Choices: Evidence from European markets
Master Thesis Oleksiy Pyrozhkov University of Groningen: S2278340 Uppsala University: 9107140-P112 Supervisor: Dr. H. Gonenc Co-assessor: Dr. W. Westerman Date: 09.01.2014University of Groningen University of Uppsala
MSc International Financial Management MSc Business and Economics
1. Introduction
The academic world knows numerous studies on capital structure and on acquisition behavior separately. It has been generally established that firms have target capital structures, where all the costs and benefits of debt financing are balanced, even though there is a certain opposition to this concept. However, as Harford et al. (2009) have established, most of the companies are deviating from their target capital structure to bigger or smaller extent. The deviations happen due to a variety of reasons, both internal and external. These deviations can influence managerial decisions, for example the ones dedicated to mergers and acquisitions. This is a relatively unexplored direction of studies, mostly influenced by Uysal (2011) and a study replicating it to different extent (Vasiliou et al., 2008). Lack of research exploring this topic is a major reason for the relevance of the current paper as well.
It is important to study the cross-border component since cross-border acquisitions have never been a major focus of researchers in previous studies (which would usually contain a sample of domestic deals – Uysal (2011) or are not specific about contents of their sample – Harford et al. (2009)).
There is also mixed evidence for the direction of the influence of the cross-border acquisitions on the leverage deficit and acquisition behavior relationship. While some studies like Shimizu et al. (2004) provide evidence for the negative effect of cross-border acquisitions on the impact of leverage deficit on decision making (hence cross-border acquisitions are supposed to be less affected by the deviations from optimal capital structure), general financial literature, like Moeller and Schlingemann (2005) find that cross-border acquisitions are value-destroying, risky and more complex. Therefore, in the situation of deviation from optimal capital structure cross-border acquisitions should be less frequent, as they are subject to bigger scrutiny from both the managers making decisions and the financiers, creditors, who evaluate the opportunities the acquisitions provide and know of the possible destruction of value. In short it can be summarized as follows: cross-border acquisitions are known to be more complex, risky and potentially value-destroying; managers of overleveraged firms, who, as found by Uysal (2011) to pursue “only the most value-increasing acquisitions”, can judge cross-border acquisitions as not the most value-increasing ones.
Financial system of the country of the acquirer can be important for the relationship between deviation from optimal capital structure and acquisition decision-making by influencing the decision-making processes of the financiers. There is a branch of literature that uncovered that stronger ties with banks provide more available financing, lead to less information problems and agency problems (Weinstein and Yafeh, 1998; Boot, 1999; Degryse and Van Cayseele, 2000). Agency costs stemming from information asymmetry are found to increase the cost of debt financing and to decrease its availability. Therefore, in a situation in which those costs are lessened and potential gains from the acquisition are easier to communicate to one party (a bank), the probability of obtaining the financing and thus performing the acquisition should increase, even if the bidder is overleveraged. Hence, being in the country with a bank-based financial system may lead to companies getting better access to financing, even in the situation of deviation from optimal capital structure. This allows to expect a negative influence of company's originating from a bank-based financial system on the impact of leverage deficit.
accepted as a market-based financial system. In order to be able to draw conclusions based on his results, one has to check whether they could have been affected by the financial system of the country.
This paper checks and questions the results of Uysal (2011) on a different sample. By finding insignifi-cance of the influence of the leverage deficit on acquisition behavior in the European sample of mixed domestic and cross-border acquisitions and finding the positive and significant influence of being un-derleveraged on probability of acquisition and payment method I raise the concerns about the potential validity of Uysal’s (2011) dismissal of alternative explanation to his results. This paper also checks the possibility of introducing two potential mediators in the leverage deficit influence on acquisitions stud-ies: the financial system of the country of the acquirer and the difference between cross-border and do-mestic acquisitions. No influence of the financial system of the acquirer on the leverage deficit impact on acquisition behavior was found, while, contrary to prior predictions, cross-border acquisitions were found to be relatively less influenced by leverage deficit compared to domestic ones.
2. Theoretical background
2.1. The determinants of capital structure studies
Capital structure studies go back to the classic works of Modigliani and Miller (1958), Titman and Wessels (1988), Rajan and Zingales (1995) and others like Harris and Raviv (1991). Frank and Goyal (2003) give a good overview of the major theories that are employed when studies of target leverage are performed. Among them are The Pecking Order Theory (Myers, 1984), Market Timing (Myers, 1984; tested by Hovakimian, Opler and Titman, 2001) and a number of Trade-off Theories that cover the trade-offs in taxes vs bankruptcy costs, agency costs and stakeholder co-investment theory. I will try to briefly explain the theories and which implications they have for a first stage of our research – the determining of target capital structure of the companies of the sample. I will also overview the ma-jor empirical papers testing those theories too.
