Diversification and Shareholders’ Wealth during the 2008-2010 Financial Crisis
MSc. BA Finance Thesis
Paul van de Wiel
Student at the University of Groningen, Faculty of Economics and Business
July 2012
Supervisor: Dr. H. Gonenc
2nd Supervisor: Dr. V. Angelini
2
Diversification and Shareholders’ Wealth during the 2008-2010 Financial Crisis
PAUL VAN DE WIEL*
ABSTRACT
In this study, I investigate the value creation of corporate diversification in a trade-off model based on different pre-bid levels of acquirer diversity. I hypothesize that there exists a trade-off between the benefits and costs of diversification, predicting a positive relationship between the acquirer’s pre-bid level of diversity and acquirer gains from diversifying acquisitions during the 2008-2010 financial crisis.
Unexpectedly, I do not find support that there exists a positive relation. Thus, during the recent crisis, acquirers with higher levels of pre-bid diversity do not earn significant higher gains from diversifying acquisitions compared to acquirers with lower levels of pre-bid diversity.
Keywords: Acquisitions, Acquirer gains, Diversification, Financial crisis
JEL classification: G31, G32, G34
*
The author is grateful for helpful suggestions from Dr. H. Gonenc, lecturer at the University of Groningen, on an earlier draft of this paper. All remaining errors and omissions are the authors’ own.3 1. Introduction
Corporations diversify their operations across multiple product segments (industrial diversification)
1, across different countries (global diversification)
2, or both. Although many firms decreased their level of industrial diversity during the last decade, global diversification has increased over time (Denis et al., 2002). The integration of world markets has increased dramatically and globalization is an important strategic issue for corporations (Moeller and Schlingemann, 2005).
Nevertheless, in 2008-2010, the world experienced a global financial meltdown which has led to enormous changes in the acquisition market. While destructive and disastrous, the recent financial crisis serves as an ideal natural experiment for studying the impact of a credit-constrained environment on the performance of diversifying acquisitions (Rudolph and Schwetzler, 2011). Therefore, this paper examines the relationship between diversification and firm value during the 2008-2010 financial crisis, by using a trade-off model based on different pre-bid levels of acquirer diversity.
In general, the effect of corporate diversification on firm value is still a hot debate. The past few years have seen a growing literature on the relation between diversification and firm value, although empirical evidence remains largely mixed. Various researchers analyze this relationship by using the imputed value method. Most of them show a ‘diversification discount’ for industrially diversified firms (e.g. Lang and Stulz, 1994; Berger and Ofek, 1995; Denis et al., 2002). These studies focus on the cross- sectional relation between the market value of a firm and the sum of the imputed values for its segments as stand-alone domestic entities (Freund et al., 2007). Recently, researchers have turned to alternative valuation methods for measuring the effects of corporate diversification because of problems associated with the imputed value method
3. A complementary approach that has not received as much attention is to study the market’s response to diversification announcements using event study methods (Akbulut and Matsusaka, 2010). Also, the joint impact of industrial and global diversification on acquirer stock returns is hitherto largely unanalyzed, albeit studies controlled for the industrial diversification effect on cross-
1 Industrial diversification is also referred to as ‘unrelated acquisition’.
2 Another synonym for global diversification is cross-border acquisition, international diversification or geographic diversification. However, since geographic diversification could imply both domestic and global dimensions, I do not use this term to avoid possible ambiguity.
3 A number of studies now suggest that the observed diversification discount using the imputed value method is either not due to diversification at all, or may be result of improper measurement techniques (Martin and Sayrak, 2001; Freund et al., 2007). For example, Graham et al. (2002) find that the imputed value method might lead to a selection bias and Denis et al. (2002) admit that their diversification discount could be an endogeneity problem because firms with discounted assets might be more inclined to diversify. Campa and Kedia (2002) and Villalonga (2004) confirm this causality by showing that their sample of acquirers is already trading at a discount prior to the diversification.
4 border acquisitions (for example, Francis et al., 2008 and Freund et al., 2007). Thus, the question whether corporate diversification is shareholder value enhancing or not, remains unanswered.
Moreover, while economic theory suggests both positive and negative impacts of diversification on firm value (Bodnar et al., 1998), most existing literature appears to have overlooked the possibility of viewing the benefits and costs of diversification together. To my best knowledge, Ekkayokkaya and Paudyal (2010) were the first investigating the value creation of corporate diversification in a trade-off model based on acquirers’ existing (pre-bid) level of industrial diversity. They find that diversification gains vary across different pre-bid levels of industrial diversity. Validity of these findings for other time periods and in the context of international diversification should yield fruitful research.
This paper therefore addresses the question how the recent financial crisis affected the relationship between the acquirer’s pre-bid level of industrial and global diversity, and acquirer gains from all types of diversifying acquisitions. Diversifying acquisitions can be industrial, global, or both. I hypothesize that the relation between the acquirer’s pre-bid level of diversity and acquirer gains from diversifying acquisitions during the 2008-2010 financial crisis, resembles a positive linear shape.
Accordingly, this hypothesis predicts higher acquirer gains from diversifying acquisitions for higher levels of pre-bid diversity. Hence, this paper builds on the study by Ekkayokkaya and Paudyal (2010), extending it to a global scale. Furthermore, I follow Brown and Warner’s (1985) event study methodology to examine announcement stock returns for all types of diversifying acquisitions.
Using a sample of 220 U.K. acquisitions that were announced during the recent financial crisis
4, I do not find that there exists a positive relation between acquirers’ existing (pre-bid) level of industrial and global diversity, and announcement returns from diversifying acquisitions during the 2008-2010 financial crisis. This suggests that, during the recent crisis, acquirers with higher levels of pre-bid diversity do not earn significant higher gains from diversifying acquisitions compared to acquirers with lower levels of pre-bid diversity.
My results also provide evidence that diversifying acquisitions create value for acquirer shareholders. Acquirers experience, on average, a return of 3.62% when making an industrial diversifying acquisition. This finding is robust to various short-term event windows and consistent with existing evidence. In addition, I find a positive return of 1.44% for acquirers making an acquisition that is both
4 By exploiting the 2008-2010 financial crisis as a natural experiment to analyze how diversifying acquisitions affect acquirer gains, potential concerns about endogeneity are mitigated (Giannetti and Laeven, 2009).
