1
Amsterdam Business School
Payment methods and the long-term performance in
cross-border mergers and acquisitions
Master Thesis
Student Name: Jia Lai
Student Number: 11086750
Supervisor: Dr. Vladimir Vladimirov
Program: MSc Business Economics – Finance track
07/2016
2
Statement of Originality
This document is written by Student Jia Lai 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.
3
Acknowledgment
Firstly, I would like to express my sincere gratitude to my supervisor Dr. Vladimir Vladimirov for the continuous support of my Master thesis, for his patience, motivation, and immense knowledge. His guidance helped me in all the time of research and writing of this thesis.
My sincere thanks also goes to my friends in Amsterdam who are always encouraging and inspiring me.
Last but not the least, I would like to thank my parents and to my other families for supporting me spiritually throughout academic studies and even my whole life in general.
4 Abstract
This article studies the effects of payment methods on the long-term performance of US acquirers after cross-border mergers and acquisitions in emerging countries and developed countries. Using the event study methodology, this article investigates whether bidder firms acquiring targets from emerging markets have a significantly higher three-year buy-and-hold average abnormal returns than those acquiring developed markets targets. However, no significant difference is found. Cross-sectional regression detects whether all-stock payment will result in a better long-term performance for bidders who acquire targets from emerging and developed markets respectively. The study finds that bidder firms acquiring targets from emerging markets have a significantly positive three-year buy-and-hold abnormal returns when bidders use mixed payment of cash and stock; bidders will have negative three-year buy-and-hold abnormal returns when bidders use all-stock payment.
5 Table of contents: Statement of originality Acknowledgment Abstract 1. Introduction···6 2. Literature review···9
2.1 Long-term performance of M&A···9
2.2 Review of the effects of payment methods on performance of M&A activities··13
2.3 M&A involving emerging markets···14
3. Hypothesis and methodology···15
3.1 Hypothesis···15
3.2 Methodology···16
4. Sample and data···23
4.1 Databases···23
4.2 Sample selection criteria···24
4.3 Summary statistics···25
5. Empirical results and discussion···29
5.1 Event study results···29
5.2 Cross-sectional regression results···30
5.3 Robustness tests···33
6. Conclusion···37
6 1. Introduction
According to KPMG (2016), by the end of 2015, $4.7 trillion of global mergers & acquisitions deals were signed in the US, outnumbering 2007’s previous record for total deal values. EY (2016) reports that 50% of companies surveyed expect to actively pursue acquisitions in the next 12 months, and among these companies, 74% are considering cross-border investments. Uneven global economic growth in recent years prompts a lot of companies to seek investment opportunities outside their home countries, and the investment targets have been from not only developed markets with mature economies, but also have included many emerging countries with less mature markets. These, however, have a great potential for future economic growth. For instance, the Chinese government is rebalancing China’s economy and planning to double the gross domestic product (GDP) and per-capital income by 2020; India, whose GDP growth is above 7%, desires to boost private and inbound investment, taking advantage of the low labor-unit costs across international markets.
This thesis mainly studies the US cross-border mergers and acquisitions during the period 2000-2012, involving both emerging market and developed market targets in the research. The article aims to answer the questions below:
(I)Do bidder firms acquiring targets from emerging markets outperform those acquiring targets from developed markets?
(II)Do bidder firms with all-stock payment have a better long-term post-performance when they acquire targets from emerging countries and developed countries respectively?
The first research question is about a naive comparison between deals of acquiring firms from emerging markets and those from developed markets, while the second question will further detect how different payment methods influence post-performance of an M&A deal involving these two markets respectively. This thesis hypothesizes that there is no difference between bidders acquiring targets from emerging markets or those acquiring targets from developed markets; in terms of the second question, this thesis hypothesizes that acquiring targets from emerging markets using all-stock payment allows bidders to have a positive long-term post-performance.
7
The result of event study reports that bidder firms acquiring targets from emerging markets do not have a statistically significantly better long-term post-performance than bidders acquiring developed market targets. This result is consistent with the first hypothesis. As for the regression results, all-stock-payment deals are reported to have a 23.50 percent lower performance, which is consistent with the empirical results by Loughran and Vijn (1997). However, further regression shows a result that is completely contrary to the second hypothesis. The regression finds that bidders acquiring firms from emerging markets have 15.53 percent lower three-year average abnormal returns when the deal uses an all-stock payment method, while bidders have 9.41 percent higher three-year average abnormal returns when the deal uses a mixed payment of cash and stock. These regression results also suggest that emerging market targets deals make the negative effects of allstock payment weaker, from -23.50 percentile to -15.53 percentile.
The method of financing which a bidder firm uses in order to acquire the target firm is known as the payment method of M&A. A target firm can be paid in certain complex ways, such as through using cash, stock, debt payment and financial derivatives. Many papers study how different payment methods affect the performance of a firm. Loughran and Vijh (1997) investigate a five-year performance after acquisitions, and find that firms which use stock payment will make significantly negative excess returns of -25 percent while firms that use cash payment will make significantly positive excess returns of 61.7 percent. Netter, et al.(2011) study a large sample of M&As from 1992 to 2009, and argue against the finding that negative acquirer returns are associated with deals that pay through stock claiming that this is not universally true. More related studies will be presented in section 2.
The targets often analyzed by researchers generally come from developed countries. Yet, different markets have their own characteristics (in culture, regulation, politics, etc.) that play an important role in the post-performance of M&A. Bruner et al. (2002) argue that firm valuation will be affected by the characteristic differences between emerging countries and developed countries, such as accounting, transparency, liquidity, corruption, violability, governance, taxes, and transaction costs. Hence, it is hard to conclude that the empirical results from studying developed countries will still matter in emerging markets. From this perspective, it is worthwhile to investigate how
8
the effects of the payment method on the long-term post-performance of the acquirers differ between developed and emerging countries.
This article will contribute to and complement current literature by not only focusing on developed countries, but importantly also involving emerging markets in the research. Because emerging markets have various characteristics that are very different from those in developed countries, it is worthwhile to study whether the conclusions reached for developed countries will still matter when it comes to a very distinctive economic situation. To solve the two research questions, this article will evaluate the post-performance after M&A and the effects of payment methods, and look at whether there are different results for emerging and developed markets. If the distinctions between these two markets do exist, it is interesting to do further empirical analyses to explain which factors lead to the distinctions. In addition, addressing these two questions will help developed market bidder firms make a more sensible decision on the strategies of cross-border mergers and acquisitions.
