Factors influencing cross-border M&A: Culture, geography, and taxes - Prof. 8828, Apuntes de Arquitectura

¡Descarga Factors influencing cross-border M&A: Culture, geography, and taxes - Prof. 8828 y más Apuntes en PDF de Arquitectura solo en Docsity! Determinants of Cross-Border Mergers and Acquisitions Isil Erel Ohio State University Rose C. Liao Rutgers University Michael S. Weisbach Ohio State University and NBER March 15, 2011 Abstract Despite the fact that one-third of worldwide mergers involve firms from different countries, the vast majority of the academic literature on mergers studies domestic mergers. What little has been written about cross-border mergers has focused on public firms, usually from the United States. Yet, the vast majority of cross-border mergers involve private firms that are not from the United States. We provide an analysis of a sample of 56,978 cross-border mergers occurring between 1990 and 2007. In addition to the factors that motivate domestic mergers, national borders are associated with additional factors that also affect the likelihood that two firms choose to merge. Specifically, geography, the quality of accounting disclosure, and bilateral trade increase the likelihood of mergers between two countries. In addition, valuation appears to play a role in motivating mergers; firms in countries whose stock market has increased in value, whose currency has recently appreciated, and who have a relatively high market to book value tend to be purchasers, and firms from weaker-performing economies tend to be targets. * Contact information: Isil Erel, Department of Finance, Fisher College of Business, Ohio State University, Columbus, OH 43210: email: erel@fisher.osu.edu; Rose C. Liao, Department of Finance and Economics, Rutgers Business School at Newark and New Brunswick, Room 1152, 1 Washington Park, Newark, NJ, 07102: email: liao@business.rutgers.edu; Michael S. Weisbach, Department of Finance, Fisher College of Business, Ohio State University, Columbus, OH 43210, email: weisbach.2@osu.edu. We would like to thank Anup Agrawal, Malcolm Baker, Phil Davies, Mara Faccio, Charlie Hadlock, Campbell Harvey, Jim Hines, Andrew Karolyi, Simi Kedia, Sandy Klasa, Tanakorn Makaew, Pedro Matos, Taylor Nadauld, John Sedunov, Léa Stern, René Stulz, Jérôme Taillard, two referees, and seminar participants at Chinese University of Hong Kong, HKUST, IDC, Lingnan University, Michigan State University, Ohio State University, Ohio University, Rutgers University, Seton Hall University, University of Alabama, University of Maryland, and Washington University for very helpful suggestions. 1 1. Introduction. The volume of cross-border acquisitions has been growing worldwide, from 23 percent of the total merger volume in 1998 to 45 percent in 2007. Conceptually, cross-border mergers occur for the same reasons as domestic ones: two firms will merge when combining them increases the value (or utility) from the perception of the acquiring firm’s managers. However, national borders add an extra element to the calculus of domestic mergers, because they are associated with an additional set of frictions that can impede or facilitate mergers. For example, cultural or geographic differences can increase the costs of combining two firms. Governance-related differences across countries can motivate a merger if the combined firm has better protection for target-firm shareholders because of higher governance standards in the country of the acquiring one. Finally, and perhaps most importantly, imperfect integration of capital markets across countries can lead to a merger, in which a higher- valued acquirer purchases a relatively inexpensive target following changes in exchange rates or stock market valuations in local currency. This paper evaluates the extent to which these international factors influence the decision of firms to merge. Using a sample of 56,978 cross-border mergers occurring between 1990 and 2007, it estimates the factors that affect the likelihood that firms from any pair of countries merge in a particular year. The analysis focuses on factors that potentially affect cross-border mergers but are not present to the same extent in domestic mergers, such as cultural differences, geographic differences, country-level governance differences, and international tax effects. Of particular interest are differences in valuation, which can vary substantially over time for any pair of countries through fluctuations in exchange rates, stock market movements and macroeconomic changes. In contrast to most of the prior literature that focuses on mergers of public firms, usually involving U.S. acquirers or targets, our sample better reflects the universe of cross-border mergers, the majority of which involve private firms from outside the U.S. In our sample, 80% of completed cross-border deals between 1990 and 2007 targeted a non-US firm, while 75% of the acquirers are from outside the U.S. Furthermore, the vast majority of cross-border mergers involve private firms as either bidder or target: 96% of the deals involve a private target, 26% involve a private acquirer, and 97% have either private acquirers or targets. Hence, the inclusion of private 4 investments for the acquiring firm. Alternatively, as suggested by Shleifer and Vishny (2003), either overpricing of the acquiring firm or underpricing of the target firm could lead to a potentially profitable investment for the acquiring firm. Baker, Foley, and Wurgler (2009) suggest a test to distinguish between the two explanations based on the implication that subsequent to acquisitions occurring due to mispricing, valuations will tend to revert to their true values. We perform a similar test to that in Baker, Foley, and Wurgler (2009), and find that the wealth explanation better explains the relation between valuation differences and cross-border mergers than the mispricing explanation. We then examine at the deal level whether valuation differences affect the likelihood of cross-border M&As. We find that differences in firm-level stock returns (in a common currency) are associated with higher likelihood of cross-border deals compared to domestic deals. We further decompose valuation differences between acquiring and target firms into three components: the difference in returns of the two countries’ currencies, the differences in local stock market or industry indices, and the differences in firm-level excess returns relative to the market or industry indices. All three of these factors lead to a higher likelihood of a particular merger being cross-border than domestic, although statistical significance varies depending on the specification used. These firm-level results confirm the country-level ones, and are consistent with the view that valuation is an important factor that determines merger likelihoods. 2. Cross-Border Mergers and Acquisitions Despite the fact that a large proportion of worldwide merger activity involves firms from different countries, the voluminous literature on mergers has focused primarily on domestic deals between publicly-traded firms in the United States. While this literature is also relevant to understanding international mergers, it does not address a number of factors related to country-based differences between firms, such as cultural or geographic variables, or factors associated with the firm’s home country’s economy. In addition, public U.S. firms are unrepresentative of mergers more generally, since the majority of worldwide mergers involve non-U.S. firms, many of which are private.1 1 One recent study using a much more representative sample of mergers than is typical in merger studies is Netter, Stegemoller, and Wintoki (2009), whose primary focus, unlike ours, is on domestic mergers. These authors present evidence 5 2.1. Factors that Potentially Affect Cross-Border Merger Likelihoods National boundaries are likely to be associated with many frictions that determine firm boundaries. In general, mergers occur when the managers of the acquiring firms perceive that the value of the combined firms is greater than the sum of the values of the separate firms.2 This change in value can occur for any of a number of reasons. Contracting costs can be lower within than across firms, creating production efficiencies to combining them. Mergers can create market power since it is legal for post-merger combined firms to charge profit- maximizing prices themselves but not for pre-merger separate firms to collude to do so collectively. Mergers can lower the combined tax liability of the two firms if they allow one firm to utilize tax shields another firm possesses but cannot use. Finally, agency considerations can lead managers to make value-decreasing acquisitions that nonetheless increase managers’ individual utilities. All of these factors are relevant both domestically and internationally. National borders are associated with a set of factors that are likely to affect the costs and benefits of a merger. First of all, countries have their own cultural identities. People in different countries often speak different languages, have different religions, and sometimes have longstanding feuds, all of which increase the contracting costs associated with combining two firms across borders (see Ahearn, Daminelli, and Fracassi (2010)). Second, similar to the “gravity” literature in international trade, physical distance can increase the costs of combining two firms (see Rose 2000). Both cultural differences and geographic distance should decrease the likelihood that, holding other factors constant, two firms in different countries choose to merge. Corporate governance considerations can also affect cross-border mergers. If merging can increase the legal protection of the minority shareholders in target firms by providing them some of the rights of acquiring firm’s shareholders, then value can be created through the acquisition. In general, corporate governance arguments predict that firms in countries that promote governance through better legal or accounting standards suggesting that filters that researchers commonly use in obtaining M&A data lead to samples containing a small subset of the entire mergers universe, usually oversampling larger transactions by publicly-held companies. 