5 industrial and global diversifying. The latter result is however insignificant after controlling for other factors. Altogether, I infer that diversifiers can profit more from diversification during the 2008-2010 financial crisis, because of the ‘more-money’ effect arising from debt coinsurance features and improved efficiency of internal capital markets in times of tightened external capital markets.
This paper makes several important contributions to literature. First, this paper is the first to investigate the impact of the recent financial crisis on acquirer gains from all types of diversifying acquisitions
5. Hence, I contribute to the ongoing discussion on the real effects of the 2008-2010 crisis.
Second, evidence is provided on the relationship between the acquirer’s existing (pre-bid) level of global diversity and announcement stock returns from all types of diversifying acquisitions, by proposing a different approach to measure acquirers’ pre-bid level of global diversity: The number of countries in which an acquirer has recorded subsidiaries. The number of foreign countries distinguishes the international involvement of a firm (Stopford and Wells, 1972) and this measure might overcome potential data limitations observed in existing studies. I thereby complement the study by Ekkayokkaya and Paudyal (2010) and offer alternative evidence to the single-segment vs. multi-segment classification adopted by Lang and Stulz (1994) and Servaes (1996).
Third, I further extend literature by examining the value impact of acquisitions that are both industrial and global diversifying. In addition, a comprehensive overview is provided with the most well- known cited benefits and costs of both forms of corporate diversification.
Lastly, it has been argued that acquirer gains differ among time periods and countries. There exists a large body of evidence regarding U.S. acquisitions during the period 1960-1990, while there is hardly any evidence of recent European acquisitions. Therefore, using a sample of U.K. firms, this paper contributes to the European acquisition literature.
This paper proceeds as follows. Section 2 describes the literature background on corporate diversification and presents the research hypotheses. Section 3 reviews prior empirical evidence on acquirer gains from diversifying acquisitions. Section 4 describes the data and empirical methods. The empirical results are then reported and discussed in section 5. Section 6 concludes this study. At last, section 7 discusses the limitations of this paper together with suggestions for future research.
5 To the best of my knowledge, no other paper examines the effects of the 2008-2010 financial crisis on acquirer gains from different types of diversifying acquisitions.
6 2. Literature Background and Hypothesis Development
Although the positive and negative consequences of corporate diversification are well documented, hardly any literature exists as to their joint effect. A remarkable feature of existing studies is that most of them focus on either industrial or global diversification. In this section, a comprehensive overview is provided for the most well-known cited benefits and costs of both industrial and global diversification. Together, they form the basis for my trade-off hypothesis of corporate diversification.
2.1 Benefits of Corporate Diversification
Diversifying acquisitions have been theorized to provide several benefits to acquirers’
shareholder wealth. Most of the presumed benefits of corporate diversification fall within the scope of three main theoretical perspectives that can be used to explain why a firm might choose to diversify: the market-power view, the resource-view and the efficiency view. The market-power view argues that diversified firms have (anti-competitive) advantages over non-diversified firms because they have access to ‘conglomerate power’ (Hill, 1985; Montgomery, 1994). The resource-view assumes that firms diversify in response to excess capacity in resources and that the benefits of diversification are a function of a firm’s resource stock (Penrose, 1959; Montgomery, 1994). The efficiency theory views acquisitions as being executed to achieve different kinds of synergies
6. Although all views are consistent with profit maximization, only the market-power view is inconsistent with the efficient use of resources (Montgomery, 1994). Below, I give an extensive overview of the benefits of diversification that accrue to acquirer shareholders by first explaining general benefits that can be attributed to both forms of corporate diversification.
First, firms might exploit operational synergies in several ways with the help of industrial and/or global diversifying acquisitions. These synergies can stem from combining operations of so far separate units (Trautwein, 1990). For example, a joint sales force might lower the costs for all outputs if the cost of joint production is less than the cost of producing each output separately (Panzar and Willig, 1977). The value creation due to economies of scope could also include services of physical assets as common input (Teece, 1982).
Second, there are also different financial rationales for corporate diversification. Firms might create internal capital markets because it enables them to allocate and reallocate capital more efficient compared to external capital markets. Furthermore, capital raised internally is less costly than funds raised
6 There are three broad types of synergies: operational, strategic and financial (Markides and Ittner, 1994).
7 in the external capital market (Martin and Sayrak, 2003). Also, Hubbard and Palia (1999) find that diversifying acquisitions lead to positive abnormal returns for acquiring firms by creating internal capital markets in the absence of well-developed external capital markets. They show that the highest acquirer gains were earned when financially unconstrained bidders acquired financially constrained target firms.
Consistently, Francis et al. (2008) document positive abnormal bidder returns for diversifying acquisitions that enable targets to overcome their financial constraints. They find that an important motivation for acquisitions is to provide access to cheaper financing, thereby relaxing target firms’
financial constraints (Francis et al., 2008). Thus, internal capital markets created by diversifying acquisitions may benefit firms by overcoming information deficiencies of external capital markets and easing the effect of external financial constraints on investments (Ekkayokkaya and Paudyal, 2010).
Another financial rationale for diversification is that diversifying acquisitions could lead to risk reduction. In finance, corporate diversification is a general technique for reducing investment risk.
Diversification leads to risk reduction if the income streams of the merging firms are not perfectly positively correlated. In that case, cash flow volatility is reduced, and as a result, diversified firms will often have less risk than each of the pre-merger firms (Lewellen, 1971). The risk reduction resulting from diversification increases the collateral value of the combined debt and allows firms to increase their borrowing capacity at a lower cost of debt (Ekkayokkaya and Paudyal, 2010). This is also called the debt coinsurance effect.
Lastly, firms might use anticompetitive motives for corporate diversification. For example, firms can collude with competitors it meets in multiple markets (Martin and Sayrak, 2003). The goal of simultaneously limiting competition is called the mutual forbearance hypothesis. Further, firms can aim at deterring potential entrants from its markets by making diversifying acquisitions (Trautwein, 1990).