In order to detect the long-term performance and the effects of payment methods, this article applies two methods: event study method and cross-sectional regression method. Event study has been widely used for investigating how an economic event has an effect on the firm value. Ritter (1991) evaluates the long-run performance of initial public offerings measured by two types of returns: cumulative average adjusted returns (CAR) calculated with monthly portfolio rebalancing, and 3-year buy-and-hold returns for both the IPOs and a set of matching firms. In order to detect the performance of the firms in dividend initiations and omissions, Michaely, Thaler, and Womack (1995) apply the event study methodology, and calculate the returns from a buy-and-hold strategy. Ikenberry, Lakonishok, and Vermaelen (1995) examine long-term firm performance following open market share repurchase announcements, 1980-1990, using the average abnormal four-year buy-and-hold return. Besides, the economic event could be something which crucially changes the post-performance of a firm. Firm value can be frequently measured by stock returns on a daily, monthly or annual basis, but a number of papers also examine the post operating performance which is measured by operating income, cash flow ratio, etc.
A cross-sectional regression method is also used to analyze the research questions. In contrast to time-series regression and panel data regression which are studied over
9
multiple periods, cross-sectional regression is an econometric study method in which the independent variables and the dependent variables are studied within one period. In this article, independent variables are dummy variables that will not change over time, and the dependent variable is 3-year buy-and-hold abnormal returns. Section 3 describes further details about these two methods.
The article is organized as follows. In section 2 the related literature is reviewed. Section 3 illustrates the hypotheses of the article, and describes the methodologies the article will use. Immediately following that, section 4 aims to demonstrate sample selection criteria and summary statistics of the sample. In section 5, empirical results will be presented and discussed; to make results more plausible, robustness checks will be added after the presentation of empirical results. Based on statistics and empirical results, section 6 answers the questions mentioned in section 1 and discusses some possible further studies.
2. Literature
2.1 Long-term performance of M&A 2.1.1 Overview of M&A
Looking into the basic concepts and history of mergers and acquisitions, this section demonstrates how cross-border mergers and acquisitions become increasingly common and mature all over the world, and explains concepts of M&A which should be made clear before the research on this field is discussed. The overview of M&A helps the readers get to know some important basic concepts of M&A, and thus realize why this research is increasingly becoming necessary to study.
The term mergers and acquisitions means the process of consolidation of companies or assets. A merger means that two companies are combined to become one company, while an acquisition means that a company purchases another company but no new companies are formed. From the perspective of the relationship between two companies, mergers and acquisitions have three main types: horizontal, vertical and conglomerate. Horizontal mergers mean that two companies with similar business in the same industry merge and form a new company. This merger type aims at
10
establishing a large firm in order to gain a higher market share in the markets. Vertical mergers refer to the integration of two companies within the same supply chain. Because a company is the upstream or downstream of another merged company, the new company has the internal gain which mainly refers to a lower cost of production or transaction than that of original companies. Conglomerate integration takes place when two companies with unrelated businesses merge. These two companies are from different industries and supply chains, and the merge aims to enter a new business. Starting a new business by a firm itself is costly and takes a long time to develop, while conglomerate integration allows the firm to take advantage of the resources in the merged firm which has already constructed a mature business system in the industry.
There are mainly six merger waves in the US history where cross-border mergers and acquisitions became common and matured increasingly. The first wave, from 1895 to 1905, is known as “the Great Merger Movement”. This wave was mainly horizontal mergers. Small firms with little market share consolidated with similar firms to establish larger firms. World War I ended the first wave. The second wave was mainly a vertical wave which ended in 1929 due to financial crisis. The third wave, between 1955-1970, witnessed numerous companies entering the new markets through mergers and it was known as conglomerate wave. The era of 1974 to 1989, was an era when a number of firms conducted hostile takeovers with high leverage. This high leverage provided an amount of funds for firms to expand their businesses, but also ended the wave when the banks could not provide any liquidity for the transactions anymore. During the periods 1992-2000, the fifth wave saw a great number of cross-border mergers and most of these deals were under friendly terms. Under the circumstance of globalization, these mergers were mostly strategic and aimed at participating in a worldwide business. However, the internet bubble in 2000 burst and ended this great wave. The sixth wave also centered on globalization but was destroyed by the global financial crisis in 2008.
Mergers and acquisitions facilitate the readjustment of economic structures. Specifically speaking, a firm’s structure will be changed after the firm takes over another firm from different industries or different supply chains. A firm will be expanded after it acquires a target in the same industry and supply chain. Actually,
11
firms take over other firms out of a variety of reasons. The motivations include economy of scale, economy of scope, cross-selling, synergy, taxation, experts, etc.
2.1.2 Review of cross-border M&A
M&A include domestic deals and cross-border deals. From the five waves of M&A already discussed, these activities mainly took place in the home country due to the undeveloped Internet, transportation and businesses in the past. With the advancement of science, technology and economy, the markets can become more mature and the business becomes more diverse. These factors facilitate the fast development of cross-border mergers and acquisition.
Cross-border takeovers refer to mergers and acquisitions across different countries. Increasing globalization provides numerous opportunities for business cooperation around the world. In order to expand business, many firms have their subsidiaries in different countries, taking advantage of the cheaper resources and broader markets in the local countries. Starting up a subsidiary abroad is undoubtedly costly and needs a relatively long time to develop, but acquisition of a local firm allows the acquirer to utilize the resources of the target firm, getting access to the local markets more rapidly and easily and thus working more efficiently.
Scholars mainly study the affects of successful cross-border takeovers and how the performance will be affected by these takeovers. Quah and Young (2005) posit that in cross-border acquisitions, the management of both cultural and organizational integrations (and their interactions) plays a significant role in making the acquisitions successful. Moeller and Schlingemann (2005) studied 4430 acquisitions during the period 1985 to 1995 from the perspective of US acquirers. They found that US firms acquiring cross-border targets have roughly 1% lower announcement stock returns and lower changes in operating performance. Also, they argue that the degree of economic restrictiveness in target countries is negatively related to the returns of acquirer. This thesis also investigates the long-term performance of cross-border mergers and acquisitions, but more specifically and differently, this thesis involves the deals of emerging market targets as well, and studies the different post-performance between emerging markets and developed markets targets.
12
Some scholars also investigate the factors that play an essential role in post-performance of cross-border takeovers. Bris et al. (2008) demonstrate that Tobin’s Q of an industry rises when the targets within that industry are acquired by a bidder from countries with better shareholder protection and better accounting standards. Starks and Wei (2013) examine the effect of corporate governance on firm value, and find that acquirers’ abnormal returns at the merger announcement are increasing in the quality of corporate governance for stock offers. Also, they report that foreign acquirers from countries with better corporate governance prefer to pay by stock. In this thesis, payment method is viewed as a critical factor that influences the long-term post-performance of cross-border M&A.
