Master Thesis
Degree Program: Master in Finance Chali Guo, S2442019
1
Do companies benefit from past experience when acquiring in
emerging markets? A study of listed UK acquirers in 2000-2013
Chali Guo1
Abstract
In this paper I study the influence of prior acquisition experience on acquisition performance. Drawing on behavioral learning theory by Haleblian and Finkelstein (1999), I predict that prior experience’s influence on acquisition performance may range from negative to positive, depending on the extent to which prior targets are similar to the focal target. I use the event study approach and cross-sectional regression in measuring and estimating the acquisition performance of 317 emerging-market acquisitions announced between 2000 and 2013 by listed UK acquirers. The estimation results show that the correlation between prior overall acquisition experience and subsequent acquisition performance in emerging markets is significant and negative. A further examination confirms the findings of Haleblian and Finkelstein (1999). Differentiating acquisition experience in emerging market and developed economies reveals that acquisition experiences in the target country and other emerging markets positively influences emerging-market acquisition performance, though not significantly. Moreover, acquisition experiences in the developed economies have significant negative effects on emerging-market acquisition performance. This suggests that acquisition managers may be overconfident in applying previous acquisition experiences in developed countries to emerging market acquisitions.
Keywords:
Acquisition Experience, Emerging Market, Behavior Learning Theory.JEL code: F21, G14, G34
1. Introduction
Do acquisitions create value? Numerous studies have been done in order to answer that question. Most empirical studies fail to relate acquisitions to shareholder value creation, and therefore conclude that acquisitions do not create value, and may even destroy value (Lubatkin 1983; Cakici, Hessel, and Tandon, 1996; Meschi and Metais 2006). From the acquirer’s perspective, it seems that the efforts in carrying out acquisitions are not always justified with an increase in shareholder’s wealth, but sometimes lead to negative or no return. An acquisition is often a bald move that could go wrong for many reasons. Acquirer could pick the wrong target, pay too much for the target or poorly manage the acquisition process afterwards (Haspeslagh and Jemison, 1991). For a company that is actively expanding and healthily growing, throughout its history, acquisitions may occur multiple times. Thus, it appears that if companies could generate knowledge and could learn from previous acquisitions, and
1
2
apply the knowledge appropriately to subsequent acquisitions, it is possible that such experience would lead to superior acquisition performance.
Previous studies yield various results regarding the effect of acquisition experience. Although intuitively knowledge and routine derived from prior experiences would enhance future acquisition performance (Haleblian and Finkelstein 1999), empirical findings often prove otherwise. Von Eije and Wiegerinck (2010) assume a linear relationship between acquisition experience and subsequent acquisition performance and find that more acquisitions in one specific country result in inferior acquisition performance. Kusewitt (1985) lists the number of prior acquisitions as one of seven main factors that could determine acquisition performance and finds the relationship to be significantly negative, which he explained could be a result of “corporate indigestion” rather than the inability to learn from prior experience. Hayward (2002) observes that a company can learn from previous acquisitions, but the ability to generate superior acquirer’s return depends on the nature and timing of the acquisition. The gain from experience is most pronounced if the previous acquisition is associated with a small loss. On the other hand, Fowler and Schmidt (1989) find that acquiring firms’ previous acquisition experiences significantly improve subsequent acquisition performance, from which they conclude that managers could learn how to consolidate firms successfully via past experience.
However, a more complex form of relationship has been observed. Meschi & Metais (2006) observe a curvilinear (inverted U-shaped) relationship between the acquisition performance and acquisition experience of French acquirers in US markets, where similar and low announcement returns are observed for acquirers with no acquisition experience, or to the very opposite, namely considerable acquisition experiences. To the contrary, Haleblian and Finkelstein (1999) observe that relatively inexperienced acquirers, also termed “beginners”, inappropriately derive “experience” to subsequent dissimilar acquisitions, while relatively experienced acquirer, or “experts”, properly discriminate between acquisition experiences, which leads to a U-shape relationship between acquisition experience and subsequent acquisition performance.
Although the findings from these studies fail to converge, they repeatedly point to the importance of acquisition experience and its relationship to acquisition performance. One possible explanation for the various findings is that learning relates to not only the quantity of acquisition experiences, but more importantly, the quality of the experience (Hayward 2002). How firm can adaptively learn from prior acquisition depends on the nature, timing and performance of the experience (Hayward 2002), and its similarity to focal acquisitions (Haleblian and Finkelstein 1999).
3
industry level. In this study, drawing on the study of Haleblian and Finkelstein, I examine prior experience’s influence on acquisition performance in the context of emerging-market acquisitions. Moreover, I extend their study by differentiating between acquisition experiences in emerging, developed economies and finally acquisition experience within the same target country to examine whether the similarity in terms of the development level of the target market makes a difference. The remainder of this paper is organized as follows. Section 2 provides the literature on emerging-market acquisitions, organizational learning theories and behavioral learning theories and gives my hypotheses. Section 3 describes the research methods, namely event study and cross-sectional analysis. Section 4 presents the sample characteristics. Section 5 sets out the results and Section 6 concludes the paper.