Pecking Order Hypothesis involves the notion that the firms or, more precisely, the firm insiders do not view three major sources of financing – retained earnings, debt and equity as similarly beneficial and preferable. Equity is seen as the riskiest and the least favorable one, followed by the debt and by the re-tained earnings. Therefore, in any particular financing choice managers will prefer the rere-tained earn-ings; when those are not available they use debt; when further borrowing is not an option the equity is issued. Hence, the actual leverage of the firm is the result of all past decisions and no optimal leverage level exists. However, this does not mean that Pecking Prder Theory does not have any implications for establishing the relationships between certain variables and leverage of the company. For example, ac-cording to it higher profitability leads to higher retained earnings and thus is negatively associated with the leverage ratio (Fama and French, 2002). This relationship between profitability and leverage has been also confirmed by certain studies using profitability as a control variable (Uysal, 2011). Capital expenditures are, on the other hand, a cash outflow for the company, increasing its financial needs and thus should increase the leverage ratio (Shyam-Sunder and Myers, 1999 through Frank and Goyal, 2003). Same reasoning can be applied to the dividends as they are also a cash outflow for the company and raise the need for additional financing.
equity; in times of the recessions with the government lowering the interest rates to stimulate economy the leverage of the companies increases (Frank and Goyal, 2003). However, incorporating such effects into explanatory model is complicated.
The family of trade-off theories give bigger implications for the purposes of the current study. One ex-ample is the tax savings vs bankruptcy cost duality. Higher leverage leads to higher tax savings, while increasing the potential bankruptcy costs as well (Barclay and Smith, 1999). However, as noted by Gra-ham (2000) (through Frank and Goyal, 2003), bankruptcy costs are deferred in time and uncertain com-pared to tangible and immediate tax savings from the leverage, which leads to a certain bias for the managers. The empirical evidence in favor of this this theory are limited, since the tax variables are generally of minor significance in existing research (Myers, 1984) and there is a contradiction to well-confirmed outcome of the pecking order theory: the negative relation between profitability and lever-age. According to tax shield theory higher profitability should be associated with higher leverage, since there is a need to protect the profits from taxation. Studies, like the ones by Titman and Wessels (1988) and Fama & French (2002), have proved the contrary.
con-sumers. A bankruptcy is a loss of (unique) investments of the stakeholders, thus any investments like that (R&D, Selling expenses, investments in human capital) are likely to be negatively associated with leverage – which has also been checked by Uysal (2011), who has been using the product uniqueness as control variables.
Misvaluation hypotheses are the ones that connect the notions of target leverage with the acquisition behavior and thus have to be explained in detail here. Uysal (2011) controls for misvaluation in his studies. While it is a marginal point in the study, it is important to give more attention to this theory since it can potentially provide alternative explanations to the results. Misvaluation has to do with the way that perceived value of the target company influences the probability of it being acquired and a choice of payment method, which in turn influences subsequently the target leverage. The complete and concise model has been suggested by Vishny and Scheifer (2003). Such behavior is put in scale by Rhodes-Kropf and Viswanathan (2004) who find the presence of merger waves, waves of acquisitions, waves of payments in cash or debt – which in turn influence capital structures of firms in question. Their study raises up the importance of controlling for merger waves while performing the study on de-viation from target capital structure effects on mergers and acquisitions.
As one can see, certain theories contradict each other, thus making the task of determining a model of optimal capital structure a difficult, sample-dependent task.