2 See Jensen and Ruback (1983), Jarrell, Brickley, and Netter (1988) and Andrade, Mitchell, and Stafford (2001) for surveys of the enormous literature on mergers. 6 will tend to acquire firms in countries with lower-quality governance.3 The level of development of the markets is another factor that could affect cross-border mergers. In particular, developed-market acquirers are likely to benefit more from weaker contracting environments in emerging markets.4 A potentially important factor in international mergers is valuation. Given that markets in different countries are not perfectly integrated, valuation differences across markets can help to motivate cross-border mergers. Suppose, for example, that a firm’s currency rises for some exogenous reason unrelated to the firm’s profitability. This firm would find potential targets in other countries relatively inexpensive, leading some potential acquisitions to be profitable that would not have been profitable at the old exchange rates. Therefore, we expect to observe more firms from this country to engage in acquisitions, since they will be paying for these acquisitions in an inflated currency.5 The logic by which valuation differences can lead to cross-border mergers depends on whether participants believe these movements are temporary or permanent. If the valuation differences are temporary, then cross-border mergers effectively arbitrage these differences, leading to expected profits for the acquirers. Shleifer and Vishny (2003) develop a behavioral model in which firm values deviate from their fundamentals. Managers of an overvalued acquirer consequently have incentives to issue shares at inflated prices to buy assets of an undervalued or at least a less overvalued target. This transaction transfers value to the shareholders of the acquiring firm by arbitraging the price difference between the firms’ stock prices. The key component of this model is that the source of the valuation difference is private information owned by managers. 6 While it is implausible that one particular firm’s managers have superior information about the valuation of the overall market or any particular currency, Baker, Foley, and Wurgler (2009) argue that cross-border mergers could similarly occur because of mispricing of securities from fluctuations in risk-aversion by local investors or 3 Rossi and Volpin (2004), Bris and Cabolis (2008), and Bris, Brisley, and Cabolis (2008) all provide support for this argument using samples of publicly-traded firms. 4 See Chari, Ouimet, and Tesar (2009) for more discussion and evidence on this point. 5 A recent example of this phenomenon occurred when the Japanese yen appreciated relative to other major currencies in the summer of 2010, leading Japanese firms to increase their number of cross-border acquisitions substantially (see The Economist, August 5, 2010 or The New York Times, September 15, 2010, p. B1). 6 A similar argument in which a firm’s managers have superior information about rational stock movements than other market participants has been proposed by Rhodes-Kropf and Viswanathan (2004). Using a sample of U.S. domestic mergers, Rhodes-Kropf, Robinson, and Viswanathan (2005) provide empirical support for this argument. 9 3. Data Our merger sample is taken from Security Data Corporation’s (SDC) Mergers and Corporate Transactions database announced between 1990 and 2007 and completed by the end of 2007. We exclude LBOs, spin-offs, recapitalizations, self-tender offers, exchange offers, repurchases, partial equity-stake purchases, acquisitions of remaining interest, and privatizations, as well as deals in which the target or the acquirer is a government agency, or in the financial or utilities industry. We then drop deals from countries with incomplete stock market data between 1990 and 2007.9 After excluding these deals, we end up with a sample of 187,841 mergers covering 48 countries, with the total transaction value of $7.54 trillion, 56,978 of which are cross-border mergers, having a total transaction value of $2.21 trillion. We collect a number of data items from SDC, including the announcement date, the completion date, the target’s name, its public status, its primary industry measured by the four-digit Standard Industrial Classification code, country of domicile, the name of the acquirer and its ultimate parents, its public status as well as its primary industry and country of domicile. We collect the deal value in dollar terms when available, the fraction of the target firms owned by the acquirer after the acquisition, as well as other deal characteristics, such as the method of payment made by the acquirer. We acquire monthly firm-level and country-level stock returns both in local currency and in U.S. dollars from Datastream. We also obtain the national exchange rates from WM/Reuters (WMR) through Datastream, whose quotes are based on 4:00pm Greenwich Mean Time. We then calculate nominal exchange rate returns by taking the first difference of the monthly natural logarithm of the national exchange rates. To calculate real stock market returns and real exchange rate returns, we obtain from Datastream the monthly Consumer Price Index (CPI) for each country in each month and convert all nominal returns to the 1990 price level.10 When calculating real exchange rate returns for the Economic and Monetary Union (EMU) countries, we use the Euro and the local owners. Finally, Coeurdacier, DeSantis and Aviat (2009) use a database on bilateral cross-border M&As at the sector level (in manufacturing and services) over the period 1985-2004, and find that institutional and financial developments, especially the European Integration process, promote cross-border mergers and acquisitions. 9 This filter on dropping deals from countries without stock market returns excluded 4,061 deals worth cumulatively $145 billion, or 2% of the original sample count. 10 For Australia and New Zealand, we only have quarterly prices. When extrapolating to monthly prices, we assume that prices are as of the end of month/quarter. 10 corresponding CPI for EMU countries after 1999. This approach implies that all EMU countries have the same exchange rate movements in our database after 1999. We obtain ratings on the quality of accounting disclosure from the 1990 annual report of the Center for International Financial Analysis and Research as well as a newly assembled anti-self dealing index from Djankov, La Porta, Lopez-de-Silanes, and Shleifer (DLLS, 2008). Our culture variables, language (English, Spanish or Others) and religion (Protestant, Catholic, Muslim, Buddhist or Others), are from Stulz and Williamson (2003). We obtain the latitude and longitude of capital cities of each country from mapsofworld.com and calculate the Great Circle Distance between a country pair.11 The data on the average corporate income-tax rates are from OECD. We obtain annual Gross National Product (in U.S. dollars) normalized by population and annual real growth rate of the Gross Domestic Product from the World Development Indicator report. To control for the quantity of business done between a country pair, we include bilateral trade flows, calculated as the maximum of bilateral imports and exports between the two countries. Bilateral imports (exports) is calculated as the value of imports (exports) by target country from (to) acquirer country as a percentage of total imports (exports) by target country, all of which are from the United Nation Commodity Trade Statistics database (see Ferreira, Massa, and Matos (2009)). Following Bekaert, Harvey and Lundblad (2005) and Bekaert, Harvey, Lundblad and Siegel (2007), we construct an index of the quality of a country’s institutions based on the sum of the International Country Risk Guide (ICRG) political risk subcomponents: Corruption, Law and Order, and Bureaucratic Quality. We also use the investment profile subcategory in the ICRG political risk ratings as a measure of the state of investment environment in a country. For the public firms in our M&A sample, we obtain accounting and ownership information from Worldscope/Datastream. In particular, we use firm size (book value of total assets), book leverage (long-term debt divided by total assets), cash ratio (cash holdings divided by total assets), the two-year geometric average of sales growth, and return on equity as well as the market-to-book ratio of equity. To calculate country-level market-to- 11 The standard formula to calculate great circle distance is: 3963.0 * arcos [sin(lat1) * sin(lat2) + cos (lat1) * cos (lat2) * cos (lon2 - lon1)], where lon and lat are the longitudes and latitudes of the acquirer and the target country locations, respectively. 11 book ratio, we follow Fama and French (1998) and sum the market value of all equity for all public firms in a country and divide it by the sum of their book values. The details on the definitions of these variables can be found in Table A1. 