2.1.1 Benefits of Industrial Diversification
There are also benefits that can be attributed more specific to industrial diversification. For
instance, industrial diversifying acquisitions could benefit acquirer shareholders by realizing synergies
with respect to management capabilities. Examining acquisition announcements during the conglomerate
merger wave of the late 1960s, Matsusaka (1993) find that conglomerate mergers – mergers between
firms that are involved in different business – benefit acquirer shareholders by exploiting managerial
synergy. His results show that the market responded positively to acquirers who retained the management
of their unrelated target. Firms might realize these managerial synergies because acquirers’ managers may
possess superior management skills that benefit the target firm and ultimately acquirer shareholders
(Trautwein, 1990). The idea is that firms which are said to have superior management capabilities can be
8 profitable in multiple business segments or industries and that diversification is a search process by which firms seek business segments that are good matches for their management skills (Matsusaka, 2001).
Further, the creation of internal capital markets through corporate diversification might benefit firms in the form of cross-subsidization. According to the efficient internal capital market theory, industrially diversified firms can create value by actively reallocating capital across divisions. The cash flows in these diversified firms are not automatically returned to their sources, but instead are exposed to an internal competition (Williamson, 1975). Evidence of this investment interdependence between divisions can be found in Lamont (1997) and Shin and Stulz (1996). Moreover, Stein (1997) analyzes internal capital markets and document that firms could benefit by allowing their headquarters to allocate funds to the business segments with the best investment opportunities. For example, firms could transfer funds to divisions where the returns are highest. Consistent with this rationale, Billet and Mauer (2003) show that subsidies to small financially constrained divisions with good relative investment opportunities significantly increase firm value, while transfers of funds from divisions with good relative investment opportunities significantly decreases firm value. By overcoming these financial constraints, firms are able to undertake positive NPV projects which would otherwise have been forgone (Francis et al., 2008).
2.1.2 Benefits of Global Diversification
The operational synergies that stem specially from global diversification are mostly rationalized
in the context of the theory of transaction-cost economics (Markides and Ittner, 1994). For example, firms
might be motivated to expand international if this provides them access to cheaper labor or other
production inputs, especially from less developed (low-cost) countries (Hood and Young, 1979; Morck
and Yeung, 1991). Firms could also benefit by transferring resources and knowledge abroad in order to
exploit firm-specific assets. The internalization theory posits that firms with specific intangible assets are
inclined to expand cross-border. According to Caves (1971), firms can increase their market value by
internalizing markets for certain intangible assets, such as superior production skills, patents, brand
names, marketing skills or consumer goodwill. Also, Morck and Yeung (1991) find that the effect of a
global acquisition on firm value is driven by the intensity of the firm’s R&D or advertise spending. These
results suggest that cross-border acquisitions enhance the scope for using the firm’s intangible assets,
thereby creating competitive advantages over other firms. The reasoning is that firms that possess unique
intangible assets can create more value through internalizing the market for their specific assets in foreign
subsidiaries at low cost, than by attempting to capture market value for these assets in other ways
(Markides and Ittner, 1994; Malone and Rose, 2006). To benefit acquirer shareholders, these advantages
have to outweigh the costs of transferring these assets cross-border (Trautwein, 1990).
9 Global diversification could also be annotated by a financial perspective. Global diversification theories hypothesize that firms can benefit by meeting the international diversification portfolio objectives of investors (Freund et al., 2007). As long as economic activity in different countries is less than perfectly correlated, cross-border acquisitions should improve individual investors’ risk-return opportunities (Markides and Ittner, 1994). The rationale is that individual investors are not able to diversify global on their own account under the same terms as firms can, due to institutional constraints on international capital flows and information asymmetries. Globally diversified firms benefit their shareholders by offering them international diversification opportunities via direct investments abroad through cross- border acquisitions (Morck and Yeung, 1991). Therefore, global diversifying acquisitions should generally be leading to positive acquirer returns for multinational firms if individual investors recognize this service
7.
Cross-border acquisitions may also benefit firms in the form of a cross-listing option. Because this option is not available for a high number of firms, global diversifying acquisitions could provide this potentially valuable option for firms that do not meet cross-listing requirements. For example, some stock market exchanges (e.g. New York Stock Exchange) require firms that cross-list to have a certain income.
Also, the costs of cross-listing could be quite significant for some firms (Francis et al., 2008).
Moreover, it has been argued that global diversifying acquisitions create more tax arbitrage opportunities for firms. Multinational firms benefit shareholders because they have more possibilities for tax avoidance using transfer pricing and finding tax havens (Morck and Yeung, 1991). For example, globally diversified firms could concentrate their debt financing in highly taxed countries. Furthermore, multinational firms can shit taxable income within its group of companies to reduce its overall tax burden (Harris et al., 1991). Empirical evidence points out that those tax considerations have an important influence on internal pricing policies of globally diversified firms. Harris et al. (1991) find that firms with subsidiaries in low-tax countries have relatively low tax payments per dollar of assets or sales, while Grubert and Mutti (1991) show that firms declare more income in low-tax jurisdictions. These findings are consistent with income shifting because of tax purposes.
2.1.3 Benefits of Corporate Diversification during the 2008-2010 Financial Crisis
The recent financial crisis has possible effects on the relationship between corporate diversification and firm value. It has been argued that the value of diversification might change due to changes in the efficiency of internal capital markets. For example, Hovakimian (2011) and Yan et al.
7 Several studies have been supporting this prognosis (e.g. Severn, 1974; Errunza and Senbet, 1981; Kim and Lyn, 1986).
10 (2010) show that diversified firms improve the efficiency of internal resource allocation in times of tightened external capital markets (Rudolph and Schwetzler, 2011). Moreover, Kuppuswamy and Villalonga (2010) examine two channels through which the 2008-2010 financial crisis might have triggered a change in the value of diversification for U.S. firms and find that the value of diversification increases due to financing and investment advantages. First, they show an increase in the value of the
‘more-money’ effect arising from the debt coinsurance feature of diversified firms. The rationale is that investors may prefer to lend their funds to safer diversified firms than to riskier non-diversified firms (Kuppuswamy and Villalonga, 2010). Second, they find that access to internal capital markets became more valuable and that the efficiency of internal capital allocation increased significantly during the recent crisis. Furthermore, Rudolph and Schwetzler (2011) extend the study by Kuppuswamy and Villalonga (2010) to a global scale and find an increase in the value of diversification for Asian and U.K.
firms, while the impact of the financial crisis upon the value of diversification for Continental European firms is insignificantly. Their results show that the change in the value of diversification varies across countries and that these differences can be explained by the degree of capital market development. In sum, above findings suggest that the benefits of corporate diversification are higher during the recent recession period for U.K. firms.