2. Literature Review and Testable Hypothesis
2.1 Learning from Acquisition Experience
If a company has acquired once, is it possible that the company can learn from prior acquisition and generate skills and experiences that lead to a superior performance in the subsequent acquisition? Intuitively our answer to this question would be yes. At first glance it seems obvious if we simply apply the “learning by doing” perspective. An organization can realize improvements in productivity through accumulated experience with the production. As organization produces more of a particular kind of product, the unit cost of this product will decrease. Such pattern, often termed “learning curve” and “learning by doing”, have been commonly presented and documented, but mainly in manufacturing industry (Epple, Argote, and Devadas, 1991). Although the precise learning curve cannot be observed directly, generally scholars describe a learning curve with diminishing marginal return of learning (Hayward, 2002; Boone, Ganeshan, and Hicks, 2008).
Outside the manufacturing industry, organizational learning behavior is much harder to predict, and studies so far often yield contradictory results. Boone, Ganeshan, and Hicks (2008) observed “learning curve” pattern in professional services firms, and that organizational learning serves as an important source of their competitive advantage, which supports organizational learning. However, Argote and Epple (1990) find that organizations exhibit considerable variation in the patterns and rates at which they learn, depending on the nature of the activity. It is questionable whether a “learning curve” would hold for strategic organizational decisions like acquisition, as it is a sophisticated and dynamic process that requires close cooperation of highly trained professionals from various disciplines such as accountants, lawyers, consultants, financial advisors, etc.
4
which could be explained by a positive learning effect. For a company to generate superior performance and to gain competitive advantage through accumulated experience, a company needs to (1) summarize and acquire knowledge from prior acquisition (2) preserve past knowledge and protect such knowledge from organizational knowledge depreciation (3) properly apply accumulated acquisition knowledge to the target acquisition. Meanwhile, companies are faced with various obstacles in bridging the gap between learning and actually benefiting from prior experiences. First, an acquisition is carried out by highly skilled professionals. The person who carried out the previous acquisitions is primarily responsible for generating knowledge from prior projects and solving current problems. If these people leave the company after a period of time, the unique intellectual property also leaves, and thus the knowledge stock and accumulated experience depreciates (Boone, Ganeshan, and Hicks, 2008). Second, inferences are particularly valuable when they are generated in a timely fashion (Hayward, 2002). If the interval between previous and current acquisition is too long, the experiences accumulated from previous acquisition may fade, and thus the firm’s ability to utilize previous experience also weakens. On the other hand, if the interval between the previous and current acquisition is too short, the knowledge and pattern may not be properly digested or have not taken root.
This part of analysis gives an overview about previous experiences’ influence when applying organizational learning and “learning by doing” perspective. However, these theories only explains the circumstances where applying experience have a positive influence on the performance of subsequent behavior, and fails to offer insight on cases where negative influence are observed (for example, von Eije and Wiegerinck, 2010; Kusewitt, 1985).
2.2 Acquiring in Emerging Markets
In the twenty-first century, the foreign direct investment (FDI) into emerging markets experienced a sharp increase from 1393 billion dollars in 2000 to 6059 billion dollars in 2012 (UNCTAD, 2013). Within this global trend, cross-border mergers and acquisitions are clearly the favorite corporate growth strategy of multinational corporations. This study is carried out in the context of emerging-market acquisition. I choose emerging-market acquisition as an appropriate context of my analysis for mainly two reasons. First, as stated above, acquisitions in emerging markets is clearly on the rise and an extensive analysis of acquisition experience and organizational learning would provide valuable inferences for future acquisitions. Second, compared to focal acquisitions, past acquisition experiences may vary from similar to dissimilar (i.e. in an emerging market or not in an emerging market respectively). This sets the ground for in-depth analysis of experiences’ influence, which may also range from positive to negative.
5
Monetary Fund’s (IMF) classification2 for both developed countries and emerging markets. 3 In 2012, the International Monetary Fund (IMF) labels 36 economies as "advanced economies" and 25 countries as "emerging economies”, among which the four largest emerging economies are Brazil, Russia, India and China, also known as the BRIC countries. Regarding the global financial crisis in 2008, emerging economies recover faster and more strongly in terms of gross domestic product and industrial production while advanced economies generally remained in a state of recession or poor growth (Didier and Schmukler, 2011). Among the developed economies who are actively engaged in foreign acquisition, UK is the largest acquiring country worldwide by 2000, accounting for 31% of the total value of all cross-border acquisitions (Conn, Cosh, and Guest, 2005). Entering the twenty-first century, UK acquirers has grown more cautious and slowed down the pace of foreign investment to observe the global M&A markets. For the past 14 years, both the number and total value of acquisitions abroad by UK companies has decreased dramatically. Meanwhile, UK, as a frequent player in the global M&A market, remained the fourth largest acquirer and the third most targeted nation (UNCTAD, 2013).
For foreign investors, emerging markets are both appealing and risky. Scholars identify the emerging markets by their rapid growth and industrialization. Meanwhile, compared to developed countries, they also display a higher level of risk and volatility of the economy. Despite the fact that cross-border acquisitions targeting developed countries often lead to negative announcement returns for acquirers, emerging-market acquisitions are usually associated with an increase in shareholder wealth. Chari, Ouimet, and Tesar (2004) find significant and positive value effects for both acquiring company and target firm in the case where an acquirer from developed countries acquires in emerging economies. This could be explained by the fact that acquirers can benefit from tax advantages captured from reductions in foreign profit taxes (Norback, Persson, and Vlachos, 2009), or that the emerging-market targets has the potential to benefit much from the transfer of technology and skills from developed-market acquirers (Soussa and Wheeler, 2006).