4. Results 4.1. Stylized Facts about Cross-Border Mergers Mergers involving acquirers and targets from different countries are substantial, in terms of both absolute number, and as a fraction of worldwide M&A activity. Figure 1 plots both the number (Panel A) and dollar value (Panel B) of cross-border deals over our sample period. Both figures show similar patterns. The volume of cross- border mergers increases throughout the 1990s peaking in 2000, declines after the stock market crash of 2000, and increases again from 2002 until 2007. As a fraction of the total value of worldwide mergers, cross-border mergers typically amount to between 20 and 40 percent. The fraction of cross-border deals follows the overall level of the stock market; the fraction drops in the early 1990s, increases in the later 1990s to a peak in 2000, and then increases again with the stock market between 2004 and 2007. Table 1 characterizes the pattern of cross-country acquisitions during our sample period. The columns represent the countries of the acquiring companies while the rows represent those of the target companies. The diagonal entries of the matrix are therefore the number of domestic mergers for a particular country and the off- diagonal entries are the number of deals involving firms from a particular pair of countries. The totals reported in the bottom row and the right column exclude domestic mergers. Hence, these totals represent the number of cross- border mergers to and from a particular country. The country with the largest number of acquisitions is the U.S.; U.S. firms were acquirers in 15,034 cross-border mergers and were targets in 11,886 cross-border mergers. These numbers are substantial but do not represent the majority of the 56,978 cross-border mergers. A casual glance at Table 1 indicates that geography clearly matters. For every country, domestic mergers outnumber deals with any other country. Of the cross-border mergers, there is a large tendency to purchase companies in nearby countries. For example, of the 226 cross-border acquisitions by New Zealand companies, about two-thirds, 145, were of Australian companies. By far the largest target of Hong Kong based companies 14 average stock market return difference is negative and significant but this effect seems to be driven only by private target-private acquirer pairs. However, the average country-level market-to-book ratio has a significantly positive coefficient. Second, consistent with Rossi and Volpin (2004), having a higher quality of accounting disclosure system increases the likelihood that firms from a country will be purchasers of firms from another country. This effect appears to be driven by deals with public acquirers, which are most affected by disclosure requirements (see Columns 7-10).16 Third, the regional effect discussed above is evident; holding other things constant, the shorter the distance between two countries, the more likely there are acquisitions between firms in these countries.17 Finally, larger differences in corporate-income-tax rates attract foreign investment. There is no evidence that sharing a common language or religion has any impact on merger propensities once other factors are controlled for. [See Ahearn, Daminelli, and Fracassi (2010) for more analysis of this issue.] 4.3. Differences in Valuation Using Country-Level Panel Data: Univariate Evidence To understand the role of valuation differences in motivating cross-border mergers, we present data on measures of valuation of acquirer and target firms. As measures of valuation, we focus on differences in real exchange rate returns, differences in real stock returns in local currency, and differences in market-to-book ratios prior to the acquisition. Because only a small minority of the deals in our sample contains both acquirers and targets that are publicly-traded, we present these measures both at the country and firm levels. We first calculate these return differences for the entire sample of cross-border mergers.18 For both the recent change in valuation (local stock market returns and exchange rate appreciation) and the level of valuation (market-to-book ratio), acquirers are more highly valued than targets. The exchange rate of acquiring companies appreciates relative to that of the target companies, by 1.12% in the year prior to the acquisition, by 2.13% in the two-year period and by 3.43% in the three-year period prior to the acquisition. In addition, the average local stock 16 A potential concern with the quality of accounting disclosure effect is that it might be up an “emerging markets” effect in that disclosure quality could proxy for the level of economic development. To address this possibility, we examine whether the accounting disclosure effect exists within subsamples of developed and emerging country targets (see Table A.2). The results suggest that disclosure quality matters in each subsample, though with a larger magnitude when the target is from an emerging market. 17 This result parallels those from a growing literature on the effect of geography in domestic acquisitions. For example, Kedia, Panchapagesan, and Uysal (2009) find that in domestic acquisitions, acquirers experience higher returns when they are geographically closer to targets, potentially due to better information sharing between firms that are closer to one another. 18 We present detailed statistics on the valuation differences between targets and acquirers in Table A.3. 15 market returns are higher for acquiring firm countries than target firm countries, by 0.3% in the year prior to the merger, by 0.92% in the two-year period prior to the merger and by 2.12% in the three-year period prior to the merger. Finally, the market-to-book ratio averages almost 10% higher for acquiring countries than for target countries. All of these results are consistent with the view that firms purchase firms when they are relatively highly valued. For the subsample of mergers for which the acquirers and targets are each publicly traded and hence have firm-level stock returns, acquirers substantially outperform targets prior to the acquisitions. The differences are much larger than the country-level differences, about 10% in the year prior to the acquisition, 19% in the two-year period prior to the acquisition and 23% in the three-year period prior to the acquisition. This relation is again consistent with the valuation arguments and is similar to what others have found for domestic acquisitions (see Rhodes-Kropf, Robinson, and Viswanathan (2005), Dong, Hirshleifer, Richardson, and Teoh (2006), and Harford (2005)). This pattern can be clearly seen in Panel A of Figure 2. Prior to the month of the acquisition, differences in both the local currency stock returns and exchange rate returns are positive, meaning that the stock market of the acquirer’s country outperformed that of the target country and that the acquirer’s currency appreciated relative to the target’s during the three years prior to the acquisition. Subsequent to the acquisition, however, the stock return difference disappears, implying that the target country’s stock market outperforms the acquirer’s during the three years subsequent to the acquisition. However, the acquirer’s currency continues to appreciate, leaving the common-currency returns in the two countries’ stock markets approximately the same following the acquisitions. The post-acquisition appreciation of the acquirer’s currency relative to the target’s probably reflects the composition of acquirers and targets; acquirers are more likely than targets to be from developed economies and over the sample period, developed economies’ currencies tended to appreciate relative to those of developing countries. This pattern emphasizes the importance of controlling for country-pair effects econometrically when estimating the determinants of cross-border merger propensities (as we do below). We also break down the sample by whether the acquirer and target are from developing or developed countries, using the World Bank definition of “high income” economies. The pre-acquisition local return 16 differences are positive for each category, although they are substantially larger when a developed acquirer buys a developing target (12.79% difference in pre-acquisition returns) than when a developing acquirer buys a developed target (9.54% difference). However the currency movements prior to the deal go in opposite directions for these two categories. When a developing acquirer buys a developed target the acquirer’s currency actually depreciates prior to the acquisition (-23.32% pre-acquisition exchange rate difference). On the other hand, when a developed acquirer buys a developing target, it generally follows a period of strong relative appreciation (34.22% difference). This pattern, which can be seen in Panel B of Figure 2, could reflect a general appreciation of currencies in developed countries relative to developing ones over our sample period and emphasizes the importance of controlling for these effects econometrically. 4.4. Differences in Valuation Using Country-Level Panel Data: Multivariate Evidence To evaluate the hypothesis that relative valuation can affect merger propensities formally, we rely on a multivariate framework that controls for other potentially relevant factors. It is not obvious, however, what the most natural approach is to address this question. One possibility is to use deal-level data on the acquirer and target’s market valuations. This approach has the advantage of utilizing the most accurate measure of firm values in the comparison. However, it has the disadvantage of only being usable for the subsample of public acquirers and public targets. As discussed above, the vast majority of cross-border acquisitions have either private acquirers or targets (or both), so using deal-level data necessitates discarding the vast majority of the sample. An alternative approach relies on country-level data. This approach has the disadvantage of ignoring firm-level information (where available) but has the advantage of being able to utilize the entire sample of deals. In addition, a number of hypotheses of interest, in particular those concerning currency movements and country-level stock market movements, are testable using country-level data. Since each approach has both advantages and disadvantages, we use both: We first estimate equations using the entire sample of deals using country-level data on market indices, valuation levels, and exchange rates. We then estimate equations with deal-level data on the smaller sample of deals involving public acquirers and targets. We estimate a specification in which the dependent variable is the number of deals between an ordered particular country pair, normalized by the sum of the total number of domestic deals in the target country and the 19 otherwise economically-sensible mergers more attractive and they should not be thought of as a pure financial arbitrage. 