2.1.3 Review of long-term performance of M&A activities
Long-term performance of a company can be measured by a variety of variables, such as stock returns, cash flow to total assets and some other financial ratios. Some variables indicate the performance of a company from a general view, such as stock returns which embody a firm’s market value. Some variables, such as cash flow to total assets, are often viewed as a ratio measuring the operating performance of a company. In this article, monthly stock returns are chosen to measure a firms performance.
Former studies on the long-term post-performance after the mergers and acquisitions present inconsistent results. In the US market, Heron and Lie (2002) find that the long-term operating performance is improved after acquisitions, but Moeller & Schlingemann (2005) report no improvement in performance. In the Canadian markets, Dutta, Saadi and Zhu (2013) focus on 1300 completed deals by Canadian acquirers between 1993 and 2002 to examine the effects of payment methods in the cross-border M&A deals, and investigate the long-term operating performance of cash- and stock-financed deals, but they do not find a significant difference. In the UK markets, Powell and Stark (2005) observe that the performance of the acquirers slightly increases, while in terms of the continental Europe, Gugler, Mueller, Yurtoglu, and Zulehner (2003) find no evidence that the post-acquisition profit improves at all. As for the Australian markets, Sharma and Ho (2002) show that performance has no significant improvement after acquisition. In the Asian markets, Rahman and
13
Limmack (2004) report significant improvements in operating performance for Malaysian acquirers.
These inconsistent results raise many questions such as: why do different papers present different results? Do different markets or different empirical methods matter to the different results? These questions stimulate interest into studying the performance of different markets. This article divides the targets of M&A into emerging market targets and developed market targets since the characteristics of these two markets are very distinctive in their culture, politics, economy, markets, etc.
2.2 Review of the effects of payment methods on performance of M&A activities
Different payment methods will have different effects. For example, cash payments require a firm to have an amount of cash from internal or external financing. Internal financing means the use of cash holdings in the bidder firm, while external financing refers to getting funds from loans, such as through a bank loan. Both of these two cash payments will pose a great financial pressure on a bidder firm, and thus the firm’s future operation and investment will be influenced under the risk of a low net cash inflow in the future. Additionally, stock payment seems to be a desirable method for a bidder because it will not cause a great pressure on a firm’s future cash flow. However, a lot of studies demonstrate that stock-financed deals have their own drawbacks. Myers and Majluf (1984) argue that asymmetric information could be revealed by the all-stock payment method, which suggests a relatively overvalued price for the bidder’s stock and a relatively undervalued price for the target’s stock. According to this asymmetric signaling, investors will adjust overvalued prices when the bidder chooses stock payment.
There are some articles studying the motivations for acquirers to choose different payment methods. Myers and Majluf (1984) argue that companies prefer cash payment to stock payment and prefer internal financing to external financing, because asymmetric signaling suggests an overvalued share price of a bidder when it chooses the stock payment method. This is called pecking order theory. Faccio and Masulis (2005) explore how acquirers make financing decisions using a sample of European deals over the period from 1997 to 2000. They find that corporate control incentives to pay by cash are strong when a bidder’s controlling shareholder has a voting power from 20% to 60%. When targets shareholdings are highly concentrated, bidders prefer
14
a cash payment for the deal. Actually, the choice of payment method by a company is motivated by combined factors, including tax effects, deal financial costs under asymmetric information, agency and corporate control motives, and behavioral arguments. For example, the US companies have different tax treatment when they use different payment methods. The capital gains are required to be paid immediately by the targets if they are in an all-cash purchase; however, using an all-stock method allows the taxes to be deferred. Then tax effects could become a significant factor for the takeovers. Rossi and Volpin (2004) analyze a sample of mergers and acquisitions of firms from 49 major countries, and they observe that targets in countries with low investor protection prefer cash payment due to the risk of expropriation for being minority shareholders.
With regard to the effects of payment methods on the performance of M&A, researchers get different results. Loughran and Vijh (1997) investigate a five-year performance after acquisitions, and find that firms with stock payment will make significantly negative excess returns of -25 percent while firms with cash payment will make significantly positive excess returns of 61.7 percent. These results are consistent with those found from regression results in this article, in which negative abnormal effects are also reports for all-stock payment. Savor and Lu (2009) find that stock-financed deals generate negative reactions around the announcement dates and also underperform significantly in the long-run. Netter, et al. (2011) study a large sample of M&A from 1992 to 2009, and they report that it is not a universal finding that negative acquirer returns are associated with deals where stock is a means of payment. Further, the use of stock is as frequent in the greatest value reducing deals as in the deals that create the most value. Thus, market timing by managers cannot fully explain the motivation for the use of stock.
2.3 M&A involving emerging markets
Emerging markets are increasingly drawing attention from investors all over the world. Cross-border investments in emerging markets are distinctive from those in developed markets. To invest in the former markets, firms need to take into account how emerging markets characteristics have an effect on their investments. Bruner et al. (2002) argue that firm valuation will be affected by the characteristic differences between emerging countries and developed countries, such as accounting,
15
transparency, liquidity, corruption, violability, governance, taxes, and transaction costs. For example, the authors investigate firm-level data from 46 countries, and find that firms from countries with a high level of corruption obtain a statistically significantly lower market multiples.
Some scholars study the way in which the performance of cross-border acquisitions will be affected when targets are from emerging markets. Narayan and Thenmozhi (2014) use the industry-adjusted operating performance to measure acquisition gains, and find that involving emerging market firms shows a statistically significant impact on the deal characteristics on post-acquisition operating performance of cross-border takeovers.
Chari et al. (2010) detect the returns over a three-day event window of acquirers from developed countries who undertake acquisitions in emerging markets, and report that acquirers gain significantly positive abnormal returns of 1.16% on average when the acquirers have control of targets. However, when the acquisitions are not of majority stake but minority stake, the returns are not statistically significant.
3. Hypothesis and methodology 3.1 Hypothesis
The objective of this thesis is to investigate: (1) whether bidder firms acquiring targets from emerging countries will have a better long-term post-performance after M&A than bidders acquiring targets from developed countries; (2) whether bidder firms acquiring targets will have a better long-term post-performance after M&A when they pay the deals by pure stock.
Regarding the effects of payment methods on the long-term post-performance after M&A, Loughran and Vijh (1997) investigate a five-year performance after acquisitions in the US markets, and find that firms with stock payments make significantly negative excess returns of -25 percent while firms with cash payments will make significantly positive excess returns of 61.7 percent. For the relevant research on emerging markets, Narayan and Thenmozhi (2014) used the industry-adjusted operating performance to measure acquisition gains, and find that involving
16
emerging market firms shows a statistically significant impact from the deal characteristics on post-acquisition operating performance of the cross-border takeovers.