In examining the differences between experiences generated from those two types of acquisition I apply a process perspective proposed by Very and Schweiger (1998). They argue that numerous activities are involved in the process of making an acquisition and among them, a number of key activities determines whether acquirer’s investments can be realized, and whether the acquisition is value creating. The key activities at the initial stage involve identifying and selecting acquisition candidates as well as valuing and pricing the target, while after the transaction the key activity involves managing the integration of the acquired firm (Very and Schweiger, 1998). Accordingly, whether companies can benefit from prior acquisitions depends on the
2
Source: World Economic Outlook Update, IMF (July 2012) http://www.imf.org/external/pubs/ft/weo/2012/update/02/index.htm
3
6
acquirer’s ability to properly generate inferences and patterns from the acquisition process and whether these inferences and knowledge are applicable to subsequent acquisitions.
Experiences generalized from developed-market acquisitions may not be applicable to emerging-market acquisitions, as we can identify a number of potential risks associated exclusively to acquiring in foreign countries, and particularly emerging markets. First, in order to negotiate a proper deal, the acquirer must be able to form an accurate estimate of the target’s fundamental value. For firms from developed countries to acquire in emerging countries, the ability to form a close estimate of the target’s true value is particularly important, and is usually jeopardized as a result of information asymmetry (Very and Schweiger 1998). Second, even if the developed-market acquirers accurately value the target firm, they often find themselves in a disadvantaged position when competing with the local emerging-market bidders (Very and Schweiger 1998). When the price is already not favorable, it is not uncommon that the acquirers do not want to walk away from the opportunity to expand into emerging markets and end up paying too much. Third, acquirers generally experience larger cultural, political, legal and social differences when operating the combined company with targets from emerging markets than with domestic targets or other developed-country targets (Very, Schweiger 1998; Fiordelisi, Martelli 2011; Barkema, Bell, and Pennings 1998). These characteristics are often country specific, and make emerging-market acquisitions fundamentally different from developed-market acquisitions.
2.3 Behavioral Learning Theory
Following the fundamental differences observed in Section 2.2, as the analysis is carried out in the context of emerging-market acquisitions, I define the firm’s prior acquisitions targeting emerging markets as similar and the acquisitions targeting developed countries as dissimilar. Drawing from the behavioral learning theory proposed by Haleblian and Finkelstein (1999), I predict that prior acquisition experience’s influence on subsequent acquisition performance may range from positive to negative, depending on the similarity or dissimilarity between prior targets and focal target.
7
outcomes. As Very and Schweiger (1998) has observed, local experience accumulated by an acquirer about a target country increases the chance of subsequent acquisitions to turn out successful. They explain that when the acquirer has established operations in an emerging market, not only the company will be familiar with the regulations, culture, language and social conditions of that market, but the local staff could also provide assistance in finding local targets, analyzing, negotiating and managing the acquisition (Barkema, Bell, and Pennings 1998). Thus my first hypothesis is as follows:
Hypothesis 1: When acquiring in an emerging market, the acquirer’s performance is positively related to prior acquisition experiences in the target country.
It follows that though not targeting the same country, the present acquisition also shares more similarity with the previous emerging-market acquisitions. Thus my second hypothesis is:
Hypothesis 2: When acquiring in an emerging market, the acquirer’s performance is positively related to prior acquisition experiences in the emerging markets.
Contrary to the appropriate generalization hypothesis, Haleblian and Finkelstein (1999) believe that when the present target is highly dissimilar to previous targets, inappropriately generalizing acquisition experience lead to inferior acquisition performance. They argue that acquisitions are essentially heterogeneous, as they are carried out to serve various expansion needs (Hayward, 2002) and the targets each possess unique resources and capabilities (Shimizu et al., 2004). With inappropriate generalization, when acquirers recognize similar superficial feature, they have the tendency to apply the learned pattern to subsequent events without properly differentiate and distinguish the dissimilar structural features (Hearst and Koresko, 1968). Thus when a past event is dissimilar and irrelevant to the new event, generalizing past experience could be misleading and may lead to negative outcomes (Novick, 1988). Here I introduce the concept of acquirer “overconfidence” 4. Overconfidence is the subjective bias where individuals overestimate their knowledge and ability, or are too optimistic about the future and their ability to control it (Ackert and Deaves, 2009). It is well-established and extensively documented, for instance in a social experiment in the context of horse racing, where after receiving more information, participants’ confidence in their judgment is reliably greater than their objective accuracy and they exhibit excess optimism and illusion of control over the outcome. Dissimilar acquisition experience in this case serves to boost acquirer’s confidence in acquiring in an emerging economy while failing to offer really valuable inferences. Acquirers with more experience in acquiring developed-market targets may be over-optimistic about their ability to acquire in emerging market and be under the illusion of control over the integration process. In this case acquirers pick the target in emerging market even if it may not be the best option, and this results in more risk taking, which leads to negative outcome in acquisition. Given this, the
4
8 appropriate generalized hypothesis is as follows:
Hypothesis 3: When targeting an emerging market, acquisition performance is negatively related to prior acquisition experiences in developed countries.
3. Research Method
The approach here is to test whether differences in acquisition performance can be explained by the acquiring companies’ prior experiences. A positive and significant relationship between acquirer’s prior experiences and acquisition performance would suggest that firms could generate superior performance by learning from prior acquisitions.