2.2. Deviations from target capital structure studies
Among the most important papers to be discussed is the work of Harford et al. (2009). The authors were among the first ones to investigate the actual existence of target leverage, provided the methodology that was later widely used in subsequent studies and found certain important results. The scholars determined the existence of target leverage ratios for the companies, they found that deviations from target leverage, being overleveraged significantly influence the financing of the acquisitions: overleveraged firms tend to finance acquisitions with equity instead of debt. They also found that deviations from target leverage are value-destroying by studying bid announcement returns. Finally, they found that deviations from target leverage are not static; when firms take on additional debt to finance acquisitions they tend to revert to the target leverage ratio over the following 5 years. Important implication of this study in the context of the current study is the fact, that Harford et al. (2009) found that cash acquisitions are associated with higher leverage, potentially pushing the firm far from its target leverage level – and require efforts from the management in order to return to the optimum. The paper is also important from the methodological perspective: it provides clear and established methodology of determining the deviation from capital structure and provides a set of explanatory variables for the target leverage determination. It is also clear that Harford et al. (2009) are more interested in confirming or denying the existence of the concept of leverage deficit per se, while their study acquisitions and acquisition behavior is just a method of reaching the original goal. It is important to notice that Harford et al. (2009) use book leverage as a proxy for capital structure of the company. While, as mentioned by Frank and Goyal (2003), it is a possible proxy, it is not the only one possible and the preference of it over the proxies based on the market value is questionable. Another important fact is the use of the US sample of large companies and their acquisitions.
Harford et al. (2009) do not mention whether they use cross-border acquisitions in the sample, or it is restricted to domestic deals only.
debt and tend to repurchase equity other than debt, effectively moving to higher target leverage. Leverage deficit was found to be more important in influencing the repurchase decisions, contrary to issuance decisions, which are more affected by stock price variables.
Another study which has to be mentioned and is even more important for the needs of current study is the research of Uysal (2011). This study is of special importance, since it serves as a seminal paper and is referred to multiple times. Uysal was researching the effect of deviations from target capital structure on the frequency of the mergers and acquisitions, on the payment method (by estimating how likely is underleveraged/overleveraged firm is to pay all cash in the deal), on the relative size of the deal, on returns from the acquisition and on managerial adjustments to capital structure. Uysal draws from target capital structure literature, including Harford et al. (2009), from capital structure literature (including nearly everything I mentioned above) and from acquisition literature and uses a relatively large US sample of domestic acquisitions (as he specifically mentions). The findings are coherent with the ones of previous studies: leverage deficit exists, it negatively influences the frequency of mergers and acquisitions, their relative size and the probability of bidder paying in cash. In the situation of leverage deficit managers are inclined to pick the target more carefully which is reflected in significantly higher cumulative abnormal returns after the acquisition for overleveraged acquirers. Managers also are found to re-balance their leverage to bring it closer to the target after the acquisition takes place. It is important to note that Uysal (2011) estimated both the impact of leverage deficit per se. - the difference between the actual leverage and the target leverage of the company for a specific year, and the impact of being underleveraged or overleveraged (top or bottom quantile of leverage deficit for the sample for a particular year) and found that while the effect of being overleveraged is significant and strong, the influence of being underleveraged on acquisition behavior is generally not significant. This has important implication for the sign of the leverage deficit which stems from its definition: the higher the deficit is, the more overleveraged firm becomes, hence a negative relationship. However, if in certain sample and in certain conditions the influence of being undeleveraged would of similar magnitude and directed oppositely, the direction of the impact of the leverage deficit would be not so predictable.
have the same leverage level, however differences in their target leverage impact the respective leverage deficits, which will in turn lead to different influence on acquisition behavior. Already [too] highly leveraged firms are expected to have problems raising more debt, since the perceived bankruptcy costs both for the management and for the creditors increase substantially. Therefore, the only thing that can challenge this relationship is some factor altering those cost perceptions and perceived bankruptcy risks. This is especially important for current study.
The findings of the Uysal (2011) can be partially challenged. For example, Morellec and Zhdanov (2009) found, that leverage of the winning bidder in the acquisitions is generally below the industry average. This can be interpreted in the way that the companies with relatively low leverage are more likely to get the deal. Since the overleveraged companies as defined by Uysal (2011) are in the top quartile of leverage deficit, one could expect the correlation between the leverage deficit quartile and the actual leverage quartile. So those companies might be actually inclined to make the deals (instead of choosing not to, as claimed by Uysal), but are unable to due to effects studied by Morellec and Zhdanov (2006).
The works of Uysal (2011) were in progress over several preceding years with publications in lower-tier journals. Therefore, even before publishing the final paper Uysal encouraged certain scholars to replicate his studies on different samples or using different models, so he could be cited by, for example, Vasiliou et al. (2008). The preceding paper was published in 2007 (Uysal, 2007), over the revisions Uysal mostly refined the methodology and argumentation in rebuking possible alternative explanations for his findings.
optimal capital structure, acquisitions and European sample.