However, no past literature discusses the research questions addressed in this thesis. Hence, the hypotheses of this thesis will be based on some other suggestions. Narayan and Thenmozhi (2014) argue that value will be eroded when emerging markets bidders acquire targets from developed markets, because bidders from emerging markets generally have limited experience in dealing with cross-border mergers and acquisitions. They also report that developed market firms acquiring emerging market firms show a 50 percent chance of value of creation for both acquirers and targets. It can be explained that investment in the emerging markets by developed market acquirers is full of uncertainties, and these factors make a 50 percentile possibility of a better performance after M&A. The study result suggests that the possibility of showing a better long-term performance after acquiring firms from emerging markets is 50%.
Therefore, the first hypothesis in this thesis is as follows:
Hypothesis I: There will be no difference for US bidders’ long-term
post-performance between acquiring firms from emerging countries or from developed countries.
With regards to the studies on the differences between emerging markets and developed markets, Rossi and Volpin (2004) find that in a country with low investor protection, target shareholders tend to choose cash payment instead of equity of the bidders in the mergers and acquisitions because of the risk of expropriation 1for being minority shareholders. The minority shareholders are more likely to be expropriated when they have fewer rights, and furthermore, lower shareholder protection prompts relatively less effective corporate control but higher private benefits of control. Their research results suggest that all-stock payments for targets from emerging markets with lower investor protection will be more beneficial to the bidder firms of developed countries. Accordingly, the second hypothesis is as follows:
1 In a country with low shareholder protection, there are large private benefits of control(Nenova, 2003; Dyck and Zingales,
17
Hypothesis II: The bidder firms acquiring the targets from emerging markets with
all-stock payment will have a better long-term performance.
3.2 Methodology 3.2.1 Event study
The paper uses the event study methodology (Mackinlay,1997; Barber and Lyon, 1997; Kothari and Warner, 2004) in order to investigate the long-term effects of M&A. The event study method is often used to examine the behavior of firms’ stock prices around corporate events, but much literature also examines other variables instead of stock prices, such as the operating performance (e.g., Barber and Lyon, 1996).
Long-term event study research can be derived from the late seventies and early eighties when the efficient markets hypothesis was considered suspect by a lot of researchers. Since these scholars were skeptical of the hypothesis, market inefficiency theory was developed. Hence, short-horizon event studies may not fully capture the full valuation impact of certain events.
This section will illustrate the process of event study, including three parts: identifying the event, measuring normal stock returns, and measuring long-run abnormal stock returns.
3.2.1.1 Identifying the event
The event could be any economic event that could influence the change of a firm’s value. In this article, the event will be cross-border mergers and acquisitions whose bidder firms are from the US during the period 2000-2012. The targets could be from emerging countries or developed countries (excluding the US). The article considers the event date (t=0) to be the announcement date of M&A deals. The event window in this article will be [0, +3] where +3 means 3 years after the announcement date of the event.
18
For long-run event studies, buy-and-hold returns of the event firms are always subtracted from the buy-and-hold returns of plausible benchmark returns (Barber and Lyon, 1996). This can be referred to as Equation Ⅱ. The benchmark return, or normal return, can be defined as a characteristics-matched portfolio. For instance, market value, market-to-book-ratio, and industry are typically characteristics and are used for matching control firms in this thesis.
Detecting normal stock returns for the long-term event study is always challenging and it is always hard to find a suitable benchmark. For the short-horizon tests, risk adjustment is effective and typically unimportant. The error is very small in calculating the abnormal returns when there are errors in a short-term test. However, normal returns calculation in a term test will be very different. Because long-term event studies are always detecting events over one year, even a small error in risk adjustment will result in an economically large difference in abnormal returns. Furthermore, there are no researchers obtaining a “most correct” model for normal returns, the precision of the results of abnormal returns will be very dependent on the choice of normal returns model. In the past years, a lot of researchers have worked on improving the model precision, and the works of Barber and Lyon (1996) are widely used in many papers.
Barber and Lyon (1996) test different methods to detect long-term performance of firms after merger and acquisitions, and find evidence that the well specified test statistics are yielded when event study chooses buy-and-hold abnormal returns with matching firm method to examine normal returns. This method can avoid three biases including new listing bias, rebalancing bias and skewness bias.
New listing bias arises when the long-term performance event study applies a reference portfolio, such as market index, in order to detect normal returns, because the firms constituting the market index always change when new firms are listed and some firms are unlisted in the long run. This change will lead to biases of normal returns. Rebalancing bias means that because the returns of sample firms are compounded without rebalancing, but that of reference portfolio are compounded assuming periodic rebalancing, the normal returns calculation will be biased. Skewness bias arises because long-term abnormal returns are positively skewed.
19
Out of three biases mentioned above, the article will choose the matching firms method to measure the normal returns. Referred to by Barber and Lyon (1997), the article matches on both firm size (defined as market value) and market-to-book ratio. In the first place, all firms with a market value of equity between 70% and 130% of the market value of the sample firm will be identified. Secondly, from this set of firms, the firm with the market-to-book ratio closest to that of sample firms will be chosen. Additionally, the article will have robustness checks (section 5.3) in which control firms will be identified through another benchmark: firm size and industry. This method is also mentioned and motivated by Barber and Lyon (1997). The method will match on the closest firm size in the same industry. “The same industry” here means the same two-digit SIC (Standard Industrial Classification) code. SIC code classifies industries by four-digit code. For instance, SIC code 2000-3999 signifies manufacturing, and 5200-5999 signifies retail trade. Michaely, Thaler, and Womack (1995) also calculate the returns of firms after dividend initiations and omissions using a matching firm in the same industry (two-digit SIC code).
3.2.1.3 Measure long-run abnormal stock returns
After identifying the event and measuring the normal returns, this part illustrates how the article calculates long-run abnormal stock returns.
Mackinlay (1997) defines abnormal returns of a firm after a specific event as follows.
Equation Ⅰ：
= − ( + )
, where R is equal to the actual return of the security i and R is the return of the market portfolio. This equation illustrates the very basic concept of “abnormal returns” which is always defined as the difference between returns of the sample firms and returns of the benchmark firms. This thesis, nevertheless, uses a different set of abnormal returns from the returns in Equation Ⅰ. It is called buy-and-hold abnormal returns.