2.1.4 Law of Diminishing Marginal Returns
According to modern economists, the effect of diversity benefits can be explained by the law of diminishing marginal returns. This law is a fundamental principle of economics and states that in all productive processes, adding more of one factor of production will at some point yield lower per-unit returns (Samuelson and Nordhaus, 2001). The law of diminishing returns is rooted in the work of Turgot (1767) who argued that doubling the expenditure on agriculture do not double the product. Applied to corporate diversity, this means that the aggregate marginal benefit of diversity should increase at a decreasing rate with the level of diversity (Ekkayokkaya and Paudyal, 2010). The reasoning is that, in general, it will be more difficult for highly diversified firms to exploit further diversity benefits than for moderately diversified firms. Therefore, acquirer stock returns from diversifying acquisitions should diminish marginally with the firms’ existing (pre-bid) level of diversity.
2.2 Constraints and Costs of Corporate Diversification
Diversified firms are often more complex than focused firms and thereby more difficult to manage. There are several factors that might be expected to exert a negative impact of diversity on acquirers’ gains, or at least to impose limits on the potential benefits for diversified firms (Grant, 1987).
Therefore, the potential costs associated with managing a diversified firm also define the benefits of
11 maintaining a non-diversified firm (Martin and Sayrak, 2003). The costs of diversity are not exactly differentiated between industrial and global diversification because, in general, these costs can be attributed to both types of diversification.
The most popular explanation in finance literature for the costs of diversity is that firms face agency problems. The agency theory emphasizes that managers may extract private benefits at the expense of shareholders, for example, by making diversifying acquisitions. In this view, corporate diversification exacerbates managerial agency problems (Martin and Sayrak, 2003). The agency view is therefore closely related to management behavior and is inconsistent with either profit maximization or resource efficiency (Montgomery, 1994). Extensive literature attributes the costs of diversification to several agency problems. For instance, managers might engage in diversifying acquisitions in order to increases the firm’s demands for his or her particular skills and knowledge. Hereby, managers can reduce the probability of being replaced, extract higher wages and obtain more latitude in determining corporate strategy (Shleifer and Vishny, 1989). This behavior is called managerial entrenchment. It suggests that managers have the power to make manager-specific investments, such as diversifying acquisitions, which are not in the interests of their shareholders. Furthermore, Bodnar et al. (1998) argue that managers may deliberately engage in global diversifying acquisitions to make shareholder monitoring of their actions more difficult. The reasoning is that multinational firms are arguable more complex than domestic firms because of language, accounting, cultural and political differences between markets (Morck and Yeung, 1991; Freund et al., 2007). These monitoring difficulties provide managers more opportunities to extract private benefits at the expense of shareholders.
Managers might also diversify in order to get more assets under control because of the power and
prestige associated with directing larger firms (Jensen, 1986; Stulz, 1990). Furthermore, managerial
compensation is often related to firm size (Jensen and Murphy, 1990). It has been argued that the
incentives of managers with low ownership may not be aligned with the best interest of their shareholders
and that these managers engage in diversifying acquisitions because they do not bear the costs of
diversification (Jensen and Meckling, 1976). Corporate diversification in the view of the agency theory
could also be explained by reducing managers’ employment risk and/or by the need for personal
diversification. Managers may pursue diversifying acquisitions as their employment risk decreases with
firm-specific risk and as they can reduce the risk of their undiversified personal portfolios (Amihud and
Lev, 1981). These acquisitions improve only managers’ personal positions and do not benefit acquirer
shareholders.
12 Alternatively, it might be the case that diversified firms have more difficulties in allocating their resources than non-diversified firms, implying inefficiency problems rather than agency problems (Martin and Sayrak, 2003). A general consensus is that inefficient investments will lead to less valuable firms. In support of this view, Rajan et al. (2000) show that when diversity in investment opportunities increases, resources can flow toward the most inefficient divisions of a firm. Moreover, Scharfstein and Stein (2000) argue that rent-seeking behavior by division managers in diversified firms can destroy the workings of an internal capital market, leading to inefficient capital allocation. Further, Jensen (1986) and Stulz (1990) find that firms with excess cash may overinvest in business divisions with poor investment opportunities.
These results suggest that capital allocation in diversified firms tends to be inefficient. Altogether, the common implication is that excessive diversity is likely to cause politics among divisions and put severe strain on firms’ monitoring capability, which will ultimately cost acquirer shareholders (Ekkayokkaya and Paudyal, 2010).
Importantly, Ekkayokkaya and Paudyal (2010) point out that diversification gains also depend on the acquirer’s capability in integrating and managing newly acquired firms. They argue that the diversifier’s organizational capability can affect its gain from diversifying acquisitions independently of the expected benefits. Thus, firms with superior managerial skills will gain the most from diversification and the differences in acquirer gains will become greater as the level of diversity increases. The argument is that high levels of diversity cause severe constraints on firms’ ability in coordination and resource allocation, and making managing more difficult and costly (Ekkayokkaya and Paudyal, 2010). An explanation for this phenomenon is bounded rationality. Bounded rationality asserts that managers are intended to be rational, but they sometimes fail because of human cognitive and emotional architecture (Jones, 1999). Consistent with this view, Grant (1987) argues that there are limits to the capacity of managers to cope successfully with greater complexity. Furthermore, Siddharthan and Lall (1982) find that managerial constraints increase with international diversification. In line, Geringer et al. (1989) report that as firms increasingly diversify globally, the costs associated with global dispersion began escalating.
Indeed, firms experience increasing transaction costs with greater levels of global diversity (Hitt et al., 1997).Moreover, Teece (1982) indicates that if diversification is based on scope economies, the problems associated with accessing common inputs will be greater as the degree of diversity increases. This is especially the case when the common input is knowledge.
