requirement to change the organizational structures or names. Ma et al. (2009) mentioned in their studies that it is not necessary to distinguish between the two concepts or to study them separately. No matter if it is a merger or an acquisition, bidding firms will obtain targets’ resources including their talents, technologies, market shares, brand value, etc.

The reason why corporations seek M&A transactions is that the management team believes in the potential benefits generated by strategic transactions. Acquirers are aiming to make profits out of their investments and realize their potential after integration (Shimizu, Hitt, Vaidyanath & Pisano, 2004). As a result, the announcement of M&A transactions is supposed to stimulate investors and as a result increase the firm value and stock prices of the acquirer. However, this is not always the case as shown in prior studies.

If investors would react to a special event, the announcement of mergers or acquisitions, abnormal returns should be observed around the event date. Therefore, I will investigate whether there are significant abnormal returns in the samples I studied. Also, I will test whether the abnormal return is positive or negative for bidding firms.

2.2 Cross-border M&A

Due to the dynamic circumstances of global business, the popularity of cross-border mergers and acquisitions has increased. Searching for international opportunities are strategically beneficial for bidding firms. Nevertheless, if compared to local M&A transactions, cross-border M&As have more uncertain elements such as a different economic and cultural environment for bidding companies to learn (House, Javidan, Hanges & Dorfman, 2002). Information asymmetry in foreign markets also

(9)

makes successful integration more difficult (Zaheer, 1995). Legitimacy issues and government constraints are also important factors for operational success in foreign countries. The more similar background the two players have, the easier they can cooperate and perform well after the mergers or acquisitions. “Liability of foreignness” and “double-layered acculturation” were introduced by former scholars to summarize the difficulties of cross-border M&A transactions (Zaheer, 1995; Barkema, Bell & Pennings, 1996). But risks always come with opportunities. Even though Shimizu et al. (2004) pointed out the difficulty to fully accomplish strategic initiatives with different cultural and operational backgrounds, they also believe M&A transactions can bring more opportunities in the targets’ market and capabilities of new business initiatives and projects based on greater combined resources. In the paper which Vermeulen and Barkema (2001) discussed about the learning benefit that comes from acquisitions, a company’s knowledge base was addressed as the main pursuit of cross-border acquisitions.

On the one hand, local acquisitions may have less risk and investors are more willing to be optimistic regarding future returns; on the other hand, cross-border acquisitions may expand the business into two markets while local deals may cannibalize each company’s market share. Which kind of transaction is more welcomed by investors remains unclear. Therefore, I will set the characteristic of cross-border or not as a dummy variable to study whether this factor is influential for abnormal return, and also see whether the influence is positive or negative.

2.3 Industry matters of M&As

The industry of acquirers and targets might influence the effects brought about by M&A transactions (Burns & Liebenberg, 2011, p. 1031). The degree of successful integration may be related to the similarity of two parties. For example, different technologies or production processes may require different allocation of labor and physical resources. On the one hand, economies of scale could strengthen strategic power via multi-market operations (Choe & Yin, 2007). In addition to this, a positive relationship between product diversification and abnormal returns was suggested in Barkema and Vermeulen’s study (1998). On the other hand, Rajan, Servaes and Zingales (2000) pointed out that conflicts between management teams in diversified firms tended to result in a misallocation of resources.

In theory, operating in different business lines after an M&A transaction has potential to both add values as well as to have negative effects, so predictions regarding post-merger performance are not definitive. For diversified firms, productivity is higher compared to single product companies (Schoar, 2002) and resources can be more efficiently allocated (Choe & Yin, 2009). However, diversification discounts are observed and discussed in previous studies (Lang & Stulz, 1990).

(10)

Investors do not have to gain professional knowledge about technology nor access to detailed contracts. So, it is unrealistic to expect investors to know definitely how two corporations will develop based on their combined assets and resources. With possibilities of both outcomes, it is also necessary to see how the same-industry factor influences the abnormal returns.

2.4 Relative size of two parties

Normally, an M&A transaction with a transaction value of ten million would definitely post diverse impacts on stock returns for bidders with market value of either one hundred million or twenty million. A relatively more influential power of an M&A transaction is expected to be observed when the deal value is close to acquirers’ market value, which results in a meaningfully higher relative firm size. However, Kitching (1967) indicated that relatively larger acquirers had better integration capacity so they might have better post-merger performance. Similar to this, Hennart and Reddy (1997) found the difficulty of integrating targets whose enterprise values are higher. What is more, Asquith, Bruner and Mullins Jr. (1983, p. 122) explained that the abnormal return produced for the bidding firms was negatively related to the equity value of targets, if it is assumed that the same net present value accrues to the bidding firm. The reason could be that investments into smaller firms are expected to have higher return rates. Even though there is very low chance that targets are start-ups if the targets are also public corporations, small firms still have more potential to earn higher rates of return, especially for technology-oriented firms (Brouthers & Brouthers, 2000).