Barber and Lyon (1997) argue that while CARs are desirable in detecting short-term abnormal returns, it will lead to measurement bias in long-term event studies. Instead,
20
BHARs are favorable when a long-term performance is measured. A T-month BHARs for an event firm j is defined as (Equation Ⅱ):
= 1 + R − (1 + R )
While R in the equation of ARs could be a reference portfolio, it should not be in
the equation of BHARs because it will result in the three biases discussed earlier. Instead, to avoid the three biases, matching firms method is often used to measure normal returns,
To test the null hypothesis that the mean buy-and-hold abnormal returns are zero for the sample, the article employs the parametric test statistics mentioned in Barber and Lyon (1997).
= BHAR
σ(BHAR ) × 1
√n
Here BHAR is the sample average, and σ(BHAR ) is the cross-sectional sample’s standard deviation of abnormal returns for sample firms. Because the sample firms in the article are drawn randomly from a normal distribution, this test statistics is a Student’s t-distribution under the normal distribution.
3.2.2 Cross-sectional regression
The article runs a cross-sectional regression in order to analyze the second research question mentioned in section 1. The regression model is presented below and the definitions of the variables of the models are shown in the Table 1.
BHARs =
α + β × all stock payment + β × mixed payment of cash and stock + β × emerging markets + β ×
(emerging markets × all stock payment ) + β × (emerging markets × mixed payment of cash and stock ) + β × Controls + ε
21
Table 1: Description of variables of cross-sectional regression models
This table aims to define the dependent and independent variables in the regression model. The model contains variables including all-stock payment, mixed payment of cash and stock, emerging markets and control variables. The dummy variables will be expressed as value “1” or “0”.
Variables Definition
BHARs buy-and-hold abnormal returns of stock i over the event period All-stock payment 1 represents pure stock payment, and 0 represents others
Mixed payment of cash and stock 1 represents mixed payment of cash and stock, and 0 represents others Emerging markets 1 represents targets from emerging markets, and 0 represents those from
developed markets Controls control variables
The main coefficients of interests are β and β for the model. β and β capture the magnitudes of the effects of all-stock and mixed payment methods on the long-term performance respectively. β aims at double checking whether bidder firms acquiring targets from emerging markets will have a significantly better post-performance. β and β are the coefficients of the two interactions, and they aim to investigate whether bidder firms will have a statistically significantly better long-term performance when they acquire targets from emerging countries with a pure stock payment or a mixed payment of cash and stock.
Out of the heterogeneity across companies, the model will involve a set of control variables including firm size, market-to-book ratio, industry and deal attitude.
Firm size
Firm size is defined in this article as the market value of a firm. According to research by Fama and French (1995), small firms, on average, have lower earnings scaled by book value of equity than larger firms. Accordingly, many studies add this variable to their models to control the companies of different sizes. For example, Erel, Liao, and Weisbach (2012) study the determinants of cross-border mergers and acquisitions and take into account the logarithm of firm size.
22
Market-to-book ratio is referred to the ratio of the market value of a firm to its book value. Market value of a firm is defined as the market capitalization and is always determined by the stock market. Market-to-book ratio is usually used to identify overvalued and undervalued securities. Fama and French (1995) observe that high book-to-market ratio has persistent poor earnings. Hence, it is always used as a control variable in the empirical study. For instance, Chaney, Sraer, and Thesmar (2012) control market-to-book ratio in their panel-data regression models when they study the way in which real estate shocks affect corporate investment. Loughran and Ritter (1995) study the long-term performance of initial public offerings and seasoned equity offerings and also involve market-to-book ratio.
Industry
In the regression model, industries are expressed as Standard Industrial Classification (SIC) code and regarded as dummy variables. SIC code consists of four digits, and classifies the industries as: (1)Agriculture, Forestry and Fishing (0100-0999); (2)Mining (1000-1499); (3)Construction (1500-1799); (4)Manufacturing (2000-3999); (5)Transportation, Communication, Electric, Gas and Sanitary service(4000-4999); (6)Wholesale trade (5000-5199); (7)Retail trade (5200-5999); (8)Finance, Insurance, and Real estate (6000-6799); (9)Services (7000-8999); (10)Public Administration (9100-9729). Of the four digits, the former two digits represent the major industry sector to which a business belongs. Therefore, it is reasonable and sensible to be used in order to express industry dummies.
Deal attitude
For the deal of a takeover, there is a terminology named deal attitude which describes whether a bidder company obtains the agreement of the target company when this bidder is going to take over the target company. If the board of directors in the target company agrees, this deal will be expressed as “friendly takeover”; if the board does not agree, it will be expressed as “hostile takeover”. Other types of deal attitudes, such as neutral, are rare but do exist, so the article regards these rare types as “other takeovers”. Therefore, deal attitudes in the regression models include “Friendly”, “Hostile”, and “Others”.
23 4. Data and descriptive statistics 4.1 Databases
To collect data for empirical analysis, the article uses the databases including: (1) Thomson one database; (2) CUSIP database; (3) CRSP U.S. Stock database; (4) Compustat North American database.
Thomson ONE database brings together company information from various sources that are produced by Thomson Reuters, including company financial filing, deals data (M&A etc.), equity events (IPOs etc.), share ownership data and private equity data. Samples of mergers and acquisitions deals comes from Thomson one M&A database. CUSIP numbers and standardized descriptions are used by virtually all sectors of the financial industry, and are critical for the accurate and efficient clearance and settlement of securities and other financial instruments as well as back-office processing. The CUSIP number is made up of nine characters: a base number of six characters known as the issuer number, (the 4th, 5th and/or 6th position may be alpha or numeric) and a two character suffix (either numeric or alphabetic or both) known as the issue number. The ninth character is a check digit. Each company in CRSP or Compustat database will have a unique CUSIP number for identification. In order to retrieve the data from CRSP and Compustat, the raw data of deals collected from Thomson one should be given the CUSIP number, and this process can be done in CUSIP database of Wharton WRDS.
The CRSP U.S. Stock database contains end-of-day and month-end prices on primary listings for the NYSE, NYSE MKT, NASDAQ, and Arca exchanges, along with basic market indices. CRSP U.S. Stock database provides holding period stock returns for the acquiring firms and matching firms the article needs.
Compustat North American is a database of US and Canadian fundamental and market information on active and inactive publicly held companies. The article retrieves the data of market value and market-to-book ratio from this database.
In order to do event study and run the regression, this article applies the data analysis tools: Stata and Microsoft Excel.
24 4.2 Sample selection criteria
Samples of mergers and acquisitions comes from Thomson one M&A database. The sample includes mergers and acquisitions from 2000-2012 whose acquirers are all US companies in order to keep constant the legal and economic environments of the acquirer. To obtain as many cross-border deals involving developed and emerging targets as possible, the year starts from 2000 when globalization of the economy became increasingly common and mature, and a great number of companies expanded their businesses into other countries, and more companies in the developed markets began to start up their subsidiaries in the emerging markets. The sample year ends in 2012 because the research studies the three-year buy-and-hold abnormal returns, and the returns of the deals in 2012 can be detected until 2015.