3.1 Event Study
In assessing the acquisition performance, this paper uses the simple but well-established event study approach. The event study approach is commonly used by researchers who try to capture the effect of a particular economic event on the value of firms. An event study relies on the assumption that the change in the shareholder’s wealth during the event window can capture and predict the future gains and losses that would result from the event in the long run. Duso, Gugler, and Yurtoglu (2010) present empirical evidence that an event study measure has the ability to capture post-merger’s profitability. Kaplan and Weisbach (1992) also find that successful acquisitions exhibit significant higher abnormal returns around the announcement date, which indicates that the stock market response is capable of forecasting subsequent acquisition performance. These evidences support the predictive validity of event study measures and give confidence in using cumulative abnormal return as the measure of acquisition performance.
9
Given that an event is defined and the event window is identified, the abnormal return caused by the announcement can be estimated by calculating the excess returns over what would have been earned if event i did not occur. I collect the actual return from the stock market and estimate the benchmark return with the market model. The normal return is estimated from the market equation, given by:
(1)
Here is the actual stock return of acquirer i on day t. is the market return on day t. Coefficient and represents the correlation between the acquirer’s stock price and the market index. I choose an estimation window of 200 trading days (starting 220 days before the announcement day). As did von Eije and Wiegerinck (2010), the coefficients in this equation are estimated daily from 220 days to 21 days prior to the acquisition with Ordinary Least Squares Regression model. With the estimated parameters I proceed to calculate abnormal returns by subtracting the predicted normal return from the actual return for each of the days in the event window (see equation (2)). Here are assumed to be stable. Finally, cumulative abnormal returns (CARs) are calculated by summing up the abnormal returns over the event window for every firm (see equation (3)).
(2) (3)
3.2 Cross Section Analysis
The hypothesis of this study is tested within the settings of the following cross-sectional regression analysis using Ordinary Least Square (OLS). I carried out my analysis in two steps. First, I examine the influence of overall acquisition experiences by regressing acquisition performance on overall acquisition experiences as well as multiple control variables (See equation (4)). The overall effect of previous acquisition experience comprises the differed influence of acquisition experiences accumulated in the target country, emerging markets and developed countries. Then with equation (5) I distinguish the individual effects of the prior acquisition experiences that are achieved in developed economies, in emerging economies, or within the same country as the target of the current acquisition.
(4)
10
acquisitions firms completed within the past 10 years of the announcement of focal acquisition. Here measures the total acquisition experiences in emerging and developed countries. and acquisition experiences in the developed countries and emerging economies respectively. Here accounts for the acquisition experiences accumulated in all the emerging countries except the country where the target of the focal acquisition is located, for those experiences is captured by the variable , which represents the prior acquisition experiences in the target country. is the dummy variable for industry diversification; this variable
equals 1 if the first two digits of the 4 digit US SIC industry code of the acquiring firm are the same as that of the target firm, otherwise 0. It reveals whether the acquirer is in the same industry as the target firm. is the variable for relative deal size, measured by the total value of the deal divided by the outstanding market capitalization. is a dummy variable for majority control; this variable equals 1 if the acquirer’s stake in target firm after acquisition is larger than 50%, otherwise 0. is the dummy variable that equals 1 if the acquisition was announced in the period 2007-2008 during the financial crisis. It controls for the influence of macroeconomic factors. For estimated equation (4) and (5), ~ are the regression coefficients and is the error term.
3.3 Control Variables
Industrial Diversification (ID)
Industrial Diversification measures whether the target was in the same industry as the acquirer. If the company is acquiring a firm in its own industry, it is possible that the firm could benefit from his familiarity with its own industry. As acquisition is a sophisticate and dynamic process that requires close cooperation of highly trained professionals from various disciplines, other than professional skills, certain knowledge of the target industry may also necessary. An acquirer needs to possess detailed professional knowledge and information about the resources and capability of target firm to fairly evaluate the target company to negotiate a deal that is good for both, and to ensure a successful future synergy (Mcdonald, Westphal, and Graebner, 2008).
Relative Deal Size (RDS)
Relative deal size is commonly used as an important control variable in cross-sectional analysis. Von Eije and Wiegerinck (2010) found that relative deal size positively correlates with announcement returns in the US for acquirers from common law countries, as did Chari, Ouimet, and Tesar (2004), though not significant. In contrast, CONN et al. (2005) found that the relative deal size of target company have negative impact on both announcement and long run returns, regardless of payment method.
Majority Control (MC)
11
acquirer are significantly higher if majority control is shifted from the emerging-market target to the developed-market acquirer. This finding is consistent with what they observe in the market, namely that over seventy percent of the emerging-market acquisitions involve the transfer of corporate ownership. This could be explained as that the stock market values the fact that acquirer successfully extends the boundaries of the firm across border to emerging markets.
Period Effects (PER)
Zhou, Li, and Svejnar (2011) observe that during the uncertainty of a global financial crisis, firms act more conservatively and significantly reduce their acquisition behavior. Moreover, it is reasonable to assume that over that period investors may assess the risk and return of foreign acquisition differently. Therefore it is necessary to control for the potential crisis effects on acquisition performance. Consistent with von Eije and Wiegerinck (2010); Haleblian and Finkelstein (1999), I add dummy variables of period effects to the equation.
4. Sample and Data Description
4.1 Sample
The sample analyzed in this study contains cross-border acquisitions in emerging economies carried out by UK firms over the 14 year period from 2000 to 2013. The sample of acquisitions is gathered from Zephyr, a comprehensive M&A database which contains integrated detailed company information. The daily return data of the acquiring firms and the market index were obtained from DataStream. For market performance measure I use FTSE All-Share Indices (FTASI). FTSE All-Share Indices embraces the firms listed on the London Stock Exchange (LSE) and captures 98% of the UK's market capitalization5. It is considered the most suitable performance measure of UK equity market.