4.4.2. How Large is the Effect of Valuation on Merger Propensities? The estimated coefficients reported in Column 1 of Table 3 imply that a one standard-deviation increase in the real exchange rate change for a given country pair (17%) is associated with an increase of 12% in the expected number of cross-border acquisitions of firms in countries with relatively depreciated currency. 21 Similarly, a one standard deviation change in the country-level stock return difference for a given country pair (27%) is expected to lead to an increase of 6.4% in the number of acquisitions by the better-performing country’s firms of the worse performing country’s firms.22 Finally, the estimates imply that a one standard deviation increase in the market-to-book difference for a given country pair (0.72) is associated with an increase of 6.4% in the expected volume of cross-border mergers. However, the quantitative importance of the impact of valuation on merger propensities implied from the estimates presented in Table 3 varies substantially depending on the characteristics of the country pair. For a pair of counties in which the acquiring country is wealthier than the target country and which are located closer to the median distance to one another, a one standard-deviation movement in the exchange rate (17%) leads to a 36% increase in the expected ratio of cross-border mergers to domestic mergers between the two countries. In contrast, for a country pair for which the acquirer country is poorer than the target country and the countries are located relatively far away, the effect is much smaller. A one standard deviation movement in the exchange rate (17%) only leads to a 5.9% increase in the expected ratio of cross-border mergers to domestic mergers between the two countries. These calculations indicate that valuation differences can be important drivers of mergers in situations where there are other reasons for firms to merge, but are not as important in situations for which the valuation differences is the only reason for the merger. 21 The average ratio of cross-border merger to domestic mergers for a given country-pair in a given year is 0.0461. Given the coefficient of the country-level 12-month real exchange rate return difference between the target and the acquirer country from Column (1) of Table 3 (0.032), the percentage change in the ratio for an average country pair for a one standard deviation change in exchange rate returns equals (0.032*17%)/0.0461=12%. 22 The average ratio of cross-border merger to domestic mergers for a given country-pair in a given year is 0.0461. Given the coefficient of the country-level 12 month real stock return difference in Column (1) of Table 3 (0.011), the percentage change in the ratio for a one standard deviation change in stock return differences equals (0.011*27%)/0.0461=6.4%. 20 Another way to evaluate the importance of valuation on merger propensities is to reestimate the equations in Table 3 for the subsample of country pairs for which there are large currency movements in the sample. If currency movements do indeed drive cross-border mergers, we should observe these types of mergers predominately for country pairs in which there are substantial currency movements. To examine this idea, we reestimate Table 3 on subsamples of country pairs based on the average exchange rate movement between these countries. Table A.4 presents these results, first using the subsample for which the exchange rate return differential is in the top three quartiles of the sample, followed by the top two quartiles, the top quartile, the top 90 percentile, and finally top 95 percentile. The coefficient on exchange rate returns in this table increases substantially from 0.03 for those country pairs whose exchange rate differential is in the top three quartiles to 0.593 for those country pairs in the top 95 percentile. For the country-pairs whose exchange rate differential is in the top 90 percentile, the estimates imply that a one standard-deviation movement in the exchange rate (16%) leads to a 64% increase in the expected ratio of cross-border mergers to domestic mergers between the two countries. These results strongly suggest that the magnitude of the currency effect varies substantially across country pairs and is economically important for country pairs for which mergers tend to occur even in the absence of currency motives, and also in those pairs of countries that tend to experience the largest currency movements. 4.5. Differences in Valuation Using Country-Level Panel Data: Alternative Specifications To perform the analyses presented above, we had to make a number of choices about the sample and specification. Table 5 contains estimates of equations similar to those reported in Tables 3 and 4 to examine the robustness of the results to alternative specifications. The sample used to estimate the equations in Tables 3 includes only the deals that lead to majority (larger than 50%) ownership by the acquiring firm. An important issue is the extent to which they hold in cases where an acquirer purchases a large minority stake (5-49%), and whether they are different for majority but incomplete (50- 99%) acquisitions from 100% acquisitions. In Columns 1, 2 and 3 of Table 5, we provide estimates of the equation reported in Table 3 for deals that lead to minority-block ownership (5-49%), for majority but not complete acquisitions (50-99%) and for 100% acquisitions. The coefficient on the currency return difference between the acquirer and the target countries is positive in all three columns and is statistically significant at the 21 1% level while the coefficient on country-level stock return difference is statistically significant in Columns 2 and 3. These results suggest that the valuation effect appears to be robust regardless of the fraction of stock purchased by the acquirer. If valuation is a motive for cross-border mergers, then one might expect those mergers that are motivated by valuation to be more commonly financed with stock rather than cash. In Columns 4 and 5, we evaluate this argument by reestimating the equation separately for cash-financed and stock-financed deals. In these equations the difference in currency return between the target and the acquiring countries is positive and statistically significant both for cash-only and stock-only deals. The coefficients on the stock return are positive but not statistically significantly different from zero. However, information on the method of payment is missing for more than half the observations in the SDC sample. Therefore, we also perform our tests for the subsample where the method of payment is missing or is a mix of cash and stock (82% of all deals in our sample). As shown in Column 6, our main results hold for this subsample, suggesting that missing data on the method of payment prevents one from making sharp inferences about the way the valuation effect depends on the method of payment. In Column 7 of Table 5, we reestimate our equation using the value instead of the number of mergers in a particular country-pair to construct our dependent variable. Using this specification, both the coefficient on currency returns as well as on stock market returns are small and insignificantly different from zero. This finding suggests that the valuation effects are more important for smaller firms that do not have a large impact on value- weighted dependent variables. In addition, there are a substantial number of observations for which the value of the deal is missing (59% of the entire sample, 70% of private targets have missing deal values on SDC). These missing values are more likely to be associated with smaller, private firms. To explore the reasons why the value- weighted results are different from the equally-weighted results, we reestimate our tests on the subsample of mergers without deal value information (Column 8) and for the subsample with deal value information (Column 9). The coefficient on the country-level stock return difference is highly significant for the mergers with missing deal values in SDC but it loses significance when we focus on the mergers with information on deal values. The coefficient on the currency return difference is statistically significant in both subsamples but larger in magnitude for the mergers with missing deal values. These results suggest that the valuation effect is most important among 24 4.7. Differences in Valuation Using Deal-Level Panel Data We have documented that valuation appears to play an important role in determining which firms are likely to merge. Acquirers tend to be valued relatively highly compared to targets, using prior returns or market- to-book ratios as measures of valuation. The difference in valuation between acquirers and targets appears to occur due