In the first place, the raw data of M&A deals are retrieved from the Thomson one database. Initially, there were 24,039 completed cross-border deals during the period 2000-2012. The article needs to clean and select the data before the empirical analysis. The sample selection criteria are listed as follows:
(I)The deals are completed. (II)The deals are all cross-border.
(III)Percentage of shares owned after transaction should be over 50 percent.2
(IV)Acquirers have only one deal between 2000 and 2012. 3
(V)The relevant data of acquirers should be able to be retrieved from CRSP, CUSIP and Compustat databases.
(VI)The payment methods only include pure cash, pure stock and mixed payment of cash and stock.
After cleaning the raw data, there remains 10,780 deals. As mentioned in part 4.1, the data collected from Thomson one database should match the information in CRSP and Compustat databases through CUSIP number. Hence, the next step is to give a unique 8-digit CUSIP number to each company, and retrieve the data of the monthly holding
2 This criterion intends to focus on mergers and acquisitions of majority interests. 3
25
period stock returns from CRSP, and market value, book value and market-to-book ratio from Compustat. Since the article only studies the companies listed on the NYSE, NYSE MKT, NASDAQ, and Arca exchanges, some companies that are not listed on these exchanges will be dropped. After dropping these data, there remain 1,243 deals which are suitable for the following descriptive statistics, event study and cross-sectional regression analysis.
4.3 Summary statistics
This section presents the summary statistics for the sample of merger and acquisitions deals after selecting. Table 2 reports the classification of developed countries and emerging countries. Table 3 is the descriptive statistics for the sample’s financial data, such as book value of a firm, market value of a firm, and market-to-book ratio. Table 4 shows the descriptive statistics of dummy variables, such as deal attitudes and payment methods.
Because the article investigates the targets across different markets which are made up of emerging markets and developed markets, it is necessary to make clear what the emerging and developed countries respectively are and which countries belong to each group. The article refers to IMF (2016) which classifies 39 economies as “advanced economies”, and the remaining economies are classified as “emerging economies and developing economies”. The classification is based on IMF’s analytical criteria and financial criteria, such as GDP. As shown in Table 2, 31 sample target countries are developed economies while 39 sample target countries are emerging economies. Although there are more emerging countries in the sample, many more deals take place between two developed countries as shown in the following statistics.
In the Table 3, descriptive statistics of target nations are summarized. According to this table’s result, US bidder firms are more likely to acquire target firms from Canada, United Kingdom, Australia and some European countries such as Germany and France. These targets from developed countries amount to a high percentage in US cross-border M&A activities. In terms of emerging economies, China, India and Brazil contribute most M&A deals.
26
Table 2: Classification of “advanced economies” and “emerging and developing economies”
This table reports the classification of advanced economies and emerging economies (referred to emerging and developing economies), according to the classification standard by International Monetary Fund (IMF). IMF (2016) reports 39 economies as “advanced economies” which have a powerful economy, a mature market and a sound regulation. In this table, there are two variables, “Target nation” and “Emerging economies”. “Target nation” is the name of a certain country, while “Emerging economies” is dummy variable whose value is 1 when it is an emerging economy and otherwise its value is 0. This table only lists countries that are in the sample of the data in this article, other countries that are not in the sample are delisted from this table.
Target nation Emerging economies Target nation Emerging economies Target nation Emerging economies
Argentina 1 Hong Kong 0 Puerto Rico 0
Australia 0 Hungary 0 Qatar 1
Austria 0 Iceland 0 Romania 1
Bahamas 1 India 1 Russian Fed 1
Belgium 0 Indonesia 1 Saudi Arabia 1
Bermuda 1 Ireland-Rep 1 Serbia 1
Brazil 1 Israel 0 Singapore 0
British Virgin 1 Italy 0 Slovak Rep 1
Canada 0 Japan 0 South Africa 1
Cayman Islands 1 Lithuania 0 South Korea 0
Chile 1 Luxembourg 0 Spain 0
China 1 Malaysia 1 Surinam 1
Colombia 1 Malta 0 Sweden 0
Costa Rica 1 Mexico 1 Switzerland 0
Czech Republic 0 Monaco 1 Taiwan 0
Denmark 0 Neth Antilles 1 Thailand 1
Egypt 1 Netherlands 0 Tunisia 1
El Salvador 1 New Zealand 0 Turkey 1
Estonia 1 Norway 0 United Kingdom 0
Finland 0 Pakistan 1 Uruguay 1
France 0 Peru 1 Utd Arab Em 1
Germany 0 Philippines 1 Vietnam 1
Greece 0 Poland 1
Honduras 1 Portugal 0
Table 2 and Table 3 mainly focus on demonstrating the classification of the developed economies and the emerging economies, and summarizing the statistics of target nations of the sample. Table 4 gives general descriptive statistics for firm-level variables, including total book value, total market value and market-to-book ratio. As mentioned in section 3, firm size is defined as the total market value of a firm, and these three variables are collected in order to select matching firms in the event study,
27
Table 3: Descriptive statistics of target nations of the sample
This table presents the frequency and percentage of target nations for the sample of the M&A deals. The countries are sorted by the first letter of their names. There are three variables in this table. “Target nation” signifies the name of a country. “Freq.” is short for frequency, meaning that how many deals involve the target from this country. “Percent” variable further summarizes the percentage of frequency.