Moeller, Schlingemann and Stulz (2004) also studied the relationship between the relative size of bidders and targets and the excess return of acquisitions. Moeller et. al. observed higher significant abnormal returns for bidders with smaller equity value and this finding was robust over time, ignoring form of payments or the other deal characteristics (2004, p. 226). The significantly positive relationship between cumulative abnormal returns and relative firm size were also observed with similar research (Jarrell & Poulsen, 1989). With regard to the significance, Travlos (1987) could not find the significance of the coefficient for the variable of relative firm size in all of his regressions.

Therefore, in order to test whether the relationship between abnormal returns and relative firm size has changed, I will also regress the cumulative abnormal return on the relative size ratio between bidding firms and target firms. There are two parts of study, the first one is to test the positive or negative effects and the second one is to see whether the effect is significant.

(11)

2.5 Research questions

The reason why I chose American acquirers is that the United States (“U.S.”) has a long history of mergers and acquisitions, so data availability is better on such transactions. In addition, the U.S. government also has strict laws and regulations against hostile takeovers. Furthermore, the U.S. stock exchange is supervised and managed better than that of any other country. Therefore, the unexpected and unusual effects, or even illegal manipulation of stock prices, are highly prohibited. We could assume that all the movements of stock prices are the result of the behaviour of rational investors.

In this paper, the research questions on U.S. acquirers are whether the abnormal returns around the announcement of M&As are affected by cross-border (𝐻3), same industry indication (𝐻4) and ratio between transaction values and the market values of acquirers (𝐻5). Before this, I would test whether there is significant abnormal return of the event (𝐻1) and whether the abnormal return is positive or negative (𝐻2).

(12)

CHAPTER 3: Methodology

To answer the research questions with the event study methodology, the assumption of efficient capital markets is adopted here. Stock prices in efficient capital markets instantaneously reflect all of the publicly available information of publicly traded companies (Markides & Ittner, 1994, p. 349). As a result, investors’ expectations are assumed to be directly reflected in the stock prices. Markides and Ittner also summarized that in most cases, the event study model includes the historical relationship between the market return and a firm’s individual return, which could be used to estimate the “normal” return during an event window.

In order to establish the function to determine the estimation window to be used for prediction, the market model introduced by MacKinlay (1997) is used in this paper. The market model is used for each sample and the market model is as below:

𝑅𝑖𝑡 = 𝛼𝑖 + 𝛽𝑖𝑅𝑚𝑡+ 𝜀𝑖𝑡

where:

𝑅𝑖𝑡 = return on the security of firm i at time t;

𝑅𝑚𝑡 = return on the market portfolio at time t. The market index we adopt in this research is the S&P 500 Index, and the return on the market portfolio at time t is the daily movement of S&P 500 Index;

αi and βi = parameters of the relationship between the return on the individual security and that of the market; and

𝜀𝑖𝑡 = residual of the relationship at time t, assumed to be distributed normally with mean equal to zero, a constant variance over the control and prediction periods, and zero correlation between residuals over time (𝜀𝑖𝑡 ~ 𝑁(0, 𝑆2).

With regards to the selection of the estimation window in prior studies, Brown and Warner (1985) selected the window of (-244, -6); Boehmer et al. chose the estimation period (-249, -11); while Ma, Pagan and Chu (2009) used the window of only 120 trading days, (-125, -6). Moreover, Campbell and MacKinlay stated in their book (1997) that a normal estimation window in event study was between 120 days and 210 days. Therefore, the estimation window used in this research is 150 days, which is from 160 trading days before the announcement to the 11 trading days before the announcement, simply noted as (-160, -11). After calculating 𝛼𝑖 and 𝛽𝑖, we can use the model above to calculate the expected daily return based on market return. Then the part of actual returns beyond the estimated returns are regarded as “abnormal returns” and could be expressed as the function:

(13)

where:

𝐴𝑅𝑖𝑡: the abnormal return for firm i on day t; 𝑅𝑖𝑡: the actual return for firm i on day t;

𝛼̂𝑖, 𝛽̂𝑖: the value of coefficient and constant part for firm i resulted from market model in estimation window.

Even though Markides and Ittner (1994) pointed out that the two-day abnormal return is sufficient to study the impact of the M&A transaction announcements, I also used a five-day event window, (-2, 2). Accordingly, the window is from two days prior to the announcement to two days after the announcement. Moreover, to compare it with a shorter event window of three days, I would also study the event window of (-1, 1). The reason why I also include days before the announcement is to absorb the impact of rumors regarding the transaction prior to the official announcement.