to both stock market and currency effects. Yet, the results presented so far utilize country-level data. Consequently, they do not control for firm-level factors that potentially affect the decision to merge, including the firm’s own valuation. To control for firm-level factors, we consider the subsample of firms for which we have public data on both acquirers and targets. Unfortunately, this subsample is both relatively small and unrepresentative of the overall sample of mergers, because firms in this subsample are much more likely to be from developed rather than developing countries. Of the 56,978 cross-border mergers in our sample, only 1,178 have both public acquirers and targets, and also have data available on firm-level variables that we use to control for other factors that potentially affect mergers. Of these 1,178 mergers, 877 have acquirers from developed countries and 780 have targets from developed countries. While these mergers are interesting in their own right, they are nonetheless not representative of cross-border mergers in general. To estimate the factors that affect the likelihood of a merger, one would ideally like to consider every possible pair of firms that could conceivably merge and estimate the likelihood that any two of them actually do merge. Unfortunately, this approach would be infeasible as the number of possible combinations would be extremely large relative to the number of actual mergers. Instead, we adopt two alternative approaches, each of which allows us to draw inferences about the factors leading one firm to buy another. 4.7.1. Cross-Border vs. Domestic Mergers We first consider the sample of all mergers of publicly traded firms (including domestic ones), and estimate the characteristics of the firms involved with the merger that lead a particular merger to be either cross- border or domestic. We estimate logit models that predict whether an observed merger is domestic or cross-border as a function of deal characteristics. Intuitively, this approach presumes that domestic mergers can provide a benchmark for understanding the nature of cross-border mergers. 25 We present marginal effects of these logit models in Table 7. The first two columns include the difference in the acquirer and target firm-level returns, converted to U.S. dollars, ((Firm USR12)j-i) as an explanatory variable. Both coefficients are positive and in the second column, which controls for whether the two firms are in a related industry and the sizes of the targets and acquirers, the coefficient is statistically significantly different from zero. The positive coefficient indicates that cross-border acquisitions tend to have larger return differences between acquirers and targets. All regressions include country-specific dummy variables and standard errors are corrected for clustering of observations at the country level. In Columns 3 and 4 we break up the return differences into three components, the difference in returns of the two countries’ currencies ((Currency R12)j-i), the differences in local stock market indices ((Market R12)j-i), and the differences in firm-level excess returns relative to the market ((Firm USR12 - Currency R12 – Market R12)j-i). 23 The coefficients on all three variables are positive, but often statistically insignificantly different from zero. We then use a finer measure of local stock market index: industry index in local currency, which will not suffer from composition bias, where industry representation varies in different stock market indices. In Columns 5 and 6, we include difference in industry returns ((Industry R12)j-i), as well as differences in returns of the two countries’ currencies ((Currency R12)j-i) and the differences in firm-level excess returns relative to the industry ((Firm USR12 - Currency R12 – Industry R12)j-i). We find that the coefficient on the industry return is positive and statistically significant, suggesting that, in cross-border acquisitions, acquirers are from industries that outperform those of targets. 4.7.2. Identity of the Target and the Acquirer Another approach to evaluating the reasons for cross-border mergers is to consider the differences in the characteristics of targets and acquirers. If the underlying reason for the merger is to take advantage of valuation differences, then one ought to be able to predict which firms will be acquirers or targets using measures of valuation. Consequently, we consider the sample consisting of all firms involved in a public-to-public cross- border merger and estimate equations predicting whether a particular firm is a target or acquirer. Because the dependent variable is dichotomous, we estimate the equations by a logit model and present the marginal effects in 23 For the domestic deals, the differences in the local market returns and the currency returns equal zero by construction. 26 Table A.5. We estimate these equations for both domestic and cross-border mergers; the domestic mergers are in Columns 1-4 while the cross-border ones are in Columns 5-10. As in Table 7, we first break up the firm return differences into three components, the difference in returns of the two countries’ currencies, the differences in local stock market indices, and the differences in firm-level excess returns relative to the market (Columns 7 and 8). We then use a finer measure of stock market index to decompose firm-level stock returns into the difference in industry-level index return in local currency, the difference in currency returns, and the residual (Columns 3, 4, 9 and 10). All regressions include country dummies and standard errors are corrected for clustering of observations at the country level. The results in Table A.5 indicate that for both domestic and cross-border mergers, acquirers outperform targets prior to the acquisition. This finding is consistent with prior literature on domestic mergers suggesting that acquirers typically have higher valuations than targets. In Columns 7 and 8, we break down each return for the cross-border sample into three components, reflecting the local stock market index (in local currency), the currency return (relative to U.S. dollars), and the firm-specific residual in local currency. The results indicate that only the firm-specific component of returns is related to whether a firm is an acquirer or a target, not the local stock-market return or the currency return. When we use industry index in local currency to decompose firm-level returns, we find that in cross-border sample, acquirers are more likely from industries that outperform those of targets (Columns 9 and 10). We also find that there is no significant difference in currency returns between the target’s countries and the acquirer’s countries. These results are consistent with what we found at the country level using only public firms sample and similar to the deal-level regressions in Table 7 using the domestic/cross-border specification. The difference between the public firm subsample and the overall sample consisting mostly of private firms is consistent with the relative wealth story suggested by Froot and Stein (1991). The underlying cause of frictions in the Froot and Stein model is asymmentric information, which is likely to be higher in private firms than in public ones. Consequently, if this channel leads to wealth effects in mergers, then it should be stronger in mergers involving private firms than in mergers of public firms, consistent with the findings reported in Table A.5. 29 References Acharya, Viral, Hyun Song Shin, and Tanju Yorulmazer, 2010, Fire-sale FDI, NYU Working Paper. Aguiar, Mark and Gita Gopinath (2005), Fire-sale FDI and liquidity crises, The Review of Economics and Statistics, 87(3): 439-542. Ahern, Kenneth, Daniele Daminelli, and Cesare Fracassi, 2010, Lost in translation? The effect of cultural values on mergers around the world, University of Michigan Working Paper. Alquist, Ron, Rahul Mukherjee, and Linda Tesar (2010) “Fire sale FDI or business-as-usual,” University of Michigan Working Paper. Andrade, Gregor, Mark Mitchell, and Erik Stafford, 2001, New evidence and perspectives on mergers, The Journal of Economic Perspectives 15, 103-120. Baker, Malcolm, C. F. Foley, and Jeffrey Wurgler, 2009, Multinationals as abitrageurs: The effect of stock market valuations on foreign direct investment, Review of Financial Studies 22, 337-369. Bekaert, Geert, Campbell Harvey, and Christian Lundblad, 2005, Does Financial Liberalization Spur Economic Growth, Journal of Financial Economics 77, 3-55. Bekaert, Geert, Campbell Harvey, Christian Lundblad and Stephan Siegel, 2007, Growth Opportunities and Market Integration, Journal of Finance, 62, 1081-1138. Bris, Arturo, Neil Brisley, and Christos Cabolis, 2008, Adopting better corporate governance: Evidence from cross-border mergers, Journal of Corporate Finance 14, 224-240. Bris, Arturo, and Christos Cabolis, 2008, The value of investor protection: Evidence from cross-border mergers, Review of Financial Studies 21, 605-648. Chari, Anusha , Paige Ouimet, and Linda Tesar, 2009, The value of control in emerging markets, Review of Financial Studies23, 1741-1770 . Coeurdacier, Nicolas , Roberto A. DeSantis, and Antonin Aviat, 2009, Cross-border mergers and acquisitions and european integration, Economic Policy January, 55-106. Cushman, David O., 1987, The effect of real wages and labor productivity on foreign direct investment, Southern Economic Journal July, 174-185. Demirgüç-Kunt, A., and R. Levine, 2001. Financial structure and economic growth (MIT press). Desai, Mihir, C. Fritz Foley and James R. Hines Jr., 2001, Repatriation Taxes and Dividend Distortions, National Tax Journal 54, 829-851. Dewenter, Kathryn L., 1995, Do exchange rate changes drive foreign direct investment?, The Journal of Business 68, 405-433. Djankov, S., R. La Porta, F. Lopez-de-Silanes, and A. Shleifer (DLLS), 2008. The Law and Economics of Self- Dealing. Journal of Financial Economics 88, 430-465. 