Target
Nation Freq. Percent
Target
Nation Freq. Percent
Target
Nation Freq. Percent
Argentina 12 0.97 Iceland 3 0.24 Saudi Arabia 4 0.32
Australia 46 3.7 India 27 2.17 Serbia 2 0.16
Austria 7 0.56 Indonesia 1 0.08 Singapore 14 1.13 Bahamas 4 0.32 Ireland-Rep 14 1.13 Slovak Rep 1 0.08 Belgium 15 1.21 Israel 31 2.49 South Africa 11 0.88 Bermuda 6 0.48 Italy 24 1.93 South Korea 6 0.48
Brazil 43 3.46 Japan 22 1.77 Spain 12 0.97
British Virgin 2 0.16 Lithuania 2 0.16 Surinam 1 0.08 Canada 236 18.99 Luxembourg 3 0.24 Sweden 27 2.17 Cayman
Islands 1 0.08 Malaysia 3 0.24 Switzerland 28 2.25
Chile 1 0.08 Malta 1 0.08 Taiwan 9 0.72
China 38 3.06 Mexico 18 1.45 Thailand 3 0.24
Colombia 4 0.32 Monaco 1 0.08 Tunisia 1 0.08
Costa Rica 3 0.24 Neth Antilles 3 0.24 Turkey 4 0.32 Czech
Republic 8 0.64 Netherlands 34 2.74
United
Kingdom 209 16.81 Denmark 25 2.01 New Zealand 13 1.05 Uruguay 2 0.16
Egypt 1 0.08 Norway 14 1.13 Utd Arab Em 1 0.08
El Salvador 1 0.08 Pakistan 2 0.16 Vietnam 1 0.08
Estonia 1 0.08 Peru 1 0.08
Finland 8 0.64 Philippines 6 0.48 Total 1,243 100
France 60 4.83 Poland 3 0.24
Germany 96 7.72 Portugal 3 0.24 Greece 1 0.08 Puerto Rico 2 0.16
Honduras 1 0.08 Qatar 1 0.08
Hong Kong 13 1.05 Romania 5 0.4 Hungary 2 0.16 Russian Fed 10 0.8
and functioning as control variables in the cross-sectional regression. Besides, variables are winsorized at 5% and 95% in order to rule out the impacts from outliers. As can be seen from Table 4, the number of observations is 1,243 which means that 1,243 US mergers and acquisitions deals are selected through the sample selection criteria mentioned in section 4.1. The standard deviation of market value is 8,240 in millions which is very large, because the market value of the sample ranges from around 60 million to 31,780 million. The bid gap between the minimum and maximum value, nevertheless, makes the sample firms diversified in size.
28
Table 4: Descriptive Statistics
This table presents descriptive statistics for the sample of M&A deals in the US during the period 2000-2012. The data is retrieved from Compustat database. The variables include the acquirers’ book value, market value and market-to-book ratio. The table contains the information of number of observation, median value, mean value, minimum value, maximum value and standard deviation. Variables are winsorized at 5% and 95% in order to rule out the impacts from outliers
Variables Number Median Mean Min Max SD
Book value( million) 1,243 561 2,256 24.8380 16,905 4,187 Market value( million) 1,243 1,227 4,881 60.6294 31,780 8,240 Market-to-book ratio 1,243 0.4754 0.5449 0.0998 1.3349 0.3300
Table 5 shows the descriptive statistics of the dummy variables. The dummies in this article contain targets’ original countries (emerging countries and developed countries), deal attitudes (friendly, hostile and others), and payment methods (all-cash payment, all-stock payment and mix of cash and stock). From the table, the number of
Table 5: Descriptive Statistics-Percentage
This table illustrates variables including targets’ original countries, deal attitude, and payment method. The information contains their number of observations and percentage. In the sample, targets of M&A deals are from emerging markets and developed markets; deal attitudes consist of “friendly”, “hostile”, and “others”. Payment methods include pure cash, pure stock and the mix of cash and stock.
Variables Number Percent(%)
Targets
Targets from Emerging markets 245 19.71
Targets from Developed markets 998 80.29
Total 1,243 100 Deal attitude Friendly 1,221 98.23 Hostile 10 0.08 Others 12 1.69 Total 1,243 100 Payment method all cash 234 18.83 all stock 44 3.54 Mix 965 77.63 Total 1,243 100
29
observation and percentages of these dummies are presented. As shown in the table, roughly 80.29% of US bidder firms acquired firms from developed countries, while only 19.71% of these bidder firms acquired firms from emerging countries. US bidders, it seems, prefer to invest in countries with more mature markets in order to avoid high risk in the emerging markets with unstable economies and immature markets. In addition, almost all the bidder firms’ deal attitude is friendly, while only 0.08% of the bidder firms’ deal attitude is hostile. In terms of payment method, US firms are more likely to choose mixed payment of cash and stock, and the second most common payment method is all cash.
5. Empirical results and discussion 5.1 Event study results
The article investigates the long-term post-performance of mergers and acquisitions through event study methodology. Table 6 presents the event study results for US bidder firms.
To rule out any possible impacts from outliers, the thesis winsorizes buy-and-hold abnormal returns for all deals at the 5 percentile and the 95 percentile. From the table, the BHARs are not statistically significant for acquiring either emerging targets or developed targets, although three-year buy-and-hold average abnormal returns are positive and acquiring targets from emerging markets get higher BHARs than from acquiring targets from developed markets.
Hence, the thesis cannot reject the null hypothesis that bidder firms acquiring targets from emerging markets do not have a better long-term post-performance than bidders acquiring developed markets targets.
30
Table 6: Three-Year Average Buy-and-Hold Abnormal Returns
This table describes the event study results. The event study of the article uses three-year buy-and-hold average abnormal returns to define post-performance of cross-border mergers and acquisitions of US bidder firms. The results include bidders acquiring targets from emerging markets and developed markets respectively. To detect buy-and-hold abnormal returns, benchmark is matching firms which are selected based on the methods of identifying a control firm(Barber & Lyon, 1997). According to one of the methods, matching on both firm size and market-to-book ratio, the article first identifies all firms with a market value of equity between 70% to 130% of the market value of equity of the sample firm; from the set of firms, the article selects a firm with the closest market-to-book ratio to that of the sample firm. Barber and Lyon(1997) find that this matching method yields test statistics that are well specified. The three-year average buy-and-hold abnormal returns and its t-statistics are presented in this table. Besides, buy-and-hold abnormal returns for all deals are winsorized at 5% and 95% in order to rule out the impacts from outliers.
Emerging targets Developed targets
Benchmark Control firms Control firms
Mean 5.76% 0.55%
t-statistics 0.1151 0.0106
Conclusion(5%
level) Not Significant Not Significant
Note:
* ** ***
Indicates statistical significance at 10% level. Indicates statistical significance at 5% level. Indicates statistical significance at 1% level.
5.2 Cross-sectional regression results
In this section, cross-sectional regression on the effects of the different payment methods is run and the empirical results are illustrated in the Table 7. The cross-sectional regression aims to answer the second research question. This research question is about whether all-stock-financed deals outperform the deals using other types of payment methods. Table 7 reports the empirical results of regression as follows.
BHARs =
α + β × all stock payment + β × mixed payment of cash and stock + β × emerging markets + β ×
31
(emerging markets × all stock payment ) + β × (emerging markets × mixed payment of cash and stock ) + β × Controls + ε
As shown in the regression model, β tries to answer the general question: whether all-stock-financed deals outperform the deals using other types of payment methods. β shows the effect of using mixed payments of cash and stock. As for β , the article aims to see whether bidder firms acquiring targets from emerging markets will have a significantly positive three-year buy-and-hold abnormal returns. The result of β can be regarded as a robustness check of the result of the event study shown in the part 5.1. β and β are the main coefficients of interest in the first regression model which will answer the research question 2 specifically. The estimates of these two coefficients aims to look at whether bidder firms acquiring targets from emerging countries will have significantly positive long-term buy-and-hold abnormal returns when they choose all-stock payment (for β ) or mixed payment method (for β ).