2.3 Trade-Off Theory and Hypothesis
A solution to the diversification puzzle might be given by the trade-off theory of corporate
diversification. As noted before, existing literature has ignored the possibility of viewing an optimum
13 level of existing (pre-bid) diversity, where the benefits and costs of diversification are in balance. While there are many explanations of why firms make diversifying acquisitions, most theories fail to specify whether some optimal level of diversity may exist. Therefore, a theoretical framework is needed for the relationship between diversifying acquisitions and acquirer gains that includes both benefits and costs of corporate diversification over different phases of diversification (Lu and Beamish, 2004). An analysis must be performed that show how the benefits and costs jointly behave when a firm with a given (pre-bid) level of diversity makes a diversifying acquisition, because viewing them together might reveal a trade- off between the benefits and costs of corporate diversification (Ekkayokkaya and Paudyal, 2010). The argument is that there are limits to firm’s organizational capabilities to cope with increased diversity. At some point, the coordination required costs more than the benefits derived from diversification and these costs begin to produce diminishing gains to diversifying acquisitions (Hitt et al., 1997). So, the desire to diversify through diversification stems from a trade-off between the potential benefits of diversification and the potential losses in shareholder value (Denis et al., 1999).
There are a few studies which examined the relationship between the degree of diversity and firm performance. For example, Geringer et al. (1989) and Hitt et al. (1997) show a nonlinear inverted U- shaped relation between the level of global diversity and firm performance for midsize and large multinationals. Their findings show that as the degree of global diversification reached higher values firm performance also exhibited increased values, but then peaked and exhibited diminished levels of firm performance. Hitt et al. (1997) argue that the costs of global diversification eventually exceed the benefits of diversification due to increased managerial information-processing demands and transaction costs.
More interesting, they show that industrial diversification moderates the relationship between international diversification and firm performance, and that the slope and shape of this nonlinear relationship varies with the level of industrial diversification. This might imply that experience with industrial diversification can build managerial capabilities that allow more effective management of global diversification (Hitt et al., 1997). On the other hand, Capar and Kotabe (2003) find a U-shaped curvilinear relationship between global diversification and performance in service firms. They claim that global diversification would reduce firm performance up to a certain point owing to diseconomies of scale, but begins to increase at higher levels of global diversity due to benefits of economies of scope.
Furthermore, Lu and Beamish (2004) show a horizontal S-shaped relationship between multinationality
and performance for Japanese firm, which seem to be at odds with the inverted and upright U-curves from
prior evidence.
14 There are also few studies analyzing the relationship between industrial diversity and firm performance. For example, Palich et al. (2000) examine the results from 55 previously published studies and find that moderate levels of industrial diversification yield higher levels of performance than either limited or extensive diversification, consistent with the curvilinear model. In addition, Kahloul and Hallara (2010) also find evidence for a non-linear relationship between industrial diversification and performance. Contrary, Miller (2006) shows a positive relationship between diversification based on technological diversity and market-based measures of firm performance
To my best knowledge, Ekkayokkaya and Paudyal (2010) are the first who applied the trade-off theory to diversifying acquisitions to investigate the relationship between acquirers’ existing (pre-bid) level of industrial diversity and acquirer gains from industrial diversifying acquisitions. They find that announcement returns from domestic industrial diversifying acquisitions follows a curvilinear inverted U- shape. Their findings show that acquirer gains from industrial diversifying acquisitions are significant for single-segment acquirers and increases for moderately diversified acquirers (operating in two to five segments). Eventually, when a firm becomes highly diversified (six or more segments), the gain significantly drops and eventually becomes a loss. This suggests that industrial diversification benefit acquirers during the early stages of diversification, but ultimately lead to severe organizational constraints and higher costs if they diversify further.
However, it has been argued that the 2008-2010 financial crisis has possible effects on the relationship between diversification and firm value for U.K. firms (Rudolph and Schwetzler, 2011). As noted before, prior evidence shows that, during the recent crisis, the value of diversification has increased because of higher financing and investment advantages. These findings imply that the benefits of diversification are higher during recession periods, while the costs of diversification are largely the same.
This supports the prediction of a positive linear relationship between the existing (pre-bid) level of diversification and acquirer gains from diversifying acquisitions during the 2008-2010 financial crisis.
This position rests upon several assumptions derived mainly from internal market efficiency arguments (Palich et al., 2000). This prediction implies that firms with high levels of pre-bid industrial diversity also could gain from diversifying acquisitions because they have much greater flexibility in capital formation due to their access to internally generated resources in times of tightened external capital markets.
Therefore, with respect to industrial diversity, I hypothesize:
Hypothesis 1: “During the 2008-2010 financial crisis, the relationship between the acquirer’s existing
(pre-bid) level of industrial diversity and acquirer gains from diversifying acquisitions is positive”.
15 This leads to the question whether the trade-off hypothesis for the pre-bid level of diversity is also relevant in the context of global diversification, since the acquisition of a foreign target can be considered global diversification at the firm level. Therefore, I extend the trade-off hypothesis to a global scale and hypothesize:
Hypothesis 2: “During the 2008-2010 financial crisis, the relationship between the acquirer’s existing (pre-bid) level of global diversity and acquirer gains from diversifying acquisitions is positive”.
3. Empirical Evidence on Acquirer Gains from Diversifying Acquisitions
This section presents a short overview of empirical evidence on acquirer gains. Subsection 3.1 report evidence on acquirer returns for industrial diversifying acquisitions, whereas in subsection 3.2 acquirer gains from global diversifying acquisitions are discussed. As indicated earlier, existing evidence regarding the effect of diversification on acquirer gains is somewhat contradictory and there are different explanations for the observed differences in acquirer gains
8. Altogether, the empirical evidence on the impact of corporate diversification on acquirer stock returns remains an unresolved puzzle.