Therefore, the cumulative three-or- five-days’ abnormal return of firm i is:

𝐶𝐴𝑅𝑖(−𝑑,𝑑)= ∑ 𝐴𝑅𝑖𝑡 𝑡=𝑑

𝑡=−𝑑 where d equals to 1 or 2.

Next step is to conduct the “t-test” to study whether 𝐶𝐴𝑅𝑖(−𝑑,𝑑) is equal to 0, which is the first hypothesis of this paper (𝐻1). There are plenty of different t-statistic methods used in previous studies. For instance, widely adopted methods included the student-t statistic, Patell t-statistic (Patell, 1976), Brown and Warner t-statistics (1980) and cross-sectional t-test. Boehmer et al. (1991) compared the rejection rates of various significance tests and proposed a hybrid statistics combining standardized-residual and the ordinary cross-sectional test, called standardized cross-sectional test. Boehmer et al. pointed out that previously used event-study methods too frequently reject the null hypothesis and they demonstrated the stronger power of the standardized cross-sectional test. Therefore, I will conduct the t-statistcs as Boehmer et al. proposed, formulated as below:

𝑡 𝑠𝑡𝑎𝑡𝑖𝑠𝑡𝑖𝑐 = 1 𝑁 ∑ 𝐴𝑖 𝐸 𝑁 𝑖=1 /√ 1 𝑁(𝑁 − 1)∑(𝐴𝑖 𝐸− ∑ 𝐴𝑖 𝐸 𝑁 ) 𝑁 𝑖=1 2 𝑁 𝑖=1

For the first hypothesis (𝐻1: 𝐶𝐴𝑅𝑖(−𝑑,𝑑) = 0), in order to test whether the cumulative abnormal return is significantly different from zero, a two-tailed critical value for t-statistics should be used. Furthermore, in order to compare the sign of cumulative abnormal returns with previous studies, a one-tailed critical value for t-statistics is used for 𝐻2(𝐶𝐴𝑅𝑖(−𝑑,𝑑) < 0). Once we can conclude from the hypothesis that there is a significant non-zero effect of announcement on the return of acquirers (the cumulative

(14)

abnormal return is significantly different from zero), we can study the factors of abnormal returns we discussed before. So we do regression of 𝐶𝐴𝑅𝑖,(−𝑑,𝑑) on 𝑆𝐼𝐶𝑖 and 𝐺𝐸𝑂𝑖:

𝐶𝐴𝑅𝑖,(−𝑑,𝑑)= 𝛼1+ 𝛽1∗ 𝑆𝐼𝐶𝑖+ 𝛽2∗ 𝐺𝐸𝑂𝑖+ 𝛽3∗ 𝑇𝑅𝐴𝑁𝑆𝑖+ 𝜀𝑖 where:

𝛼1: constant term for regression;

𝐺𝐸𝑂𝑖: Dummy variable, 𝐺𝐸𝑂𝑖 = 1 when the headquarter of acquirer i is based in the same country as target i in event i, otherwise 𝐺𝐸𝑂𝑖= 0;

𝑆𝐼𝐶𝑖: Dummy variable, 𝑆𝐼𝐶𝑖 =1 when the SIC code is the same for target i and acquirer i; otherwise, 𝑆𝐼𝐶𝑖 = 0. (SIC code for firms is provided by Thomson Mergers and Acquisitions One Database, indicating the industry the company operates in);

TRANSi: the ratio of value of transaction and the market value of acquirer at the end of previous month of event date;

𝜀𝑖: a noise term including other factors for 𝐶𝐴𝑅𝑖,(−2,2);

With regards to the measurement of relative size of bidders and targets, the ratio (𝑇𝑅𝐴NSi) is calculated as the value of transaction divided by the market value of equity of the acquiring companies. Specifically, the numerator is retrieved in conjunction with the transaction data, while the denominator is the result of multiplying stock price and the number of shares outstanding of the last day of the previous month of the event date. The reason why I choose transaction value is that it depends on the equity value of the target and it also reflects the acquirers’ expectations regarding the targets. The more positive attitude acquirers have, the higher the premium of the transaction. Moreover, this definition is similar to previous studies as Moeller et al. (2004) used the sum of all consideration payment and Asquith et al. (1983) used the market value at the end of previous year of event date.

In order to conduct the regression models, I will use the linear regression model embedded with OLS assumptions in Stata. With the output of the Stata linear regression program, the p value for each coefficient of control variables could be used to evaluate null hypotheses.