As one can see, this area of study can benefit from additional research that would check the findings of the previous scholars on different samples with altered methodology and would try to find moderators and mediators of the influence of leverage deficit on decision-making in mergers and acquisitions. This is done in the following part of the paper, where one explores the theoretical foundations of the effect of cross-border acquisitions and of financial system of the bidder on decision-making in acquisitions. After that the appropriate hypotheses are developed.
2.3 Hypothesis formulation
2.3.1 Hypothesis 1 – Replicating the results of Uysal (2011)
There is a need to test, at least partially, the findings of Uysal (2011) on a different sample, that would include cross-border and domestic acquisitions of European companies as opposed to domestic acquisitions of US companies. As mentioned earlier, the only partial test like this was performed by Vasiliou et al. (2008), however it cannot be deemed sufficient due to restriction to one European country, questionable choice of estimation models and lack of descriptive statistics for major variables. Another important reason to double-test the findings of Uysal (2011) is the possibility of alternative explanations to them. Uysal (2011) mentions several of them including free cash flow hypothesis (Jensen, 1986) and wealth transfer hypothesis (that makes acquisitions more valuable to underleveraged firms) (Billet, King and Mauer, 2004). While Uysal's (2011) results reject both of them, one should test them to see if such results are sample-dependent in order to consider them absolutely reliable. Generally, Uysal's (2011) findings (the part which is relevant for the needs of the current study) can be summarized as follows: leverage deficit has a negative and significant effect on acquisition probability, relative size and preference of cash as a payment method. Overleveraged companies tend to engage into acquisitions which are less frequent, smaller and payed for with equity. The mechanism of the effect lies in probability to get financing and in scrutiny over managers from the debtholders when the company is overleveraged. As a starting point I can adopt the findings of Uysal in order to formulate the hypothesis. I should keep in mind, however, that we are testing only the results regarding to probability, size and payment methods of the acquisitions due to data and time limitations.
Hypothesis 1A. There is significant and negative relationship between the leverage deficit of the company and its probability to engage into acquisitions.
Hypothesis 1B. There is significant and negative relationship between the leverage deficit of the company and the size of the acquisitions it is engaging at.
Hypothesis 1C. There is significant and negative relationship between the leverage deficit of the company and its probability to pay cash in the acquisitions.
2.3.2. Hypothesis 2 - Financial system and the influence of leverage deficit on decision-making in acquisitions
It has been established in the scientific community, that financial system of each particular country can be either attributed to bank-based ones, or to market-based ones. This is generally dependent on the major sources of financing that businesses traditionally employ in the country – whether they are provided by the banks in the forms of loans or by equity and bond markets. Kunt and Levine (1999) provide a broad overview of the criteria that can be used to consider a particular country as market-based or bank-based economy and also study the determinants for a country to become more market-oriented or bank-oriented. While the determinants are not especially relevant for the needs of current study, the distribution of the countries is later on used in the analysis. Lee (2012) also contributes to setting up the model to define the status of the country, however doubts the possible positive implications of adopting bank-based vs market-based perspective and the possible benefits of one system in relation to another. It is important to notice that market-based system implies wider usage of market-based equity and debt as a source of financing, so a country with a developed stock market, but undeveloped bond market (most of debt is still supplied by the banks) can be still attributed to market-based financial systems.
While talking about the influence of financial system on the leverage deficit relationship with acquisition behavior one should take into account two branches of research: agency theory and research on relationship lending.
management, which leads to higher agency costs of equity. The same approach can be potentially applied to bonds and bondholders – dispersed bondholder base can potentially lead to higher agency costs of debt and to lower probability of actually getting the debt financing. This is especially true when the company has to communicate the specific purpose of obtaining the money to unestablished, broad mass of individual economic agents. On the other hand, when the company is operating in a country with a prevalently bank-based financial system will more likely have to communicate the financial needs to one or a few specific entities only, which can potentially lead to higher probability of obtaining the debt financing for the investment (or acquisition), even though the company can be relatively more overleveraged. Banks have higher control over the loan and have more opportunities for an extensive analysis of both the company and the deal for which financing is intended, compared to numerous individual bondholders.