3.1 Empirical Evidence on Acquirer Gains: Industrial Diversification
Most studies to date have examined the effects of industrial diversification on acquirer announcement returns. Appendix A provides an extensive literature review on acquirer stock returns for industrial diversifying acquisitions. As can be seen, both positive and negative acquirer gains have been found and different definitions of industrial diversification have been used. Doukas et al. (2002) and Chevalier (2004), among others, find that industrial diversifying acquisitions lead to negative market reactions during 1980-1995. Interestingly, Doukas et al. (2002) find that acquirer gains are higher for firms that are associated with conglomerate groups. In addition, Morck et al. (1990) analyze industrial diversifying acquisitions in the U.S., and document more negative acquirer returns for diversifying acquisitions compared to related acquisitions. However, they find positive acquirer returns for the sub- period 1975-1979. Also, Akbulut and Matsusaka (2010) examined 57 years of diversification literature and report significant negative announcement returns for unrelated acquisitions. Remarkably, they find
8 For example, Hitt et al. (1997) argue that the relation between diversification and firm value is more complex than has been theoretically argued and empirically tested so far. Akbulut and Matsusaka (2010) report that conflicting findings could be the result of examination of different time periods because corporate diversification is an evolving practice which might be sensitive to time variation. Also, Francis et al. (2008) point out that increasingly large reputable multinational firms engage in global diversifying acquisitions that have led to changes in acquirer characteristics over time. Lastly, different definitions of industrial diversification may lead to different acquirer gains.
16 that industrial diversifying acquisitions were less harmful than related acquisitions over the last 57 years.
Besides, their results show that acquirer gains turn positive when paying with cash (instead of common stock) for industrial diversifying acquisitions.
On the contrary, several papers show that industrial diversification lead to positive acquirer gains.
For example, Matsusaka (1993) find significant positive announcement returns for unrelated acquisitions during the conglomerate merger wave in the 1970s, while Hyland and Diltz (2002) report significant positive acquirer gains for the full sample in the 1980s.
3.2 Empirical Evidence on Acquirer Gains: Global Diversification
More recently, researchers have been analyzing the effects of global diversifying acquisitions.
Morck and Yeung (1992), Kiymaz and Mukherjee (2000) and Freund et al. (2007) all find that cross- border acquisitions are value enhancing for acquiring firms (see Appendix B). However, other scientific papers show contradicting findings. For example, Aybar and Ficici (2009), Eun et al. (1996) and Kuipers et al. (2009) all show negative acquirer returns for global diversifying acquisitions. Moreover, Eun et al.
(1996) report significant differences between acquirer countries: Gains are positive for acquirers from Japan, but negative for U.K. acquirers.
There is also a stream of literature reporting lower acquirer gains for cross-border acquisitions when compared to domestic acquisitions. For instance, Moeller and Schlingemann (2005) examined 4,430 U.S. acquisitions between 1985 and 1995 and find that acquirers experience significantly lower stock returns for cross-border than for domestic acquisitions. In a similar study, Conn et al. (2005) find cross-border discounts for U.K. acquirers.
4. Data and Methodology 4.1 Data Selection
The sample data is collected from the Zephyr database of Bureau Van Dijk and includes all acquisitions made by U.K. firms that are announced on and after January 1, 2008 and up to and including December 31, 2010. This time period corresponds to the period of slow economic growth due to the financial crisis in the U.S. and Europe (Ings and Inoue, 2012).
The sample is drawn from domestic and cross-border acquisitions made by U.K. public firms
which have their stock price data held on the Thomson Reuters DataStream database. The target firms in
the sample can be public or private. To ensure that the acquirer does not already own a controlling interest
in the target, I define acquisitions as deals where the acquirer owns less than 10% of the voting shares of
the target firm prior to the acquisition and ends up by owning more than 50% of the target’s voting shares.
17 To be included in the sample, I further require that the acquisition has been completed and that information is available about: acquirers’ market capitalization, deal value, deal method of payment, acquirer ISIN number, acquirer- and target U.S. SIC-code(s) and target country origin. In addition, I follow Moeller and Schlingemann (2005) and exclude acquisitions with a deal value less than €1 million in order to avoid the potential marginal impact of small targets on estimates.
Among the remaining sample are several acquisitions consisting of multiple acquirers and/or targets. I exclude these acquisitions because the gain from a diversifying acquisition per firm cannot be distinguished. The resulting sample consists of 220 completed acquisitions that were announced during the 2008-2010 financial crisis.
4.2 Methodology
4.2.1 Abnormal Returns
An event study methodology is used to test market reactions in response to acquisition announcements because firms commonly diversify through acquisitions (Graham et al., 2002). I follow Brown and Warner’s (1985) event study methodology to analyze daily stock returns around acquisition announcements. The trade-off hypothesis of corporate diversification is examined by measuring abnormal returns around the acquisition announcement date. This methodology has some strong scientific characteristics. First, it helps overcome potential endogeneity problems associated with the imputed value method. Second, announcements returns from (diversifying) acquisitions provide fairly clean estimates of the changes in expected values of diversified firms and the effect of diversification is isolated from other potential influences (Akbulut and Matsusaka, 2010). Third, the estimate is forward looking. Lastly, according to Healy et al. (1992), announcement stock returns seem to predict subsequent operational performance.
To measure the acquisition’s impact, I follow Fuller et al. (2002) and Moeller and Schlingemann
(2005) and use the market-adjusted model (MAM) to calculate abnormal returns. The MAM is preferred
over other models (e.g. market model) because many diversifying acquirers are multiple acquirers and
therefore it is possible that a firm’s systematic risk (beta) is influenced by other acquisition
announcements in the estimation period. This makes the estimated beta less meaningful for analysis
(Conn et al., 2005; Fuller et al., 2002). Also, Brown and Warner (1980) show that for short event
windows, adjusting for firm’s systematic risk does not significantly improve estimation. Lastly, Draper
and Paudyal (2006) report that the estimates of abnormal returns to U.K. acquirers are not sensitive to the
18 choice of return benchmark model. Using the MAM for calculating abnormal returns, actual returns are benchmarked against their contemporaneous market return
9. The MAM is given by equation (1):
(1)
Where is the actual return to firm on day and is the return of the value-weighted market index on day . then is the market-adjusted abnormal return for acquirer on day . Subsequently, abnormal returns are aggregated through time by calculating the cumulative abnormal return (CAR) over the three-day (-1,+1) and the five-day (-2,+2) event windows around the announcement date. CARs are estimated using equation (2):
(2)
The event window (t1, t2) is the period around the announcement date. The announcement date is denoted as day 0 (event date) and represents the day at which the acquisition is announced to the market, or the next trading day if the announcement date is not a trading day. As indicated, two different short- term event windows (-1,+1) and (-2,+2) are used to capture the announcement effect of an acquisition on stock prices. Although a variety of event windows is used in existing work, the vast majority of diversification studies use a short-term event window for analyzing abnormal returns (e.g. Moeller and Schlingemann, 2005; Francis et al., 2008). Moreover, Conn et al. (2005) argue that the advantage of short-term event windows is that its results are typically insensitive to the benchmark return used for adjusting actual returns. The aggregation of abnormal returns through time and across firms (cumulative average abnormal return) is calculated by using equation (3):
(3)
To test if mean CARs differ significantly from zero, I first perform a parametric t-test. In addition, I use the nonparametric Wilcoxon signed rank test to test if median CARs differ significantly
9 Stock return data is retrieved from the Thomson Reuters DataStream database and is based on the Total Return Index which includes dividends and capital gains. The Local Market Index for this sample is the U.K. value- weighted FTSE All Share Index. Moreover, both stock and market returns are measured as continuously compounded return.