The research question was investigated in the context of publicly traded UK acquirers during the period between January 1, 2000 and December 31, 2013. To start with, I collected data of all cross-border acquisitions announced and carried out by listed UK acquirers in emerging markets during the past 14 years from Zephyr. I limited the sample to those acquisitions where our UK firm is the only or primary acquirer in the acquisition. By excluding the purchases of less than 10% of the target company, I focus on large transactions that are likely to have a noticeable impact on the shareholder wealth of the acquiring company. Given the hypothesis of this study, it is reasonable to study larger acquisition only. Because small acquisitions have a smaller impact on corporate value (Chatterjee and Lubatkin, 1990), acquirers tend to draw less time and attention to them (Kitching, 1994), and it is therefore less likely to generate knowledge and experience. In order to fully capture the announcement effect,
5
Source: FTSE All-Share Indices Factsheet, FTSE (May 2014)
12
I exclude the samples whose announcement date is not the same as its rumor date, which would inevitably lead to an inability to identify a clear event date. I did not rule out the companies that went bankrupt after announcing the acquisition so the sample is free of survivorship bias. Among the acquisitions that were announced during the period, six acquisitions were terminated prematurely and were thus dropped from the sample. As a result of this screening, the sample eventually consists of 317 cross-border acquisitions in 26 emerging economies carried out by 179 publicly traded UK firms.
4.2 Sample Description
Table 1: Characteristics of the Dataset of the acquisitions in emerging markets 2000-2013
Number Percentage
Panel A: Target Region
Africa 54 17% Asia 142 44% Europe 54 17% North America 6 2% South America 61 19% Total 317 100%
Panel B: Industrial Diversification
Acquirer Industry code = Target Industry code 151 48%
Acquirer Industry code Target Industry code 166 52%
Total 317 100%
Panel C: Ownership Change
Acquired Stake <50% 65 21%
Acquired Stake >50% and <100% 94 30%
Acquired Stake=100% 151 48%
Unknown Stake 7 2%
Total 317 100%
Panel D: Acquisition Period
2000-2006 (pre-crisis) 147 46%
2007-2008 (crisis) 63 20%
2009-2013 (post-crisis) 107 34%
Total 317 100%
Panel E: Acquisition Experience
xperiencei 69 22%
xperiencei , 125 40%
68 22%
55 17%
Total 317 100%
13
diversification strategy by acquiring a target outside its core business. Regarding ownership change (Panel C), around eighty percent of the acquisitions in emerging market are majority acquisitions rather than the acquisition of minority stakes (less than 50% of target stake). Meanwhile, following the global financial crisis in 2007- 2008 there is a slowdown in global economic growth, but the acquisition behavior of UK acquirers does not seem to be dramatically influenced. The total number of acquisitions announced before and after the global financial crisis are largely even per year (Panel D).
5. Results
5.1 Cumulative Abnormal Returns
Rather than simply assume that the impact of the announcement persists for some random days, in this study I employ the same methodology as Soussa and Wheeler (2006) and determine the event window by testing the significance of CARs over 16 different event windows of different durations, in order to minimize the noise in the data. Regarding the event window, in prior studies researchers always include the day before the announcement day (Tesar, 2004; von Eije and Wiegerinck, 2010; Conn, Cosh, Guest, 2005). However, the exact end of the event window differs significantly, ranging from two days after the announcement to as long as ninety days6. The results are displayed in Table 2.
Table 2: CARs over various event windows
Start of Window T+1 T+3 T+5 T+7 T 0.015*** 0.011** 0.010* 0.009* (0.000) (0.007) (0.025) (0.026) T-1 0.016*** 0.013** 0.010* 0.011* (0.000) (0.005) (0.017) (0.019) T-3 0.017*** 0.013** 0.011** 0.011* (0.000) (0.007) (0.002) (0.021) T-5 0.016** 0.013* 0.011 0.011* (0.002) (0.019) (0.051) (0.036)
*, ** and *** indicate significance level at 0.05, 0.01and 0.001 respectively.
The cumulative average return is shown at the top of the cell. The bottom of each cell shows the probability that CAR is statistically significant (the P-value). The CARs are positive and significant in 15 of the 16 event windows. The result is consistent with Soussa and Wheeler (2006) that the event window that begin within one to three days generally is larger than those that starts on the announcement day, which indicates information leaking. Moreover, the stock price seems to adjust within one day after the acquisition is announced. This finding supports the market efficiency argument and indicates that market responds rapidly to new information, as was carefully
6 Von Eije and Wiegerinck(2010) use an event window of (-2, +2). Soussa and Wheeler(2006) use an event
14
documented by Elton and Gruber (1987). I also find that extending the event window to later days result in a decrease in the significance level yield. Eventually, if we define day T=0 as the announcement date, the event window I use in this study is (T-1, T+1), which means that I include the trading day before the announcement, the announcement day, and the next trading day. Though including more trading days prior the announcement in the event window increases the ability to capture leaking event-related gains, I prefer using a shorter event window because it reduces the noise in the CAR resulting from irrelevant events and market movements. Moreover, this three day event window is one of two event windows that is most commonly used by researchers in M&A related studies (Andrade, Mitchell and Stafford, 2001).
The first result from this analysis indicates that acquisition announcements in emerging market generate statistically significant value gain of 1.6 percent. Under the assumption of market efficiency, a positive abnormal return implies that stock market considers that the benefit of expanding into emerging markets outweighs the costs, so that acquiring in emerging market creates value.