In summary, the four null hypotheses studied in this paper are:

𝐻1 : 𝐶𝐴𝑅𝑖,(−𝑑,𝑑) = 0; 𝐻2 : 𝐶𝐴𝑅𝑖,(−𝑑,𝑑) < 0;

𝐻3 : 𝛽1 = 0; 𝐻4 : 𝛽2 = 0; 𝐻5 : 𝛽3 = 0;

(15)

CHAPTER 4: Data and empirical results

4.1 Data preparation

In my study, I required acquirers’ daily stock prices to calculate normal and abnormal returns. In addition to stock prices, I also needed fundamental information about the transaction. For example, the bidding firms and targets’ nations, the industry code of both acquirers and targets, and also other information to be used as criteria to help finalize the events list. As a result, there are two parts of data collection.

4.1.1 Mergers and acquisitions deals information collection (independent variables)

Part one is to collect the basic information of both the bidding firms and the target firms in mergers and acquisitions transactions. The database I used is the Thomson ONE Mergers & Acquisitions Database, which includes relatively complete information of every M&A transaction throughout the world. In the Thomson ONE Mergers & Acquisitions Database, not only are the announcement date, countries of acquirers and targets, deal value and public status available, but the Standard Industry Classification (“SIC”) Codes of acquirers and targets are also available to categorize corporations by industry. By using the customized data request, I have successfully retrieved all of the publicly available data of announced M&A transactions, whose acquirers are from the United States and also public, between January 1, 2010 to December 31, 2014.

To distinguish among almost twenty thousand events, I set additional criteria. First, I excluded the events of which the SIC Code of the acquirers were related to following industries: government, utilities, social welfare, financial and investment banking institutions. The reason is that the motivation of such mergers and acquisition transactions may not be exclusively market expansion, but also potentially politics- or capital-oriented. Secondly, I only kept the events whose deal value are between 100 million to 1000 million U.S. dollars. This is the most concentrated range of M&A transactions, and an abnormally low or high deal value would affect the resulting share price. Thirdly, only the transactions in which acquirers purchased 100 percent of the target’s shares were selected in order to remove the effects of different acquired portion of targets. Finally, I used only transactions in which target firms were also public to ensure investors have access to public information of their business.

After all the filters, only 151 events were left, of which 42% were events between two firms with the same SIC Codes (SIC = 1) while 20% are cross-border mergers and acquisitions transactions (GEO = 0). Transaction value came together with the retrieved information for each deal.

(16)

In order to obtain the market value of acquiring parties (the stock price multiplied by the number of shares outstanding at the end of previous month), I used the query form in Monthly Stock File of The Center for Research in Security Prices (“CRSP”) to retrieve data regarding share prices and the number of shares outstanding on the last transaction day of each month from December 2009 to December 2014. Then I matched them to 151 deals and calculated on the corresponding date for each sample.

Table 1 Data description of independent variables

Variables Obs Mean Std. Dev Min. or

percentage of Dummy=0 Max. or percentage of Dummy=1 SIC 151 0.4238 0.4958 58% 42% GEO 151 0.8013 0.4003 20% 80% TRANS 151 0.3011 0.4495 0.0007 3.0824

* Based on data retrieved from Thomson One M&A database and market value calculations. For SIC dummy variables, observations with the same SIC codes have SIC = 1; For GEO dummy variables, observations that are local acquisitions have GEO = 1.

4.1.2 Abnormal returns collection (dependent variable)

Part two of data collection was to obtain abnormal returns of each deal. Wharton Research Data Services (“WRDS”) has established an Event Study Program based on CRSP. It can help run the event study on the unique PERMNO code for company and a specific event date, providing estimation parameters and the selected risk model. After transferring acquirers’ names into PERMNO code and uploading the file containing 151 pairs of PERMNO code and event date(s), a program produces an automatic report of daily or cumulative total returns and abnormal returns.

Here below is the summary table of daily returns.

Table 2 Daily returns summary

Day Mean Total

Return

Mean Abnormal Return

Std. Dev. Standardized Cross-sectional t-statistics for

daily Abnormal Return - 2 - 0.0010 - 0.0017 0.0177 - 1.8576 - 1 0.0021 0.0008 0.0154 0.3166 0 0.0101 0.0092 0.0540 2.8734 1 0.0034 0.0031 0.0448 1.7550 2 - 0.0042 - 0.0030 0.0196 - 1.3106

(17)

* Standard deviation conducted here is based on normal calculation, attached in Appendix. Standardized cross-sectional t-test is the same as formulated in Methodology.

For cumulative abnormal returns in event windows of (-1, 1) and (-2, 2), variables’ description is reported below. I have successfully calculated cumulative abnormal returns for event windows of both 3 days and 5 days.

Table 3 Data description of dependent variables

Variables Obs. Mean Std. Dev. Min. Max.

CAR(-1,1) 151 0.0132 0.0757 - 0.3496 0.3606

CAR(-2,2) 151 0.0085 0.0768 - 0.4559 0.2486

* Standard deviation here is the same as traditional calculation attached in Appendix.