19 from zero because the Jarque Bera test for normality indicates that the CARs are not normally distributed.
Furthermore, I conduct a parametric t-test that assumes unequal variances in different subsamples, to test if the mean CARs of two subsamples differ statistically significant from each other. The Wilcoxon rank- sum test is performed to test for statistical differences in subsample medians.
4.2.2 Definition of Industrial Diversification
The standard approach is to classify acquisitions as industrial diversifying if acquirer and target operate in different 2-digit, 3-digit, or 4-digit U.S. SIC codes (Scharfstein, 1998). The U.S. Standard Industrial Classification (SIC) system is a popular method for classifying business segments in empirical work in mergers and acquisitions. The simplest SIC-based measures label an acquisition as industrial diversifying if the acquirer and target do not share the same primary 2-digit SIC code (Lien and Klein, 2006). While U.S. SIC codes are widely used, there is little consensus among researchers about which digit level should be used for measuring industrial diversifying acquisitions. Although it has been argument that the 3- and 4-digit SIC levels are potentially too detailed to identify the industrial structure of a firm (Ekkayokkaya and Paudyal, 2010), more limitations are reported about the use of 2-digit SIC codes. For example, Scharfstein (1998) argues that a shortcoming of using 2-digit SIC codes is that there can be vertical connections between business segments in different 2-digit SIC codes. Moreover, it has been reported that classification based on 2-digit SIC codes could lead to potential aggregation biases because it fails to identify horizontal and vertical acquisitions (Capron, 1999). Since theory and prior evidence do not point to any particular definition, I follow Flanagan and O-Shaughnessy (2003) and define an acquisition as industrial diversifying if the acquirer and target do not share the same primary 4- digit U.S. SIC code.
4.2.3 Definition of Global Diversification
Consistent with Moeller and Schlingemann (2005) and Francis et al. (2008), I classify the sample deals into domestic and cross-border acquisitions in order to identify global diversifying acquisitions. An acquisition is defined as global diversifying if the acquisition is cross-border and as non-global diversifying if the acquisition is domestic.
4.2.4 Measuring Pre-Bid Level of Industrial Diversity
To analyze the trade-off theory of diversification, I categorize acquirer firms in two groups with respect to their existing (pre-bid) level of diversity
10. The same 4-digit U.S. SIC codes are used as for
10 Firms are split into two groups based on the median value of the sample. This offers an alternative approach over the single-segment vs. multi-segment classification adopted by existing studies (e.g. Akbulut and Matsusaka, 2010).
20 measuring industrial diversification. Barely industrially diversified acquirers are defined as acquirers that operate in one to two segments based on the 4-digit SIC code. Highly industrially diversified acquirers are defined as those operating in more than two business segments.
4.2.5 Measuring Pre-Bid Level of Global Diversity
With respect to global diversification, I propose a different approach which makes it possible to group firms based on their existing (pre-bid) level of global diversity. The problem with prior studies is that they rely on databases which provide them little information about global segment data. For example, Compustat (database) limits the number of possible global segments to four and therefore the number of global segments in which a firm operates has limited meaning (Denis et al., 2002). Furthermore, most databases do not specify the individual countries in which firms operate, thereby making it unable to use the number of countries as a measure of global diversification. To overcome these problems, I use the Zephyr database of Bureau Van Dijk and collect information about the number of countries in which the acquirer has recorded subsidiaries. The number of foreign countries distinguishes the degree of global diversity of a firm (Stopford and Wells, 1972). I then group acquirer firms into two groups according to the number of countries in which they have recorded subsidiaries, based on the median value of the sample. Barely globally diversified acquirers are defined as acquirers that have recorded subsidiaries in one to eight countries. Highly globally diversified acquirers are defined as those that have recorded subsidiaries in nine or more countries.
4.2.6 Multivariate Analysis
To analyze cross-sectional stock-price variations, I also perform a multivariate analysis. By performing several multivariate regression models, additional insight is obtained regarding the determinants of acquirer announcement returns. The Ordinary Least Squares Method (OLS) is used to see how CARs are affected by several firm-, deal- and control variables. Moreover, the multivariate analysis accounts for possible interaction effects between various factors that shape acquirer abnormal returns (Ekkayokkaya and Paudyal, 2010). The general OLS model that will be used in this paper is as follows:
(4)
Where the dependent variable represents the three day (-1,+1) cumulative abnormal return
for acquirer , is the constant term, the vector of explanatory variables ‘X’ includes independent
21 variables as well as control variables and is the random error term. The independent variables are of main interest and contain different measures for (the existing level of) diversification. The control variables are factors that have empirically been proven to influence acquisition performance, such as method of payment and target’s listing status. The variables used for analysis will be described next.
Independent Variables: Corporate Diversification
A. Industrial Diversification
From the Zephyr database of Bureau van Dijk (primary) 4-digit U.S. SIC codes are extracted in order to create the independent variables of interest. An acquisition is defined as industrial diversifying if the acquirer and target do not share the same primary 4-digit U.S. SIC code. The dummy variable INDUS equal to 1 (0) if the acquisition is industrial diversifying (non-industrial diversifying) on the 4-digit level, respectively. With respect to the existing (pre-bid) level of industrial diversity, I follow Ekkayokkaya and Paudyal (2010) and include a dummy variable for the acquirer’s pre-bid level of industrial diversity.