5.2 General Results
Table 3: Descriptive Statistics of Variables (N=317)
Variables Mean Median Maximum Minimum Std. Dev.
1. Acquisition Performance 0.016 0.001 0.268 -0.145 0.069
2. Acquisition Experience 6.019 5 34 0 6.782
3. Experience (Emerging Country) 0.484 0 4 0 0.874
4. Experience (Developed Country) 5.280 4 32 0 6.401
5. Experience (Target Country) 0.255 0 4 0 0.635
6. Industrial Diversification 0.522 1 1 0 0.501
7. Relative Deal Size 0.166 0.016 12.529 0 1.015
8. Major or Minor Acquisition 0.783 1 1 0 0.414
15
general firms are very experienced in acquiring in the developed markets, meanwhile they lack the acquisition experiences in the emerging economies and in the target country, where it is more likely that acquirers can generate valuable inferences.
Table 4: Correlations between Variables (N=317)
Variables 1 2 3 4 5 6 7
1. Acquisition Performance
2. Acquisition Experience -0.243***
3. Experience (Emerging Country) -0.033 0.355***
4. Experience (Developed Country) -0.262 0.988*** 0.236**
5. Experience (Target Country) 0.086 0.234** 0.035 0.144
6. Industrial Diversification -0.105 0.096 0.176* 0.079 -0.008
7. Relative Deal Size -0.028 -0.107 -0.067 -0.101 -0.031 -0.076
8. Major or Minor Acquisition -0.007** 0.068 -0.087 0.082 0.022 0.068 0.073
*, ** and *** indicate significance level at 0.05, 0.01and 0.001 respectively.
Table 4 shows the correlations between the main variables in this study. The correlations between different independent variables are relatively small, except for acquisition experience and acquisition experience in developed and emerging countries. This is understandable as acquisition experience comprises experience in developed and emerging acquisitions. At first glance this may lead to multicollinearity and consequently the bias of regression coefficients. However, as can be seen from the regression equations (4) and (5) in Section 3.2, the highly correlated variables never appear as explanatory variable in the same estimation equation. Thus we are safe from multicollinearity.
We can see that the overall acquisition experience is negatively correlated with acquisition performance. This finding is consistent with Kusewitt (1985), who also observed that relatively experienced acquirers constantly underperform acquirers with limited or no acquisition experience. This allows us to differentiate the influence of emerging-market and developed-market acquisition experiences to further examine the effects of acquisition experiences, which according to my hypothesis, would range from negative to positive.
We also see that acquisition performance is negatively correlated with acquisition experience in developed markets, which supports Hypothesis 3 that dissimilar acquisition experiences, if inappropriately generalized and applied, has a negative influence on acquisition performance. On the other hand, prior acquisition experiences in the target country and in emerging markets generate positive influence on acquisition experience, which provides preliminary evidence on Hypothesis 1 and Hypothesis 2.
5.3 Results of the Cross-sectional Analysis
16
experience and acquisition performance is carried out using Ordinary Least Square (OLS) and cross-sectional regression in this section to further test the hypothesis. The main results are summarized in Table 4, which presents the Eviews output for the two estimated linear regressions.
Table 5: Cross-sectional Regression Results (N=317)
Variables Regression Model
(4) (5)
Acquisition Experience -0.247**
(0.002) -
Acquisition Experience (Emerging country) - 0.381
(0.552)
Acquisition Experience (Developed country) - -0.318***
(0.000)
Acquisition Experience (Target Country) - 1.399
(0.104)
Industrial Diversification -1.199
(0.269)
-1.269 (0.244)
Relative Deal Size -0.420
(0.432) -0.399 (0.451) Majority Control 0.335 (0.799) 0.428 (0.744) Period 0.030 (0.982) 0.582 (0.670) Intercept 3.430 (0.010) 2.931 (0.028) R-Square7 0.069 0.098 F-statistic 2.323 2.366 Prob (F-statistic) 0.046 0.025
*, ** and *** indicate significance level at 0.05, 0.01and 0.001 respectively. The coefficients are multiplied by 100 to get the percentage effect. The bottom of each cell shows the probability that the corresponding coefficient is equal to zero (the p-value).
Hypothesis 1 and Hypothesis 2 predict positive returns to acquisition experiences in emerging markets and the target country. I anticipate that acquirer could benefit from experiences generated from similar prior acquisition and outperform in subsequent acquisitions. These two hypotheses were tested in model (5). The coefficients for acquisition experiences in both target country and emerging country are both positive, though not significant, so the hypotheses are not statistically confirmed.
Hypothesis 3 predicts that prior acquisition experiences of targets that are dissimilar
7
17
to focal acquisition would have a negative influence on acquisition performance. This prediction is tested with model (5). As table 5 shows, the coefficient for developed-country acquisition experience are negative and significant, which supports Hypothesis 3. So, in line with the behavioral learning theory and investor overconfidence, inappropriately generating experience from dissimilar experience does yield negative outcomes.
In Table 5, analysis for several control variables failed to yield significant results. Consistent with prior literature, both industrial diversification and relative deal size have a negative influence on acquisition performance. Meanwhile, investors at the stock market react positively to acquiring majority control in acquisitions targeting emerging markets, though again, the relationships are not significant.