Another point, less covered by traditional theory, but covered by multiple more empirical studies is relationship lending. According to the literature (e.g. Boot, 2000) relationship lending can take place when a firm has established close ties with a particular financial institution (usually a bank). Additional important characteristic of relationship lending is that the respective financial institution engage into multiple interactions with a respective borrower over a long of period of time and through this relationship obtains unique information about the borrower (Memmel et al., 2008). This information is arguably can be seen as an aid for the lender in decision-making on whether to finance a specific acquisition or not. Since there is an information asymmetry between the mass of creditors and single banks, the banks are more likely to finance the acquisitions of overleveraged companies, since they have more opportunities to evaluate the fidelity of the borrower using proprietary information. This is supposed to influence the relationship between leverage deficit and acquisition frequency, size and cash as a payment method., effectively decreasing the effect of the deficit on acquisitions. This can be partially traced in the studies of Boot (1999) and Chakraborty and Ray (2006), who make a point, that relationship lending solves information problem and resolve agency tensions (referred to in the previous paragraph), while Degryse and Van Cayseele (2000) find that relationship lending can be associated with lower loan rate and lower probability of pledging collateral, thus indirectly giving more opportunities for overleveraged companies to borrow more.
Hypothesis 2.
The acquirers from the countries with bank-based financial systems, as opposed to market-based financial system, are significantly less influenced by the deviations from optimal capital structure in terms of acquisition frequency, size and preference of cash as a payment method.
Hypothesis 2A
There is a significant and negative relationship between the influence of leverage deficit of the company on its acquisition behavior and the company's origin from the country with relatively bank-based financial system.
It is also important to test not only the effect of financial system on the leverage deficit per se, but also its influence on the decision-making in overleveraged companies (for which leverage deficit is significantly lower than average and lower than 0). This leads us to:
Hypothesis 2B
There is significant and negative relationship between the influence of the overleveraged position of the company on its acquisition behavior and the company's origin from the country with relatively bank-based financial system.
2.3.3. Hypothesis 3 - The specifics of the impact of the leverage deficit on cross-border acquisitions
As mentioned earlier, previous studies on the influence of the leverage deficit on acquisition behavior have been, to the best of our knowledge, using predominantly the samples of domestic acquisitions. When cross-border acquisitions were included in a sample, the difference between them and the domestic ones never was a point of focus. Moreover, existing studies are focused on singled out countries. It was the US companies and deals for Uysal (2011) and for Harford et al. (2009) and it was Greek companies and deals for Vasiliou et al. (2008). However, existing research suggests that the results for a domestic sample might not be entirely similar to the ones for a cross-border sample. At very least, this similarity needs to be checked and confirmed or rejected.
cross-border acquisitions might be different from its influence on domestic deals. I will have to draw from both managerial literature and financial studies, which usually have different results.
which more depends on financiers, who also know about the characteristics of cross-border acquisitions and are possibly more reluctant to fund them when the firm is overleveraged. These two effects may be the reason for leverage deficit (and being overleveraged) to influence cross-border acquisitions stronger than domestic ones.
There is an alternative explanation to these effects, however. It is also offered by Uysal (2011), even though he finds no evidence to support it – the free cash flow hypothesis, where managers tend to create more value (hence bigger CARs from the acquisitions) when they do not have substantial free cash flow (financial slack) in their disposal (Jensen, 1986). Uysal (2011) claims these effects are not the reason for the overleveraged firms to have bigger CARs, because there is no opposite evidence – underleveraged firms are not associated with significantly lower than average CARs, while according to free cash flow hypothesis they are supposed to be. One can check if this holds true for the sample containing cross-border acquisitions and if the fact that acquisition is cross-border (hence higher risk and higher monitoring by principals) influences this relationship.
There is a certain body of counter-evidence, however, which does not allow to proceed directly to formulating the hypotheses. First of all, it has been established, that cross-border acquisition can in certain cases be the only possible mode of entry to a new foreign market (Shimizu et al., 2004), therefore even an overleveraged company will have less problems communicating the necessity of the deal in order to obtain financing. Such effects are not present for the domestic acquisitions, especially within-industry acquisitions since organic growth is usually a viable option there. Secondly, it had been established that the firms that engage in cross-border acquisitions are valued relatively higher (Erel et al., 2012) and higher valued firms typically have less obstacles in obtaining the financing (Chen, 2004). These evidence can be potentially associated with leverage deficit having less influence on cross-border acquisitions than on domestic ones.
While existing ambiguity does not allow for clear formulating of a hypothesis, I will use the evidence in favor of stronger relationship between the deviations from target leverage and cross-border acquisitions in order to formulate a set of testable hypotheses. If those are rejected, it means that the argumentation for leverage deficit having less impact on cross-border acquisitions has more ground.
domestic acquisitions.