ACQ_SEG3 equal to 1 (0) if the acquirer has three or more U.S. SIC codes on the 4-digit level, (otherwise). These acquirers are labeled as highly industrially diversified based on the median value of the sample.
B. Global Diversification
Global diversifying acquisitions are defined as cross-border acquisitions reported by the Zephyr database. The dummy variable GLOBAL equal to 1 (0) if the acquisition is global diversifying (non- global diversifying), respectively. Furthermore, a dummy variable is included in the model to represent the existing (pre-bid) level of global diversity. To measure a firm’s level of global diversity, information about the countries in which an acquirer has recorded subsidiaries is collected from the Zephyr database of Bureau van Dijk. The dummy variable ACQ_COUNTRY9 equal to 1 (0) if the acquirer has nine or more countries in which it has recorded subsidiaries, (otherwise). These acquirers are labeled as highly globally diversified based on the median value of the sample.
C. Industrial and Global Diversification
Lastly, a dummy variable is included to account for acquisitions which are both industrial and
global diversifying: the variable INDUS_GLOBAL equal 1 (0) if the acquisition is both industrial and
global diversifying on the 4-digit level, (otherwise). The inclusion of this dummy variable enables me to
avoid potential multicollinearity in the multivariate regression models because the dummy variables for
diversifying acquisitions are mutually exclusive (Freund et al., 2007). When the dummy variable INDUS
22 or GLOBAL take the value of 1, it indicates the acquisition is only diversifying in the form indicated by the variable. When an acquisition is both industrial and global diversifying, INDUS_GLOBAL takes the value of 1, but both INDUS and GLOBAL take the value of 0.
Control Variables
A. Target’s Listing Status [PRIVATE]
A well observed variable in acquisition literature is target’s listing status, also called the ‘listing effect’. Several prior studies find that acquirer stock returns are related to the target’s listing status.
Generally, these studies have examined U.S. acquisitions and indicate that the acquisition of an unlisted (private) target results in higher acquirer gains than the acquisition of a listed target. For example, Fuller et al. (2002) show that private targets (stand-alone firms or subsidiaries) provide significantly larger acquirer gains compared to listed targets. In line, Moeller et al. (2004) find that, on average, acquirers of listed targets earn a significant negative abnormal return, while acquirers of unlisted targets earn a significant positive abnormal return. Further, Faccio et al. (2003) analyze announcement period returns to acquirers in 17 Western European countries for the time period 1996-2001. Their results show that acquirers of listed targets earn an insignificant average bidder return of -0.38%, while acquirers of unlisted targets earn a significant average acquirer gain of 1.48%. Information about the target’s listing status is collected from the Zephyr database. Because only listed and unlisted targets are included in the data sample, I create a dummy variable PRIVATE that takes the value of 1 (0) if the target is unlisted (listed), respectively. Consistent with empirical evidence, I expect a positive relationship between PRIVATE and acquirer stock returns.
B. Method of Payment [CASH and STOCK]
Another area of interest is the choice for method of payment. Prior empirical evidence has shown
that the method of payment is related to announcement period returns for acquirers. It is commonly
known that acquirers experience significantly negative stock returns when they pay with stock for their
acquisition (Masulis et al., 2007). A great amount of this research focused on the theoretical relationship
between information asymmetry and the method of payment in acquisitions, also called the ‘signaling
hypothesis’ (Fuller et al., 2002). Consistent with the signaling hypothesis, Travlos (1987) find significant
differences in the abnormal returns between acquisitions financed with common stock and cash. The
reasoning is that acquisitions paid with common stock might convey negative information that the
acquirer is overvalued. Therefore, acquirers who use cash are expected to generate higher stock returns
than acquirers who use common stock as method of payment for their acquisition.
23 Moreover, Fuller et al. (2002) show that the method of payment chosen in an acquisition is partially a function of target characteristics. They find insignificant acquirer gains for public targets paid with cash, while acquisitions paid with common stock results in significantly negative acquirer gains. On the contrary, Chang (1998) and Fuller et al. (2002) report that the impact of stock offers is positive for the acquisition of private (unlisted) target firms.
For analysis, the method of payment is grouped into two categories: 100% cash offers and 100%
stock offers. Information about the method of payment is extracted from the Zephyr database. To control for the method of payment effects I include the following two dummy variables: CASH, equal to 1 (0) if the deal is financed with 100% cash, (otherwise); and STOCK, equal to 1 (0) if the deal is financed with 100% stocks, (otherwise).
C. Relative Size [RELSIZE]
Acquirer abnormal return regressions generally control for relative size. The relative size variable proxies for several effects and might affect acquiring firms in different ways. For example, the larger the relative size variable, the greater the effect of the acquisition. This likely causes greater market reaction (Fuller et al., 2002). In addition, the integration of relative large targets is arguably complex and difficult, leading to higher costs and more organizational constraints. Moeller et al. (2004) point out that empirical evidence indicates that the sign of the relative size variable varies across studies. For instance, the relative size variable is positive in Fuller et al. (2002), Moeller et al. (2004) and Moeller and Schlingemann (2005) but negative in Travlos (1987) and Conn et al. (2005).
The variable RELSIZE is measured as the acquisition value (target’s market value) divided by acquirers’ market capitalization one year prior to the acquisition. Data on the acquirer’s market capitalization is retrieved from DataStream, whereas information on the acquisition value is extracted from the Zephyr database. Based on prior evidence regarding the sign of the coefficients for the relative size variable, I do not make any predictions for the sign of this variable
D. Acquirer Size [ACQSIZE]
To control for the effect of absolute acquirer size on CARs, I include the variable ACQSIZE, measured as the natural log of the acquirer’s market capitalization one year prior to the acquisition
11. Acquirer size is hypothesized to have contradicting predictions. On the one hand, large acquirers are associated with more agency problems due to, for example, monitoring problems (Bodnar et al., 1998) and the pursuit for power and prestige by managers (Jensen, 1986). Managers of large firms might engage
11 The natural log is used to normalize the acquirer size variable since there might be large variations in acquirer size.