5.4 Robustness Tests
5.4.1 Market Adjusted Returns
Table 6: Cross-sectional Regression Results Using Market Adjusted Returns (N=317)
Variables Regression Model
(4) (5)
Acquisition Experience -0.163*
(0.035) -
Acquisition Experience (Emerging country) - -0.869
(0.154)
Acquisition Experience (Developed country) - -0.134
(0.109)
Acquisition Experience (Target Country) - 0.377
(0.643)
Industrial Diversification -1.931
(0.060)
-1.729 (0.096)
Relative Deal Size -0.262
(0.602) -0.264 (0.600) Majority Control 3.256** (0.009) 3.061* (0.015) Period -0.085 (0.947) -0.083 (0.949) Intercept 1.860 (0.132) 2.003 (0.114) R-Square 0.086 0.095 F-statistic 2.899 2.294 Prob (F-statistic) 0.015 0.029
*, ** and *** indicate significance level at 0.05, 0.01and 0.001 respectively. The coefficients are multiplied by 100 to get the percentage effect. The bottom of each cell shows the probability that the corresponding coefficient is equal to zero (the p-value).
18
over a very short window, weighting the market return with estimated alpha and beta does not significantly improve estimation results. Therefore they propose a simplified method of calculating CARs by assuming that for each stock alpha equals to 0 and beta equals to 1. The cumulative abnormal returns calculated in this way are referred to as market adjusted returns and in this part of the analysis I carry out the robustness test by using market adjusted returns as the measure of acquisition performance. I believe that using market adjusted returns will not change the estimation results significantly. The estimation results of the cross-sectional regressions are presented in Table 6.
The estimation results using market adjusted returns are largely similar to what I find using the estimated alpha and beta adjusted return. However, the sign of the coefficient of acquisition experience in emerging markets is reversed, though still insignificant. The relationship between acquisition performance and overall acquisition experience is negative and significant, which supports that misapplying experiences may negatively influence acquisition performance. However, with the new performance measure the significance of coefficients is reduced and we cannot observe a significant relationship when trying to further differentiate the particular effect of acquisition experiences in emerging countries, developed economies and in the target country. However, an additional finding is that acquiring majority control in this case significantly improves acquisition performance. This is consistent with the finding of Chari, Ouimet, and Tesar (2004).
5.4.2 U-shaped Relationship
So far in my analysis I assume that the relationship between acquisition experience and acquisition performance is linear, while prior studies sometimes suggests that a more complex form of relationship may exist (Haleblian and Finkelstein 1999; Meschi and Metais 2006; von Eije and Wiegerinck 2010). Therefore in this part of the study I add quadratic experience variable to the estimated equation to test the significance of such relationship. The estimation equations are as follows:
(6)
(7)
19
experiences may negatively influence emerging-market acquisition performance. However, the coefficient of the quadratic term of acquisition experiences in target country is not significant, which suggests that the relationship between acquisition experiences in the target country and subsequent acquisition performances is not U-shaped.
Table 7: Cross-sectional Regression Results with Quadratic Terms (N=317)
Variables Regression Model
(6) (7)
Acquisition Experience -0.486*
(0.015) -
[Acquisition Experience]2 0.009
(0.185) -
Acquisition Experience (Emerging country) - 0.596
(0.716)
[Acquisition Experience (Emerging country)]2 - -0.125
(0.820)
Acquisition Experience (Developed country) - -0.727***
(0.001)
[Acquisition Experience (Developed country)]2 - 0.018*
(0.028)
Acquisition Experience (Target Country) - 4.469*
(0.025)
[Acquisition Experience (Target Country)]2 - -1.149
(0.113)
Industrial Diversification -1.214
(0.262)
-1.339 (0.213)
Relative Deal Size -0.482
(0.368) -0.449 (0.341) Majority Control 0.248 (0.850) 0.501 (0.701) Period -0.083 (0.951) 0.632 (0.639) Intercept 4.209 (0.004) 3.639 (0.012) R-Square 0.080 0.142 F-statistic 2.241 2.481 Prob (F-statistic) 0.042 0.009
*, ** and *** indicate significance level at 0.05, 0.01and 0.001 respectively. The coefficients are again multiplied by 100 to get the percentage effect. The bottom of each cell shows the probability that the corresponding coefficient is equal to zero (the p-value).
6. Conclusions
20
emerging-market acquisition performance. Intuitively a more experienced acquirer is very likely to outperform inexperienced acquirers if we apply the simple “learning by doing” perspective. However, drawing on behavioral learning theory by Haleblian and Finkelstein (1999), I predict that the prior experiences’ influence on acquisition performance may range from negative to positive, depending on the extent to which the focal target is similar to prior targets. The study is carried out in the context of emerging-market acquisitions where a prior acquisition in emerging markets is defined as similar and a prior acquisition in developed market is defined as dissimilar. My first finding is that when UK firms acquire in emerging markets, the acquirer experiences a significant positive announcement return of 1.6 percent. After controlling for relative deal size, a majority control dummy and a diversification strategy dummy, the linear regression results show that the correlation between prior acquisition experience and subsequent acquisition performance is significant and negative, which could be a mix effect of similar and dissimilar experiences. A further examination differentiating acquisition experience in emerging market and developed economies reveals that acquisition experience in developed economies negatively influences acquisition performance, and the relationship is U-shaped (relatively inexperienced acquirer inappropriately generate inferences from dissimilar developed-market acquisitions, while experienced acquirer properly discriminate between experiences). Finally, acquisition experience in emerging market, and in the target country positively influences emerging-market acquisition performance, though the relationships are not significant.
21
References
Ackert, Lucy, and Richard Deaves. Behavioral finance: Psychology, decision-making, and markets. Cengage Learning, 2009.