4.2 Empirical results

After the collection of all of the independent and dependent variables, the data is ready to be studied to test the five hypotheses. With regard to the t-statistics, a significance level 95% is adopted in this paper. As in 𝐻1, the standardized cross-sectional t-test is used. Therefore, the rejection regions for two-tailed t-statistical are (-∞, -1.96) and (1.96, ∞). For 𝐻2, it should be a one-tailed test and on the right direction. So the rejection region for 𝐻2 is (1.65, ∞). In addition to this, 𝐻3, 𝐻4 and 𝐻5 that aim to test whether 𝛽1, 𝛽2 and 𝛽3 in regression (1) and (2) equal to zero will be measured by p-values resulted from Stata. For p-values, the rejection region with the same significance level is (< 0.05).

4.2.1 Significance test of abnormal returns

From the query report of the Event Study Program provided by WRDS, the cumulative abnormal returns are available for each deal. Based on the design of standardized cross-sectional method (Boehmer et al., 1991), the t-statistics for 𝐶𝐴𝑅(−1,1) is 3.233 and the t-statistics for 𝐶𝐴𝑅(−2,2) is 2.380. Both statistics fall into the rejection region of 95% of significance level. As a result, the null hypothesis of 𝐻1 is rejected at alpha level of 5%. This outcome is consistent with most previous scholars’ findings. In addition to this, we can see extremely significant daily abnormal return on date zero from Table 2. The rejection of cumulative abnormal returns of both three event days and five event days indicates that announcements have continuous influential power in the trading days following the event dates, not only on the day of announcement.

(18)

Table 4 Standardized cross-sectional t-statistic of cumulative abnormal returns

* Calculation are based on the daily abnormal returns provided by Event Study Program. Method of standardized cross-sectional t-statistics are detailed explained in Methodology.

From the overview of Table 2, daily abnormal returns only show significant rejection under 95% confidence level on day 0, which is exactly the event date of the announcement. However, the cumulative abnormal returns of three days and five days around the announcement date show high t-statistics. What is more, the t-statistic of the three days’ window has a higher rejection probability compared to the five days’ window. Therefore, the daily abnormal return is smaller when the date is further from the event date. The graph of cumulative returns by days is attached in Appendix, which illustrates that rumor influences the abnormal returns from one day before the announcement. In addition to this, the graph also shows that the highest cumulative abnormal return exists on day one. The reason for the downward trend of cumulative abnormal return after day one is that the abnormal return on day two is negative, which is presented in Table 2 before.

As for 𝐻2, standardized cross-sectional t-statistics of both 𝐶𝐴𝑅(−1,1) and 𝐶𝐴𝑅(−2,2) are in the rejection region of one-tailed t-test. Therefore, the hypothesis of 𝐶𝐴𝑅(−𝑑,𝑑) < 0 is rejected at significance level of 95% and together with 𝐻1, the significantly positive cumulative returns are observed for both three-day and five-day event windows.

After the rejection of null hypothesis of 𝐻1 (𝐶𝐴𝑅𝑖,(−𝑑,𝑑) = 0), more studies of the proposed factors affecting the abnormal returns around the announcement date of mergers and acquisitions could be continued.

4.2.2 Regression analysis

In the methodology part, the regression model (1) is defined as below:

𝐶𝐴𝑅𝑖,(−𝑑,𝑑) = 𝛼1+ 𝛽1∗ 𝑆𝐼𝐶𝑖+ 𝛽2∗ 𝐺𝐸𝑂𝑖+ 𝛽3∗ 𝑇𝑅𝐴𝑁𝑆𝑖+ 𝜀𝑖 (1) where:

𝛼1: constant term for regression;

𝐺𝐸𝑂𝑖: Dummy variable, 𝐺𝐸𝑂𝑖 = 1 when the headquarter of acquirer i is based in the same country as target i in event i, otherwise 𝐺𝐸𝑂𝑖= 0;

Variables Standardized cross-sectional t-statistic

CAR (-1,1) 3.2332

(19)

𝑆𝐼𝐶𝑖: Dummy variable, 𝑆𝐼𝐶𝑖 =1 when the SIC code is the same for target i and acquirer i; otherwise, 𝑆𝐼𝐶𝑖= 0. (SIC code for firms is provided by Thomson Mergers and Acquisitions One Database, indicating the industry the company operates in);

TRANSi: the ratio of value of transaction and the market value of acquirer at the end of previous month of event date;

𝜀𝑖: a noise term including other factors for 𝐶𝐴𝑅𝑖,(−2,2);

After inputting data of 151 samples and assigning three days’ cumulative abnormal returns as dependent variables in Stata, the output of Stata showed as below in Table 5:

Table 5 Regression Model of three days’ event window

Source SS df MS Number of obs. = 151

Model 0.1177 3 0.0392 F (3,147) = 7.76

Residual 0.7429 147 0.0051 Prob > F = 0.0001 Total 0.8606 150 0.0057 R-squared = 0.1368 Adj R-squared = 0.1192 Root MSE = 0.0711 CAR (-1,1) Coef. Std. Err. t P >丨 t 丨 [95% Conf. Interval]

SIC 0.0396 0.0117 3.37 0.001 0.0164 0.0627

GEO 0.0408 0.0145 2.81 0.006 0.0121 0.0694

TRANS 0.0212 0.0129 1.64 0.103 －0.0043 0.0468

_cons －0.0426 0.0142 －2.99 0.003 －0.0707 －0.0145 * Regress 𝐶𝐴𝑅(−1,1) on 𝑆𝐼𝐶𝑖 , 𝐺𝐸𝑂𝑖 and 𝑇𝑅𝐴𝑁𝑆𝑖 (the ratio between transaction value of deals and the

market value of bidding firms described in Chapter 3).

From the results above, the regression model could be translated as below:

𝐶𝐴𝑅𝑖,(−1,1)= −0.0426 + 0.0396 ∗ 𝑆𝐼𝐶𝑖+ 0.0408 ∗ 𝐺𝐸𝑂𝑖+ 0.0212 ∗ 𝑇𝑅𝐴𝑁𝑆𝑖 (a)

To interpret the equation, if 𝑆𝐼𝐶𝑖 equals to one, the acquiring firm is operating in the same industry as the targeting firm, the cumulative abnormal return of (-1, 1) would increase by 0.0396 points. Secondly, if 𝐺𝐸𝑂𝑖 equals to one, the deal is a local acquisition, the abnormal return in three days’ event window would increase by 0.0408 points. Thirdly, compared to the 𝑇𝑅𝐴𝑁𝑆𝑖 ratio of one, the cumulative abnormal return of three days is 0.0212 points higher if transaction value is two times of the acquirer’ market value (𝑇𝑅𝐴𝑁𝑆𝑖 = 2).

(20)

The positive coefficients of 𝑆𝐼𝐶𝑖 and 𝐺𝐸𝑂𝑖 are aligned with the traditional expectations for the combination of two companies. In other words, investors have stronger confidence in the future performance of domestic transactions and transactions between two parties doing business in the same industry. And the higher transaction values of mergers and acquisitions as compared to the acquiring firms’ market value, the higher the cumulative abnormal returns.

However, not all three coefficients are significantly different from zero. As for the p-value of two-tailed t-statistics, it is 0.001 for 𝛽1 and 0.006 for 𝛽2. Both of them fall into the rejection region of P with alpha level of 5%. So the null hypothesis three (𝛽1 = 0) and null hypothesis four (𝛽2 = 0) are significantly rejected. As a result, positive impacts of local acquisitions and positive impacts of same-industry deals are observed in regression (1). Even though there is positive 𝛽3 in the regression, the p-value for hypothesis five is 0.103, which is larger than the 0.05. Therefore, the hypothesis five (𝛽3 = 0) cannot be rejected in the regression (1). Asquith et al. (1983) also observed the positive coefficient for the relative size of two parties, but it is significantly different from zero.

For the five event days’ window of regression (1), output of Stata showed as below:

Table 6 Regression Model of five days event window

Source SS df MS Number of obs. = 151

Model 0.1075 3 0.0358 F (3,147) = 6.77

Residual 0.7773 147 0.0053 Prob > F = 0.0003 Total 0.8848 150 0.0059 R-squared = 0.1214 Adj R-squared = 0.1035 Root MSE = 0.07272 CAR (-2,2) Coef. Std. Err. t P >丨 t 丨 [95% Conf. Interval]

SIC 0.0439 0.0120 3.65 0.000 0.0201 0.0676

GEO 0.0375 0.0148 2.53 0.012 0.0082 0.0669

TRANS 0.0012 0.0132 0.09 0.927 －0.0249 0.0274

_cons －0.0405 0.0146 －2.78 0.006 －0.0693 －0.0118 * Regress CAR (-2, 2) on 𝑆𝐼𝐶𝑖 , 𝐺𝐸𝑂𝑖 and 𝑇𝑅𝐴𝑁𝑆𝑖 (the ratio between transaction value of deals and the

market value of bidding firms). To interpret the regression function:

𝐶𝐴𝑅𝑖,(−2,2)= −0.0405 + 0.0439 ∗ 𝑆𝐼𝐶𝑖+ 0.0375 ∗ 𝐺𝐸𝑂𝑖+ 0.0012 ∗ 𝑇𝑅𝐴𝑁𝑆𝑖 (b)

Similar with the regression function (a) for the three days’ window, there are positive coefficients for all three variables in the regression of five days’ window cumulative abnormal return. In addition to this,

(21)

the hypothesis three (𝐻3 : 𝛽1 = 0) is significantly rejected with a p value of 0.000 and the hypothesis four (𝐻4 : 𝛽2 = 0) is also significantly rejected with a p value of 0.012. Still the same as the result for function (a), the p value (0.927 > 0.05) is not small enough to reject the null hypothesis five (𝐻5 : 𝛽3 = 0).