30 Dong, Ming, David Hirshleifer, Scott Richardson, and Siew Hong Teoh, 2006, Does investor misvaluation drive the takeover market?, The Journal of Finance 61, 725-762. Fama, E. , and K. French, 1998, Value versus growth: The international evidence, Journal of Finance 53, 1975-99. Fan, Joseph P.H., Randall Morck, Lixin Colin Xu, and Bernard Yeung, 2009, Does ‘good government’ draw foreign capital? Explaining china’s exceptional fdi inflow., World Development (Special Issue on Law, Finance and Economic Growth in China) forthcoming. Ferreira, Miguel A., Massimo Massa, and Pedro Matos, 2009, Shareholders at the gate? Institutional investors and cross-border mergers and acquisitions, Review of Financial Studies 23, 601-644. Froot, K. A., and J. C. Stein, 1991, Exchange rates and foreign direct-investment - an imperfect capital-markets approach, Quarterly Journal of Economics 106, 1191-1217. Graham, Edward M., and Paul R. Krugman, 1995. Foreign direct investment in the united states (Institute for International Economics, Washington, D.C.). Guiso, L., P. Sapienza, and L. Zingales, 2009. Cultural Biases in Economic Exchange. Quarterly Journal of Economics. 124, 1095-1131. Harford, Jarrad, 2005, What drives merger waves?, Journal of Financial Economics 77, 529-560. Jarrell, Gregg A., James A. Brickley, and Jeffry M. Netter, 1988, The market for corporate control: The empirical evidence since 1980, Journal of Economic Perspectives 2, 49-68. Jensen, Michael, and Richard S. Ruback, 1983, The market for corporate control: The scientific evidence Journal of Financial Economics 11, 5-50. Kaufmann, Daniel, Aart Kraay, and Massimo Mastruzzi, 2009, Governance matters viii: Governance indicators for 1996-2008, World Bank Policy Research June. Kedia, Simi, Venkatesh Panchapagesan, and Vahap Uysal, 2009, Geography and acquirer returns, Journal of Financial Intermediation, forthcoming. Kindleberger, C.P., 1969, American Business Abroad, (Yale University Press, New Haven). Klein, Michael W., and Eric S. Rosengren, 1994, The real exchange rate and foreign direct investment in the united states: Relative wealth vs. Relative wage effects, Journal of International Economics 36, 373-90. Klein, Michael W., Joe Peek, and Eric S. Rosengren, 2002, Troubled banks, impaired foreign direct investment: The role of relative access to credit, The American Economic Review 92, 664-682. Krugman, Paul, 1998, Fire sale FDI, available at http://web.mit.edu/krugman/www/FIRESALE.htm . Kumar, Praveen, and Latha Ramchand, 2008, Takeovers, market monitoring, and international corporate governance, The RAND Journal of Economics 39, 850-874. Makaew, Tanakorn (2010) “The dynamics of international mergers and acquisitions,” University of Maryland Working Paper. 31 Neary, J. Peter, 2009, Trade costs and foreign direct investment, International Review of Economics & Finance 18, 207-218. Netter, Jeffry M., Mike Stegemoller, and Babajide M. Wintoki, 2009, Are Small Mergers Important?: A study of over 250,000 (or less) mergers and acquisitions from 1992-2009, University of Georgia Working Paper Rose, Andrew K., 2000, One money, one market: estimating the effect of common currencies on trade, Economic Policy 30, 7–45 Rhodes-Kropf, Matthew, David T. Robinson, and S. Viswanathan, 2005, Valuation waves and merger activity: The empirical evidence, Journal of Financial Economics 77, 561-603. Rhodes-Kropf, Matthew, and S. Viswanathan, 2004, Market valuation and merger waves, Journal of Finance 59, 2685-2718. Rossi, Stefano, and Paolo F. Volpin, 2004, Cross-country determinants of mergers and acquisitions, Journal of Financial Economics 74, 277-304. Sarno, Lucio, Taylor, Mark P., and Jeffrey A. Frankel, 2003, The economics of Exchange Rates, (Cambridge University Press). Shleifer, A., and R. W. Vishny, 2003, Stock market driven acquisitions, Journal of Financial Economics 70, 295- 311. Stulz, René M., and Rohan Williamson, 2003, Culture, openness, and finance, Journal of Financial Economics 70, 313-349. 34 Panel B.1 Developing Targets, Developing Acquirers. (# of obs: 311) Panel B.2 Developing Targets, Developed Acquirers. (# of obs: 3,853) -0.25 -0.20 -0.15 -0.10 -0.05 0.00 0.05 0.10 0.15 0.20 -36-33-30-27-24-21-18-15-12 -9 -6 -3 0 3 6 9 12 15 18 21 24 27 30 33 36 Market Returns Currency Returns 0.00 0.10 0.20 0.30 0.40 0.50 0.60 -36 -33 -30 -27 -24 -21 -18 -15 -12 -9 -6 -3 0 3 6 9 12 15 18 21 24 27 30 33 36 Market Returns Currency Returns 35 Panel B.3. Developed Targets, Developing Acquirers. ((# of obs: 1,056) Panel B.4 Developed Targets, Developed Acquirers. ((# of obs: 46,288) -0.40 -0.30 -0.20 -0.10 0.00 0.10 0.20 -36-33-30-27-24-21-18-15-12 -9 -6 -3 0 3 6 9 12 15 18 21 24 27 30 33 36 Market Returns Currency Returns 0.000 0.005 0.010 0.015 0.020 0.025 0.030 -36 -33 -30 -27 -24 -21 -18 -15 -12 -9 -6 -3 0 3 6 9 12 15 18 21 24 27 30 33 36 Market Returns Currency Returns 36 Table 1. Number of mergers by country pair. The columns represent the countries of the acquiring companies while the rows represent those of the target companies. The diagonal entries of the matrix are therefore the number of domestic mergers for a particular country and the off-diagonal entries are the number of deals in a particular country pair. The totals exclude domestic mergers and hence represent the number of cross-border mergers to and from a particular country. Our sample period is from 1990 to 2007. Acquirer Country Destination AR AS AU BL BR CA CC CE CH CO CT CY DN FN FR GR HK HU ID IN IR IS IT JP LX MA MX NO NT NZ PE PH PL PO RU SA SG SK SP SW SZ TH TK TW UK US VE WG Total Argentina(AR) 201 1 4 4 30 42 13 1 5 57 1 1 2 6 17 4 16 1 28 3 1 2 54 5 14 1 58 243 4 22 640 Austria(AS) 341 4 8 19 1 19 7 35 3 1 2 2 7 1 25 7 8 1 7 25 2 1 2 1 5 19 33 1 52 84 255 637 Australia(AU) 3 4,875 7 2 145 1 20 7 62 43 5 16 24 6 8 69 2 51 1 10 64 145 5 1 1 58 75 5 5 40 47 2 1 2 430 812 63 2,238 Belgium(BL) 7 13 494 12 21 12 169 4 1 9 16 2 17 24 5 9 206 1 1 3 1 6 1 9 30 18 1 148 197 79 1,022 Brazil(BR) 40 3 14 9 565 48 15 4 14 6 94 3 6 8 8 41 18 6 19 9 28 3 35 4 5 52 16 22 1 1 58 388 60 1,038 Canada(CA) 1 10 59 14 7 6,220 1 8 12 11 112 2 16 3 11 13 9 19 58 11 4 4 13 54 5 1 3 1 9 1 3 6 34 56 4 328 2,516 1 80 3,500 Czech Republic(CC) 31 1 9 1 10 143 1 14 6 38 1 1 8 5 6 1 7 3 3 9 25 8 6 1 8 16 23 47 77 76 442 Chile(CE) 6 14 1 4 39 101 1 1 8 1 1 1 4 1 7 6 10 5 3 1 2 21 3 1 1 13 82 1 8 246 China(CH) 1 2 36 14 1 43 513 9 10 31 2 214 3 9 1 2 13 53 1 27 1 6 19 3 3 2 120 34 6 11 10 4 19 58 301 22 1,091 Colombia(CO) 1 2 3 17 3 37 1 8 1 4 2 10 1 1 2 2 13 3 4 6 35 1 1 121 Croatia(CT) 12 26 3 7 1 5 1 1 4 1 2 1 1 2 8 4 6 59 Cyprus(CY) 1 37 1 7 1 1 2 1 14 Denmark(DN) 5 4 10 9 1 1 1 889 39 39 1 2 3 11 3 11 8 2 1 94 38 1 1 1 4 4 198 28 117 173 80 890 Finland(FN) 9 7 7 16 69 1,614 34 2 1 2 22 2 11 23 4 53 24 7 1 5 11 281 31 1 2 1 60 147 41 874 France(FR) 1 20 28 236 7 116 4 68 38 4,837 8 13 2 12 27 13 164 97 28 1 1 22 209 8 2 6 5 2 87 116 154 2 1 708 970 434 3,610 Greece(GR) 1 3 4 6 6 339 7 1 2 1 6 1 2 3 5 15 18 9 90 Hong Kong(HK) 28 1 22 42 10 4 20 3 348 2 1 4 2 30 73 6 2 4 80 8 2 4 7 3 1 4 67 170 14 614 Hungary(HU) 28 4 3 2 2 2 5 28 3 136 2 2 4 11 5 2 4 42 5 2 2 2 11 14 1 1 26 69 52 334 Indonesia(ID) 10 1 15 1 2 1 9 98 4 1 16 19 2 4 1 2 24 8 2 7 4 1 26 32 7 199 India(IN) 5 24 3 15 1 5 2 39 8 1 764 1 2 12 16 17 5 19 1 1 3 6 11 7 6 19 28 3 101 233 43 637 Ireland-Rep(IR) 1 5 4 8 6 1 18 2 2 4 354 1 3 5 1 1 1 6 10 1 1 3 3 2 2 5 5 1 1 265 172 15 555 Israel(IS) 2 2 10 3 8 1 1 1 160 2 1 3 3 1 1 3 3 1 1 26 171 12 256 Italy(IT) 24 17 23 6 24 3 24 19 236 13 10 1 1 10 9 9 1,633 29 16 2 7 98 1 2 5 5 2 38 60 65 1 233 428 164 1,585 Japan(JP) 1 5 6 4 4 2 31 3 12 2 1 1 3 1 5,698 1 16 1 1 2 1 12 2 11 8 1 9 46 259 42 488 Luxembourg(LX) 1 16 8 2 3 8 5 1 1 1 4 2 12 13 14 83 Malaysia(MA) 17 6 6 2 7 17 2 2 1 1 19 1,711 1 2 7 4 4 1 2 123 1 2 2 8 4 28 43 9 321 Mexico(MX) 2 4 3 6 116 4 1 2 8 2 24 3 4 4 6 7 1 188 1 18 3 1 1 2 1 35 10 8 1 33 320 1 18 650 Norway(NO) 3 4 5 2 8 1 102 55 27 1 1 2 2 5 3 4 2 1 688 30 1 1 1 2 1 2 193 21 103 130 26 739 Netherlands(NT) 16 26 95 5 44 2 1 1 36 27 122 3 9 1 9 60 9 29 39 9 4 2 18 1,512 2 2 2 1 9 3 16 66 47 1 2 3 441 436 228 1,826 New Zealand(NZ) 2 302 1 41 2 4 3 8 5 1 2 7 19 13 2 14 570 1 6 13 2 5 7 3 71 140 6 680 Peru(PE) 1 2 1 3 54 6 1 1 1 2 3 2 39 5 3 7 30 1 123 Philippines(PH) 10 9 6 5 11 10 2 1 115 12 1 1 2 2 13 32 2 119 Poland(PL) 9 1 11 12 4 2 26 16 45 3 5 6 12 3 19 2 13 16 39 227 4 2 2 1 2 13 23 10 57 76 63 497 Portugal(PO) 1 3 5 5 5 1 8 1 44 1 1 2 2 2 11 2 1 1 4 13 246 72 12 12 51 40 20 320 Russian Fed(RU) 4 2 7 24 2 8 9 23 10 5 3 2 2 11 3 9 14 18 1 6 526 4 25 15 3 59 83 30 382 South Africa(SA) 36 2 35 4 4 23 2 3 7 3 1 7 11 4 5 2 11 2 1 2 1 790 3 1 15 15 1 170 113 36 520 Singapore(SG) 28 1 7 1 1 7 4 16 35 10 22 2 5 25 98 10 8 1 1 3 614 2 1 9 7 9 6 42 116 15 492 South Korea(SK) 2 3 4 10 1 4 1 27 4 2 4 30 2 1 2 10 1 1 6 631 2 3 8 2 27 107 30 294 Spain(SP) 2 6 8 32 4 27 5 2 1 44 8 296 8 3 7 10 4 121 29 2 6 15 133 1 60 1 1 1 1,896 60 35 271 287 169 1,659 Sweden(SW) 9 10 14 24 176 198 67 3 5 4 10 4 11 23 9 1 182 65 1 1 2 3 2 1,558 31 2 218 288 1 99 1,463 Switzerland(SZ) 43 9 31 20 1 1 38 15 122 5 1 7 10 15 36 16 6 3 8 49 1 2 5 4 6 4 45 794 2 103 261 311 1,180 Thailand(TH) 5 2 2 1 2 1 9 10 1 7 1 36 1 23 1 6 2 1 1 2 37 1 3 2 194 4 22 40 9 232 Turkey(TK) 1 1 6 4 1 2 3 16 5 1 2 2 2 10 2 1 9 1 4 1 3 4 2 72 27 33 27 170 Taiwan(TW) 4 5 2 1 5 11 1 13 2 2 21 5 4 1 2 130 10 82 10 181 United Kingdom(UK) 29 177 91 2 305 3 1 2 4 158 66 485 17 60 2 2 82 410 20 92 173 21 21 2 89 313 14 1 2 8 10 86 42 9 43 206 126 7 2 5 