Andrade, Gregor, Mark Mitchell, and Erik Stafford. "New evidence and perspectives on mergers." Journal of economic perspectives (2001): 103-120.
Argote, Linda, and Dennis Epple. "Learning curves in manufacturing." Science 247.4945 (1990): 920-924.
Barkema, Harry G., John HJ Bell, and J. M. E. Pennings. "Foreign entry, cultural barriers and learning." Strategic management journal (1996): 151-166.
Bettis, Richard A. "Performance differences in related and unrelated diversified firms." Strategic Management Journal 2.4 (1981): 379-393.
Boone, Tonya, Ram Ganeshan, and Robert L. Hicks. "Learning and knowledge depreciation in professional services." Management Science 54.7 (2008): 1231-1236. Brown, Stephen J., and Jerold B. Warner. "Using daily stock returns: The case of event studies." Journal of financial economics 14.1 (1985): 3-31.
Cakici, Nusret, Chris Hessel, and Kishore Tandon. "Foreign acquisitions in the United States: Effect on shareholder wealth of foreign acquiring firms."Journal of Banking & Finance 20.2 (1996): 307-329.
Carretta, Alessandro, et al. "The impact of corporate governance press news on stock market returns." European financial management 17.1 (2011): 100-119.
Chari, Anusha, Paige P. Ouimet, and Linda L. Tesar. Acquiring control in emerging markets: Evidence from the stock market. No. w10872. National Bureau of Economic Research, 2004.
Chatterjee, Sayan, and Michael Lubatkin. "Corporate mergers, stockholder diversification, and changes in systematic risk." Strategic Management Journal 11.4 (1990): 255-268.
Conn, Robert L., et al. "The impact on UK acquirers of domestic, cross‐border, public and private acquisitions." Journal of Business Finance & Accounting 32.5‐6 (2005): 815-870.
Didier, Tatiana, Constantino Hevia, and Sergio L. Schmukler. "How resilient were emerging economies to the global crisis?." (2011).
Duso, Tomaso, Klaus Gugler, and Burcin Yurtoglu. "Is the event study methodology useful for merger analysis? A comparison of stock market and accounting data." International Review of Law and Economics 30.2 (2010): 186-192.
22
International Business Review 19.4 (2010): 360-377.
Epple, Dennis, Linda Argote, and Rukmini Devadas. "Organizational learning curves: A method for investigating intra-plant transfer of knowledge acquired through learning by doing." Organization Science 2.1 (1991): 58-70.
Fowler, Karen L., and Dennis R. Schmidt. "Determinants of tender offer post‐acquisition financial performance." Strategic Management Journal 10.4 (1989): 339-350.
Haleblian, Jerayr, and Sydney Finkelstein. "The influence of organizational acquisition experience on acquisition performance: A behavioral learning perspective." Administrative Science Quarterly 44.1 (1999): 29-56.
Haspeslagh, Philippe C., and David B. Jemison. Managing acquisitions: Creating value through corporate renewal. Vol. 416. New York: Free Press, 1991.
Hayward, Mathew LA. "When do firms learn from their acquisition experience? Evidence from 1990 to 1995." Strategic Management Journal23.1 (2002): 21-39. Hearst, Eliot, and Minnie B. Koresko. "Stimulus generalization and amount of prior training on variable-interval reinforcement." Journal of Comparative and Physiological Psychology 66.1 (1968): 133.
Kaplan, Steven N., and Michael S. Weisbach. "The success of acquisitions: Evidence from divestitures." The Journal of Finance 47.1 (1992): 107-138.
Kitching, John. "Why do mergers miscarry." Harvard Business Review 45.6 (1967): 84-101.
Kusewitt, John B. "An exploratory study of strategic acquisition factors relating to performance." Strategic Management Journal 6.2 (1985): 151-169.
Lubatkin, Michael. "Mergers and the Performance of the Acquiring Firm."Academy of Management Review 8.2 (1983): 218-225.
MacKinlay, A. Craig. "Event studies in economics and finance." Journal of economic literature (1997): 13-39.
McDonald, Michael L., James D. Westphal, and Melissa E. Graebner. "What do they know? The effects of outside director acquisition experience on firm acquisition performance." Strategic Management Journal 29.11 (2008): 1155-1177.
Meschi, Pierre-Xavier, and Emmanuel Metais. "International acquisition performance and experience: A resource-based view. Evidence from French acquisitions in the United States (1988–2004)." Journal of International Management 12.4 (2006): 430-448.
23
Novick, Laura R. "Analogical transfer, problem similarity, and expertise."Journal of Experimental Psychology: Learning, Memory, and Cognition 14.3 (1988): 510.
Rendleman Jr, Richard J., Charles P. Jones, and Henry A. Latane. "Empirical anomalies based on unexpected earnings and the importance of risk adjustments." Journal of Financial Economics 10.3 (1982): 269-287.
Rumelt, Richard P. "Diversification strategy and profitability." Strategic management journal 3.4 (1982): 359-369.
Shimizu, Katsuhiko, et al. "Theoretical foundations of cross-border mergers and acquisitions: A review of current research and recommendations for the future." Journal of International Management 10.3 (2004): 307-353.
Soussa, Farouk, and Tracy Wheeler. Do announcements of bank acquisitions in emerging markets create value?. Bank of England, 2006.
Very, Philippe, and David M. Schweiger. "The acquisition process as a learning process: Evidence from a study of critical problems and solutions in domestic and cross-border deals." Journal of World Business 36.1 (2001): 11-31.