Comparing the function (a) to the function (b), 𝛽1 is higher in function (b); this implies that the cumulative abnormal returns triggered by transactions with participants in the same industry are higher in a longer event window. In contrast to this, 𝛽2 is lower in function (b), indicating that the cumulative abnormal returns are less sensitive to the geographical location of acquirers and targets in the longer event window.

In total, with regression model (1), the null hypotheses of β1 = 0 and β2 =0 are rejected at a significance level of 95%. Moreover, the data show positive coefficients for both the three days’ event window and the five days’ event window. This finding is in line with previous studies and is not surprising based on basic logic. Investors are more willing to have confidence about future performances involving bidding organizations and target organizations whose headquarters are based in the same country or who have the majority of their operations in the same industry.

The failure of rejection to hypothesis five (𝐻5 : 𝛽3 = 0) is the most interesting part of this study. Even though both positive and negative coefficients are observed in this paper, their high p values make the influence on the cumulative abnormal returns negligible. This violates the finding of how the size of the target influences the bidding companies’ market value found in most of previous studies.

(22)

CHAPTER 5: Conclusion

In my paper, I followed the methodology of event study and studied the relation between cumulative abnormal returns and the characteristics of M&A transactions of bidding firms that are public in the U.S. from January 2010 to December 2014. Some findings are consistent with previous research and could be understood by logical deduction. But the results for the relative size of the value of a transaction and as compared to the market value of the bidder were not in line with my expectations. Compared to previous studies conducted in the previous century, investors’ expectations and actual returns from mergers and acquisition remain as concluded previously for both border-related and industry-related factors. In other words, investors still prefer M&A transactions with less risk, specifically between companies with similar geographical and industries. The reason for enjoying same-industry mergers and acquisitions might be the diversification discount mentioned before. Empirical studies suggested that diversification discount were resulted from inefficient internal capital markets (Choe and Yin, 2009).

These findings suggest that after economic crisis, investors’ expectations towards M&A transactions are still similar to what scholars observed before. For the relation between cumulative abnormal returns and the relative size of two firms, even though it is tested to insignificantly influence the cumulative abnormal returns under my model, I would still suggest including relative size ratio as a control variable in future research.

I would suggest to do further research into abnormal returns using different measurement methodologies of the relative size ratios. The reason could be that the influence of the large volume transaction values is offset by the higher risk involved. Investment and returns are always related to risks, so more control variables for risks are required to be included in the future studies. In addition to this, using inequality of SIC codes to describe the diversification should be reconsidered. Some industries might not be the same, but they are related like suppliers or distributors. Therefore, the relation of industries between two parties needs further research and a more detailed definition. In addition to this, whether the target firm is public or private could also be a dummy variable as private firms might have limited information available in market. Finally, the method of payment of mergers and acquisitions can be also studied in future research.

(23)

REFERENCES

Agrawal, A., Jaffe, J. F., & Mandelker, G. N. (1992). The post‐merger performance of acquiring firms: a re‐examination of an anomaly. The Journal of Finance, 47(4), 1605-1621.

Asquith, P., Bruner, R. F., & Mullins, D. W. (1983). The gains to bidding firms from merger. Journal

of Financial Economics, 11(1), 121-139.

Barkema, H. G., Bell, J. H., & Pennings, J. M. (1996). Foreign entry, cultural barriers, and learning. Strategic Management Journal, 17(2), 151-166.

Barkema, H. G., & Vermeulen, F. (1998). International expansion through start-up or acquisition: A learning perspective. Academy of Management journal, 41(1), 7-26.

Boehmer, E., Masumeci, J., & Poulsen, A. B. (1991). Event-study methodology under conditions of event-induced variance. Journal of Financial Economics, 30(2), 253-272.

Brouthers, K. D., & Brouthers, L. E. (2000). Research notes and communications: Acquisition or greenfield start-up? Institutional, cultural and transaction cost influences. Strategic Management

Journal, 21, 89-97.

Brown, S. J., & Warner, J. B. (1980). Measuring security price performance. Journal of Financial

Economics, 8(3), 205-258.

Brown, S. J., & Warner, J. B. (1985). Using daily stock returns: The case of event studies. Journal of

Financial Economics, 14(1), 3-31.