15,196 3,122 443 6,753 United States(US) 10 36 392 121 35 2,752 1 8 34 6 1 128 130 719 28 95 1 10 179 316 169 146 827 28 24 73 86 453 28 13 1 5 21 75 104 54 91 351 358 9 5 68 3,073 66,948 5 817 11,886 Venezuela(VE) 1 1 4 22 1 2 11 5 1 2 2 7 5 8 49 16 1 122 Germany(WG) 234 42 105 6 80 5 2 5 1 124 134 454 13 20 3 2 29 38 16 128 111 40 8 5 42 443 1 10 6 3 5 18 11 11 37 194 375 1 5 3 724 1,611 1 5,771 5,106 Total 66 557 1,360 919 133 4,236 19 60 112 16 12 23 1,199 866 3,634 150 633 40 45 473 1,044 324 1,027 1,874 242 416 160 758 2,588 226 7 51 41 142 90 333 729 171 675 2,127 1,686 54 29 145 8,468 15,034 15 3,969 56,978 39 Table 4. Panel analysis of the effect of valuation differences on cross-border M&As: Interactions with economic development, distance, and capital account openness. This table presents estimates of panel regressions of cross-border mergers and acquisitions. The dependent variable is the number of cross-border deals in year t (Xijt) in which the target is from country i and the acquirer is from country j (where i ≠ j) scaled by sum of the number of domestic deals in target country i (Xiit) and the number of the cross-border deals involving target country i and acquirer j (Xijt). Columns (1) and (2) present the interaction of valuation differences with the relative wealth of acquiring versus target country. The indicator variable equals one if the GDP of the acquirer country is larger than the GDP of the target country. Column (3) and (4) present the interaction of valuation results by the geographic distance between target and acquiring country. The indicator variable takes on a value of one if the distance between the capitals of the target and the acquirer countries is below the median (4,272 miles). Column (5) and (6) present the interaction of valuation differences with the target country’s capital account openness (Quinn, 1997). The indicator variable is one if the capital-account-openness measure (Quinn 1997) is below the median (0.68). Refer to Table A1 for the definition of variables. Country-pair and year fixed effects are included in all regressions. Standard errors are corrected for clustering of observations at the country-pair level and associated t-statistics are in parentheses. The symbols ***, **, and * denote statistical significance at the 1%, 5% and 10% levels. GDP (acquirer) > GDP (target) Below-median distance Below-median Capital Account Openness (1) (2) (3) (4) (5) (6) (Currency R12)j-i 0.002 0.018* 0.017*** (0.41) (1.85) (3.08) (Market R12)j-i 0.003 0.013*** 0.003 (1.32) (3.03) (1.18) (Market MTB)j-i -0.000 0.004*** 0.002*** (-0.01) (3.05) (2.75) (Currency R12)j-i × Indicator 0.052*** 0.037* 0.026 (3.24) (1.76) (1.48) (Market R12)j-i × Indicator 0.014** -0.005 0.018*** (2.51) (-0.81) (2.60) (Market MTB)j-i × Indicator 0.008*** 0.001 0.004** (4.25) (0.31) (1.99) Max (Import, Export) 0.178** 0.154** 0.184** 0.160** 0.179** 0.159** (2.48) (2.39) (2.57) (2.48) (2.51) (2.45) (log GDP per capita)j-i 0.042*** 0.021* 0.042*** 0.021** 0.042*** 0.021* (3.50) (1.95) (3.50) (1.96) (3.45) (1.95) (GDP Growth)j-i 0.003 0.056* -0.001 0.059* 0.000 0.059* (0.09) (1.83) (-0.03) (1.87) (0.01) (1.90) (Quality of Institution)j-i -0.001 -0.001 -0.001 -0.001 -0.001 -0.001 (-1.02) (-1.18) (-1.00) (-1.20) (-0.97) (-1.16) (Investment Profile)j-i -0.000 -0.000 -0.000 -0.000 -0.000 -0.001 (-0.24) (-0.60) (-0.12) (-0.61) (-0.27) (-0.66) Constant 0.076*** 0.034*** 0.076*** 0.034*** 0.076*** 0.035*** (7.69) (6.38) (7.67) (6.35) (7.69) (6.43) Observations 14,857 14,715 14,857 14,715 14,857 14,715 R-squared 0.497 0.512 0.496 0.512 0.497 0.512 40 Table 5. Panel analysis of the intensity of cross-border M&As: Robustness checks This table presents estimates of panel regressions of cross-border mergers and acquisitions. The dependent variable is the number of cross-border deals in year t (Xijt) in which the target is from country i and the acquirer is from country j (where i ≠ j) scaled by sum of the number of domestic deals in target country i (Xiit) and the number of the cross-border deals involving target country i and acquirer j (Xijt). Columns (1) through (3) examine subsamples of cross-border deals based on the ownership stake the acquiring firm obtains. Columns (4) through (6) examine subsamples of cross-border deals based on the method of payment: a deal is classified as a cash (stock) deal if more than 50% of the deal is paid by cash (stock). Column (7) examines the dollar value of all cross-border deals. Columns (8) and (9) examine subsamples of deals without information on deal value and those with deal value information respectively. Column (10) examines the sample of cross-border deals including withdrawn (failed) ones. Column (11) examines the subsample of countries for which exchange rate is not pegged, defining an exchange rate as pegged if the absolute values of the bilateral nominal exchange rate returns are less than 0.001 for each of 12 consecutive months. Column (12) deploys a gravity model (Rose, 2000) for bilateral cross-border mergers. Columns (13) and (14) include exchange rate volatility and interest rate differences between target and acquiring countries respectively. Refer to Table A1 for the definition of variables. Country-pair and year fixed effects are included in all regressions. Standard errors are corrected for clustering of observations at the country-pair level and associated t-statistics are in parentheses. The symbols ***, **, and * denote statistical significance at the 1%, 5% and 10% levels. 5-49% 50-99% 100% Cash Deals Stock Deals Payment missing or mixed $ Value # of Deals w/o $ Value # of Deals with $ Value Failed deals included Exclude countries with pegged EX Rose (2000)'s Gravity model Control for exchange rate volatility Control for interest rate (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (Currency R12)j-i 0.026*** 0.034*** 0.036*** 0.029** 0.017 0.029*** 0.008 0.046*** 0.019** 0.044*** 0.031*** 0.030** 0.029*** 0.040*** (2.96) (3.09) (3.58) (2.32) (1.63) (3.24) (0.79) (4.07) (2.47) (4.69) (3.34) (2.34) (3.15) (3.83) (Market R12)j-i 0.003 0.014*** 0.009** 0.005 0.004 0.010*** -0.000 0.020*** 0.002 0.010*** 0.013*** 0.011*** 0.011*** 0.012*** (0.56) (3.23) (2.41) (0.90) (0.91) (3.21) (-0.04) (5.31) (0.63) (3.56) (3.71) (2.99) (3.31) (3.41) Maximum (Import, Export) 0.243** -0.048 0.244*** 0.346*** 0.139 0.156** 0.227** 0.113 0.283*** 0.184*** 0.218** 0.620*** 0.191*** 0.228*** (2.06) (-0.36) (3.51) (2.68) (1.15) (2.32) (2.29) (1.56) (3.66) (2.83) (2.38) (9.77) (2.66) (2.72) (Quality of Institution)j-i 0.048*** 0.031** 0.048*** 0.016 -0.041* 0.042*** 0.024* 0.037*** 0.045*** 0.036*** 0.048*** 0.002 0.040*** 0.042*** (2.87) (1.99) (3.76) (0.77) (-1.66) (3.52) (1.69) (3.03) (3.46) (3.36) (3.72) (1.46) (3.38) (2.82) (Investment Profile)j-i 0.101** -0.006 -0.013 0.093 -0.100* 0.021 0.077* -0.014 0.059* -0.030 -0.009 -0.001 0.004 -0.026 (2.46) (-0.12) (-0.33) (1.57) (-1.95) (0.59) (1.93) (-0.32) (1.88) (-0.97) (-0.26) (-0.94) (0.13) (-0.61) ( log GDP per capita)j-i -0.001 0.001 -0.002** 0.000 0.000 -0.001 -0.001 -0.000 -0.001 -0.001 -0.001 0.003*** -0.001 -0.001 (-0.49) (0.90) (-2.02) (0.23) (0.12) (-1.03) (-1.01) (-0.37) (-0.74) (-0.96) (-1.28) (10.84) (-1.07) (-1.13) (GDP growth)j-i -0.002 -0.001 -0.000 -0.004*** 0.002* 0.000 0.001 -0.000 0.001 -0.000 -0.000 -0.000 -0.001 0.000 (-1.55) (-0.83) (-0.45) (-3.53) (1.72) (0.12) (0.76) (-0.02) (1.17) (-0.16) (-0.27) (-0.14) (-0.78) (0.43) Log (distance) -0.006*** (-4.73) Currency Volatility 0.273*** (4.51) (Interest Rate)j-i -0.000 (-1.37) Constant 0.067*** 0.063*** 0.074*** 0.094*** 0.011 0.073*** 0.049*** 0.075*** 0.058*** 0.062*** 0.073*** 0.006 0.068*** 0.075*** (5.73) (5.50) (6.84) (4.04) (1.18) (7.63) (4.09) (7.67) (5.43) (8.65) (7.24) (0.14) (6.66) (7.32) Observations 13,964 13,846 14,613 11,631 9,065 14,833 14,301 14,567 14,301 15,009 13,806 7,949 14,857 11,774 R-squared 0.354 0.342 0.481 0.447 0.483 0.469 0.363 0.438 0.489 0.555 0.495 0.108 0.497 0.509 41 Table 6. Explaining the effect of valuation differences on cross-border mergers and acquisitions. This table presents estimates of panel regressions of cross-border mergers and acquisitions. The dependent variable is the number of cross-border deals in year t (Xijt) in which the target is from country i and the acquirer is from country j (where i ≠ j) scaled by sum of the number of domestic deals in target country i (Xiit) and that of the cross-border deals involving target country i and acquirer country j (Xijt). We decompose market-to-book ratio of each country using future stock market returns and future exchange rate returns (Baker et al., 2009). Based on our estimates, Fitted MTB = 2.017 - 0.033 FR12 - 0.137 FR24 - 0.299 FR36 - 0.255 EXFR12 - 0.247 EXFR24 + 0.487 EXFR36 (N=642, R2=0.094). Columns (1) and (2) examine the entire sample of cross-border deals. Columns (3) through (10) examine subsamples of deals, in which various combinations of public status of the parties are selected and then aggregated to the country level. Refer to Table A1 for the definition of variables. Country-pair and year fixed effects are included in all regressions. Standard errors are corrected for clustering of observations at the country-pair level and associated t-statistics are in parentheses. The symbols ***, **, and * denote statistical significance at the 1%, 5% and 10% levels. All Private Target-Private Acquirer Private Target- Public Acquirer Public Target-Private Acquirer Public Target-Public Acquirer (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (Currency FR12)j-i 0.017*** 0.015** 0.011 0.014 0.012 (2.74) (2.17) (1.14) (1.16) (0.78) (Market FR12)j-i -0.001 0.001 -0.006 -0.012** 0.003 (-0.42) (0.39) (-1.04) (-2.04) (0.51) (Fitted MTB)j-i 0.002 0.006** -0.006 0.006 -0.001 (0.64) (2.07) (-1.03) (0.81) (-0.11) (Residual MTB)j-i 0.006*** 0.005*** 0.007*** 0.004* 0.007** (5.34) (4.01) (4.47) (1.77) (2.37) Maximum (Import, Export) 0.186*** 0.124* 0.084 0.079 0.281*** 0.191** -0.014 -0.063 0.009 0.063 (2.94) (1.94) (1.39) (1.32) (3.17) (2.09) (-0.09) (-0.33) (0.07) (0.48) ( log GDP per capita)j-i 0.041*** 0.015 0.028*** 0.008 0.048*** 0.028** -0.000 -0.005 0.008 0.001 (3.82) (1.45) (2.88) (0.72) (3.68) (2.06) (-0.02) (-0.27) (0.32) (0.02) (GDP growth)j-i 0.049* 0.079*** 0.053* 0.061** 0.087* 0.147*** -0.001 0.033 0.084* 0.031 (1.68) (2.71) (1.82) (2.11) (1.82) (2.87) (-0.02) (0.70) (1.75) (0.62) (Quality of Institution)j-i -0.000 -0.000 -0.000 -0.001 -0.001 -0.002 -0.001 0.000 0.003* 0.003* (-0.46) (-0.59) (-0.53) (-0.63) (-1.16) (-1.43) (-0.51) (0.23) (1.89) (1.78) (Investment Profile)j-i -0.000 -0.001* -0.000 -0.001* 0.001 -0.002* 0.002 0.001 -0.002* -0.003* (-0.39) (-1.79) (-0.63) (-1.89) (0.60) (-1.69) (1.50) (0.59) (-1.84) (-1.87) Constant 0.069*** 0.033*** 0.042*** 0.021*** 0.085*** 0.060*** 0.011 0.021* 0.039*** 0.028** (8.19) (6.33) (5.42) (4.46) (8.83) (8.30) (0.74) (1.70) (3.51) (2.08) Observations 14,300 12,590 13,729 12,110 13,707 12,143 6,817 6,112 7,820 6,999 R-squared 0.487 0.519 0.329 0.336 0.543 0.548 0.315 0.311 0.348 0.381 44 (Quality of Institution)j-i The sum of the International Country Risk Guide (ICRG) Political Risk (ICRGP) subcomponents: Corruption, Law and Order, and Bureaucratic Quality. Details on these subcomponents can be found in Bekaert, Harvey and Lundblad (2005) Table 1. (Investment Profile)j-i ICRG Political Risk (ICRGP) sub-component. It is a measure of the government’s attitude to inward investment. It is determined by Political Risk Service’s assessment of three sub-components: (i) risk of expropriation or contract viability; (ii) payment delays; and (iii) repatriation of profits. Each sub-component is scored on a scale from zero [very high risk] to four [very low risk]. Currency Volatility The standard deviation of the first-difference of the monthly natural logarithm of the bilateral nominal exchange rate in the five years preceding year t. (Source: Datastream) Pegged Exchange Rate A country-pair is classified as pegged exchange rate if the absolute value of the bilateral nominal exchange rate returns is less than 0.001 for 12 months continuously. (Source: Datastream) (Interest rate)j- i The differences between acquirer (j) and target (i) firm’s country of domicile in the deposit interest rate. (Source: World Bank Development Indicators) Panel B Deal-level variables Value of transaction Total value of consideration paid by the acquirer, excluding fees and expenses, adjusted to 2008 constant dollars using U.S. city average consumer price index (CPI-U) published by Bureau of Labor Statistics. (Source: SDC Mergers and Corporate Transactions database) Public target (acquirer) Target (Acquirer) is a public firm if its public status is “Public” or if its SEDOL is non-missing. (Source: SDC Mergers and Corporate Transactions database) Cross-border deal A deal occurs cross-border if target nation is different from that of acquirer’s ultimate parents. (Source: SDC Mergers and Corporate Transactions database) Related industry Target firms are in the same industries as acquirers if any line of business the target firm is in (TSIC2) overlaps with that of acquirer (ASIC2). (Source: SDC Mergers and Corporate Transactions database) Same Region A dummy variable equals 1 if the acquirer (j) and target (i) firm’s country of domicile is located in the same broadly defined continent (Africa, America, Asia, Europe). (Source: World Factbook) 5-49% stakes (50-99%, 100%) Dummy variable equals 1 if the number of common shares acquired in the transaction plus any shares previously owned by the acquirer divided by the total number of shares outstanding is between 5 and 50 percent. (Source: SDC Mergers and Corporate Transactions database) Cash (Stock) Deals A deal is classified as a cash (stock) deal if more than 50% of the deal value is paid in cash (stock). (Source: SDC Mergers and Corporate Transactions database) Failed Deals Dummy variable equals 1 if the deal is withdrawn. (Source: SDC Mergers and Corporate Transactions database) (Firm USR12)j-i The difference between the acquirer (j) and target firm (i) in annual real stock market return in US$. We obtain total return indices in US$ for all public firms (Datastream code: RI) and deflate these indices using the 2000 Consumer Price Index (CPI) in US$ to calculate real stock returns. (Source: Datastream) (Industry R12)j-i The difference between the acquirer (j) and target firm (i)’s primary industry in the annual local real stock market return. We calculate value-weighted annual local real stock market return for 48 Fama-French industries of each country. (Source: Datastream and Professor Kenneth French’s website at Dartmouth University, http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/index.html) Total assets (log) Book value of total assets in millions of constant 2000 U.S. dollars (WC07230). (Source: Worldscope) Return on assets (Net Income before Preferred Dividends + ((Interest Expense on Debt—Interest Capitalized ) * (1-Tax Rate)) / Average of Last Year's and Current Year's (Total Capital + Last Year's Short Term Debt & Current Portion of Long Term Debt) * 100 (WC08376). (Source: Worldscope) Long-term debt/assets Ratio of long-term debt to book value of assets (WC03251/WC02999). (Source: Worldscope) Sales growth Two-year local country CPI inflation-adjusted sales growth (WC01001). (Source: Worldscope) Cash/Asset Ratio of cash and liquid assets to book value of assets (WC02001/WC02501). (Source: Worldscope) 45 Table A.2. Subsample analysis of the cross-sectional determinants of cross-border mergers and acquisitions. This table presents estimate of cross-sectional regressions of cross-border M&A country pairs in subsamples of developed and emerging country targets. The dependent variable is the total number of cross-border deals between 1990 and 2007 (Xij) in which the target is from country i and the acquirer is from country j (where i ≠ j) scaled by sum of the number of domestic deals in target country i (Xii) and the number of cross-border deals between country i and country j (Xij). Columns (1) and (3) examine subsample of developed country targets while Columns (2) and (4) examine emerging country targets. Refer to Table A1 for the definition of variables. Acquirer country fixed effects are included in all regressions. Heteroskedasticity-corrected t-statistics are in parentheses. The symbols ***, **, and * denote statistical significance at the 1%, 5% and 10% levels. Developed Target Emerging Targets Developed Target Emerging Targets (1) (2) (3) (4) (Disclosure Quality)j-i 0.008*** 0.025*** 0.010*** 0.028*** (4.57) (4.47) (5.54) (3.28) (Legal)j-i -0.028 -0.273 -0.259*** -0.061 (-0.36) (-1.07) (-2.94) (-0.23) Average (Currency R12)j-i 0.063 0.002 (1.65) (0.02) Average (Market R12)j-i -0.250*** 0.053 (-4.43) (0.47) Same Language -0.005 0.020 (-0.57) (0.89) Same Religion -0.004 0.001 (-1.18) (0.09) Geographic Proximity 0.002* 0.013*** (1.91) (4.81) Max (Import, Export) 0.623*** 0.705*** 0.602*** 0.415*** (4.17) (7.61) (3.90) (4.54) (log GDP per capita)j-i 0.000 -0.007 -0.007* -0.021*** (0.12) (-1.27) (-1.95) (-3.46) (GDP Growth)j-i -0.002 0.008 0.000 0.005 (-1.35) (1.63) (0.41) (0.80) Constant 0.014*** 0.039*** 0.022*** 0.135*** (3.47) (3.29) (3.35) (5.16) Observations 662 231 662 231 R-squared 0.669 0.746 0.699 0.800 46 Table A.3. Summary statistics on valuation differences between target and acquirer. This table reports average [median] valuation differences between the acquirer (j) and the target (i) at the country and firm level. R12, R24, R36 represent real stock returns over the past 12 months, 24 months, 36 months, respectively. MTB is the market-to-book ratio of equity. For market MTB, we follow Fama and French (1998) and sum the market value of all firms within a country and divide this sum by the sum of their book value. All stock returns (both market and firm-level returns) are in local currency. The definition of developed countries is based on World Bank high-income economies. Developing Target Developed Target (1) (2) (3) (4) (5) Total Developing Acquirer Developed Acquirer Developing Acquirer Developed Acquirer Currency R12j –Currency R12i 1.12%*** 4.57%*** 10.32%*** (5.96%)*** 0.46%*** [0.26%]*** [1.24%] [6.18%]*** [(3.68%)]** [0.11%]*** Currency R24j – Currency R24i 2.13%*** 5.72%** 21.76%*** (13.40%)*** 0.79%*** [0.47%]*** [6.23%]** [18.28%]*** [(8.69%)]** [0.08%] Currency R36j – Currency R36i 3.43%*** 10.11%*** 34.22%*** (23.32%)*** 1.38%*** [0.91%]*** [16.79%]*** [31.37%]*** [(18.75%)]* [0.14%] Market R12j –Market R12i 0.30%*** 1.44% 0.05% 6.03%*** 0.20%** [0.33%]*** [(0.40%)] [(3.86%)]** [8.68%]*** [0.44%]*** Market R24j –Market R24i 0.92%*** 1.57% 2.13%*** 11.09%*** 0.64%*** [1.10%]*** [4.88%] [(1.90%)] [15.24%]*** [1.08%]*** Market R36j –Market R36i 2.12%*** 1.44% 12.79%*** 9.54%*** 1.22%*** [2.45%]*** [3.61%] [17.06%]*** [18.42%]*** [2.03%]*** Firm R12j –Firm R12i 10.38%*** 25.82%* 6.59%** 22.36%** 10.50%*** [6.01%]*** [27.80%]** [0.25%] [16.88%]*** [5.98%]*** Firm R24j –Firm R24i 19.34%*** 35.75% 11.96%** 41.81%** 19.61%*** [12.15%]*** [49.51%] [1.70%] [46.96%] [12.62%]*** Firm R36j –Firm R36i 23.36%*** 115.8%* 20.37%*** 63.13%*** 23.02%*** [17.02%]*** [116.2%] [8.98%]* [52.34%]*** [16.96%]*** Market MTBj – Market MTBi 9.93%*** 20.58%*** 64.74%*** (2.90%) 5.59%*** [7.25%]*** [31.19%]*** [64.45%]*** [(5.31%)]** [5.98%]*** Firm MTBj – Firm MTBi 28.95%*** 76.90% 47.03%** 17.27% 27.50%*** [26.23%]*** [125.7%]** [32.60%]** [(20.42%)] [25.91%]*** N. of Observations 51488 311 3853 1056 46268