As shown in the Table 7, whereas all-stock payment has a significantly negative effect on the long-term post-performance of bidders, mixed payment shows a significantly positive effect. The negative results (-23.5 percent) for stock-financed deals are consistent with Loughran and Vijh (1997), who detected 5-year post-performance of M&A and found that stock-financed deals will result in 25 percent lower returns than other types of payment methods.
However, the results do not give a statistically significant beta value for the “Emerging markets” variable. This empirical result is consistent with the findings obtained from the event study in this article. It means that no significant difference of performance has been found between acquiring firms from emerging and developed countries.
Interactions in this model present how payment methods have an effect on bidders who acquire targets from emerging markets. When bidder firms buy the targets from emerging countries using all-stock payment method, they have a statistically significantly negative post-performance, but when they choose a mixed payment of cash and stock, they have a statistically significant positive post-performance.
32
Table 7: The impact of payment methods on the long-term performance of acquisitions
This table demonstrates the results of the first cross-sectional regression. This regression studies how payment methods influence the three-year post-performance of US bidder firms who acquire targets from the emerging markets. The dependent variable is buy-and-hold abnormal returns of stock i over three years. The independent variables include dummy variables, interactions and control variables. Dummies contain all-stock payment, mixed payment of cash and stock, emerging markets and deal attitudes( “friendly” and “hostile”). When these dummies values are “1”, it means that a bidder chooses all-stock payment, mixed payment, emerging targets, friendly attitude, or hostile attitude; otherwise their value will be “0”. Interactions are products of two dummies( emerging targets and payment method). Control variables include deal attitude, industries, bidders’ firm size, and market-to-book ratio. Firm size is defined as market value of equity of a bidder; market-to-market-to-book ratio is defined as the ratio of the market value of equity to the book value of equity of a bidder. Besides, buy-and-hold abnormal returns for all deals are winsorized at 5% and 95% in order to rule out the impacts from outliers.
Variables β and standard deviation
All-stock payment -0.2350***
(0.0147) Mixed payment of cash and stock 0.0199***
(0.0068)
Emerging markets -0.0194
(0.0147) Emerging markets × all stock payment -0.1553***
(0.0343) Emerging markets × Mixed payment of cash and stock 0.0941***
(0.0162)
Friendly -0.0577
(0.0361)
Hostile 0.2920***
(0.0421) Firm size( market value of equity) 0.0000***
(0.0000)
Market-to-book ratio -0.0114**
(0.0053)
Industry dummy Yes
Intercept 0.4023***
(0.0418)
Observations 1,243
R-squared 0.1010
Note:
* Indicates statistical significance at 10% level. ** Indicates statistical significance at 5% level. *** Indicates statistical significance at 1% level.
33
The estimated coefficients of the two interactions are -0.1553 and 0.0941 respectively, which are explained by the fact that bidder firms acquiring targets from emerging markets will achieve 15.53% lower buy-and-hold abnormal returns with all-stock payment and 9.41% higher buy-and-hold abnormal returns with mixed payment of cash and stock. Obviously the results run contrary to hypothesis Ⅱ. The negative effect is consistent with the pecking order theory by Myers and Majluf (1984) who argue that overvalued stock price of a bidder firm is revealed when this firm pays by stock for M&A deals. The effect from pecking order theory is stronger than the effect from other theories which supports the opinion that stock payment will not put a great financial pressure on a bidder firm’s cash flow, and therefore will become a signaling of better future performance. The negative results cannot support Rossi and Volpin (2004). It indicates that the majority shareholders (bidder firms with all-stock payment) cannot benefit from the risk of expropriation when the minority shareholders (target firms) in emerging markets have lower investor protection and are more likely to have fewer rights.
However, comparing the estimates of coefficients of all-stock payment (-23.50 percent) and Emerging markets×all-stock payments (-15.53 percent), the empirical results show that the effects of all-stock payment become weaker when targets are from emerging markets. From this aspect, the all-stock payment for emerging market deals indeed makes a bidder have a less worse performance.
5.3 Robustness tests
In the section 5.2, one robustness check has been done; that is, the variable “emerging markets” is inserted in the regression model to detect whether bidder firms will have a significant long-term post-performance after they acquire targets from emerging markets. The result of the regression is consistent with that of the event study. Therefore, there is no significant difference of performance between acquiring firms from emerging markets and from developed markets.
In this section, robustness tests mainly deal with the event study on other benchmarks for measuring normal returns. As can be seen from section 5.1, the article matches control firms with the market-to-book ratio and market value of a firm. In this section,
34
another benchmark will be used to measure normal returns: market value of a firm and industry. Note that using industry variables to select matching firms is similar to using an industry dummy as the control variables in the regression models discussed earlier; that is, the two-digit SIC code of industry will be used to match firms. If two firms have the same two-digit SIC code and the closest market value, then they will be matched. Each company will be matched by only a single firm and cannot be matched by itself. After matching firms, the article will give the results of event study and cross-sectional regression.
5.3.1 Empirical results on other return benchmarks
Similarly to section 5.1 and section 5.2, this part looks at the event study and cross-sectional regression on the two research questions. Differently, a new benchmark will be used to detect normal returns. This new method chooses a two-digit SIC code and the market value of a firm to match firms for the samples, while section 5.1 and 5.2 uses the benchmark: market value of a firm and market-to-book ratio. Both of these two methods are motivated by Barber and Lyon (1997) who have obtained a well specified test for the two methods. Robustness checks look at whether the new benchmark will help obtain similar results to that of section 5.1 and 5.2. Similar results will make the empirical results more plausible. The new event study results are presented in Table 8, while the new regression results are presented in Table 9.
In Table 8, robustness checks are about using a different benchmark to detect normal returns, and thus obtaining new three-year buy-and-hold abnormal returns for two kinds of bidder firms. Similarly, the average BHARs of bidders are still insignificant under the new benchmark. This result is consistent with what the article obtains in section 5.1. Therefore, bidder firms buying targets from emerging countries do not have a significantly better long-term post-performance than bidders acquiring developed market targets.
Through event study, the article illustrates that there is no significant difference between acquiring firms from emerging countries or developed countries. The event study results of robustness check are consistent with the results that the article obtains in the section 5.1. As for the regression results of robustness tests, there exist similarities and distinctions between the results of Table 7 and Table 9. As for both tables, the coefficients are significantly positive for mixed payment of cash and stock,