Burns, N., & Liebenberg, I. (2011). US takeovers in foreign markets: Do they impact emerging and developed markets differently? Journal of Corporate Finance, 17(4), 1028-1046.

Campbell, J. Y., Lo, A. W. C., & MacKinlay, A. C. (1997). The Econometrics of Financial Markets (Vol. 2, pp. 149-180). Princeton, NJ: Princeton University Press.

Datta, D. K., & Puia, G. (1995). Cross-border acquisitions: An examination of the influence of relatedness and cultural fit on shareholder value creation in US acquiring firms. MIR:

Management International Review, 337-359.

Firth, M. (1996). Dividend changes, abnormal returns, and intra-industry firm valuations. Journal of

financial and Quantitative Analysis, 31(2).

Franks, J., Harris, R., & Titman, S. (1991). The postmerger share-price performance of acquiring firms.

Journal of Financial Economics, 29(1), 81-96.

Harris, R. S., & Ravenscraft, D. (1991). The role of acquisitions in foreign direct investment: Evidence from the US stock market. The Journal of Finance, 46(3), 825-844.

Hennart, J. F., & Reddy, S. (1997). The choice between mergers/acquisitions and joint ventures: The case of Japanese investors in the United States. Strategic management journal, 18(1), 1-12. Hitt, M. A., Harrison, J. S., & Ireland, R. D. (2001). Mergers & acquisitions: A guide to creating value

(24)

House, R., Javidan, M., Hanges, P., & Dorfman, P. (2002). Understanding cultures and implicit leadership theories across the globe: an introduction to project GLOBE. Journal of World

Business, 37(1), 3-10.

Jarrell, G. A., & Poulsen, A. B. (1989). The returns to acquiring firms in tender offers: Evidence from three decades. Financial management, 12-19.

Jarrell, G. A., & Bradley, M. (1980). The economic effects of federal and state regulations of cash tender offers. The Journal of Law & Economics,23(2), 371-407.

Kitching, J. (1967). Why do mergers miscarry. Harvard Business Review,45(6), 84-101.

Lang, L., Ofek, E., & Stulz, R. (1996). Leverage, investment, and firm growth. Journal of financial

Economics, 40(1), 3-29.

Ma, J., Pagan, J. A., & Chu, Y. (2009). Abnormal returns to mergers and acquisitions in ten Asian stock markets. International Journal of Business,14(3), 235.

MacKinlay, A. C. (1997). Event studies in economics and finance. Journal of Economic

Literature, 35(1), 13-39.

Markides, C. C., & Ittner, C. D. (1994). Shareholder benefits from corporate international diversification: Evidence from US international acquisitions. Journal of International Business Studies, 343-366. Moeller, S. B., Schlingemann, F. P., & Stulz, R. M. (2004). Firm size and the gains from

acquisitions. Journal of Financial Economics, 73(2), 201-228.

Morck, R., & Yeung, B. (1992). Internalization: an event study test. Journal of International Economics,

33(1), 41-56.

Patell, J. M. (1976). Corporate forecasts of earnings per share and stock price behavior: Empirical test. Journal of Accounting Research, 14(2), 246-276.

Rajan, R., Servaes, H., & Zingales, L. (2000). The cost of diversity: The diversification discount and inefficient investment. The journal of Finance,55(1), 35-80.

Scharfstein, D. S., & Stein, J. C. (2000). The dark side of internal capital markets: Divisional rent‐ seeking and inefficient investment. The Journal of Finance, 55(6), 2537-2564.

Schoar, A. (2002). Effects of corporate diversification on productivity. The Journal of Finance, 57(6), 2379-2403.

Shimizu, K., Hitt, M. A., Vaidyanath, D., & Pisano, V. (2004). Theoretical foundations of cross-border mergers and acquisitions: A review of current research and recommendations for the future.

Journal of International Management, 10(3), 307-353.

Stein, J. C. (1997). Internal capital markets and the competition for corporate resources. The Journal of

Finance, 52(1), 111-133.

Travlos, N. G. (1987). Corporate takeover bids, methods of payment, and bidding firms' stock returns. The Journal of Finance, 42(4), 943-963.

Vermeulen, F., & Barkema, H. (2001). Learning through acquisitions. Academy of Management

(25)

Zaheer, S. (1995). Overcoming the liability of foreignness. Academy of Management Journal, 38(2), 341-363.

(26)

APPENDIX Data and Explanation

 Cumulative abnormal return graph:

(27)

 Traditional (normal) standard deviation formula: 𝑆𝐷 = √∑ (𝑋𝑖− 𝑋̅) 2 𝑛 𝑖=1 𝑛 where:

𝑋̅: the mean of all values in the data set 𝑛: the number of observations

Abnormal returns to U.S. bidding firms of M&A transactions from 2010 to 2014 : the impact of border, industry and relative firm size