Hypothesis 3A. The fact that acquisition is a cross-border one is associated with significantly bigger negative impact of the leverage deficit on its probability, size and preference of cash as a payment method.
There is also a need to check the relationship between the overleveraged position of the company (leverage deficit significantly lower than average and lower than 0) and its engaging into cross-border acquisitions. This leads us to:
3. Data and methodology
The sample consists of European (14 countries) listed companies and their domestic and cross-border acquisitions over the period from 2002 to 2012. Data is taken from Datastream (for companies) and Zephyr (for the deals) databases. The company data sample and the acquisition data sample were normalized for the availability of the data. Following Uysal (2011) financial firms, banks and regulated utilities were excluded from the sample. The data was winsorized at 1% level. The acquisition data was taken from Zephyr database and consisted of both the Asset and Firm acquisitions over the specified period. This leads us to 968 companies and 2506 acquisitions. The data used is both the historic data on the companies in order to calculate their optimal capital structure and acquisition-level data (combined with the data on acquisition targets) that will be used for finding the link between acquisitions and deviations from capital structure.
As suggested by Uysal (2011) the analysis consists of 2 major parts (originally offered by Hovakimian, Opler, and Titman, (2001)).
First of all, yearly OLS regressions with industry dummies are used to determine target capital structure of the companies of the sample for all the years. I follow Uysal (2011) in doing so. The general regression equation form is the following:
Market Leverageit=λ Xi ,t −1+ε1i (1) where Xi ,t −1 are one-year lagged variables suggested by existing theory. Some of them, like
Variables of the first stage (all variables are one year-lagged):
Table 1
Variables of the first stage of the analysis, their meaning and sources
Variable Name Proxy for Literature Source
Size (LN (Sales)) Firm Size Rajan and Zingales, 1995
MtB ratio (Market Value as specified above over Total Assets) (MTB)
Growth opportunities Goyal, Lehn and Racic, 2002
CAPEX/TA Current Growth Frank and Goyal, 2009
Fixed Assets/Total Assets (Fixed/TA)
Assets Tangibility Titman and Wessels, 1988, Uysal, 2011,
Nunkoo and Boateng, 2010
EBITDA/TA Profitability Uysal, 2011, Nunkoo and Boateng, 2010
Market Leverage (Book Debt over Market Value as specified earlier)
Firm fixed effects Lemmon, Roberts and Zender, 2008
Dividend Dummy (1 for the firm-year with dividends paid) (DIV)
Dividend Payments Frank and Goyal, 2009
Industry (Fama & French 48 industries)
The correlation table for the major variables is presented below.
Table 2
Correlation table for major variables of the analysis
CAPEX/TA EBITDA/TA
FIXED
ASSETS / TA LEVERAGEMARKET TO BOOKMARKET SIZE
CAPEX/TA 1.0000 0.1324 0.4899 0.1047 0.0466 0.0948 EBITDA/TA 0.1324 1.0000 0.1354 -0.0026 -0.1431 0.3185 FIXED ASSETS / TA 0.4899 0.1354 1.0000 0.3172 -0.0278 0.1483 MARKET LEVERAGE 0.1047 -0.0026 0.3172 1.0000 -0.2306 0.2568 MARKET TO BOOK 0.0466 -0.1431 -0.0278 -0.2306 1.0000 -0.2080 SIZE 0.0948 0.3185 0.1483 0.2568 -0.2080 1.0000 DIVID. DUMMY 0.1112 0.3495 0.1728 0.0185 -0.1085 0.4016
As one can see, certain variables, like Fixed Assets to Total Assets and Capex to Total assets are relatively correlated, so are Profitability, Size and Dividend Payment Frequency. However, the level of such correlation is not high enough to suspect the problem of multicollinearity. Moreover, even the possible presence of near multicollinearity does not violate fundamental assumptions of OLS estimation.
After conducting the regression I follow Hovakimian, Opler, and Titman, (2001) and Fama & French (2002) and use the coefficients to obtain the fitted variable for each company – its target market leverage for a specific year. Subtracting then from it its actual market leverage for that year I obtain a measure of deviation from target capital structure (Market Leverage Deficit) which will be used later for stage 2 of the analysis.
levels of target leverage can lead to absolutely different levels of leverage deficit and hence different underleveraged/overleveraged status.
At the second stage regression analysis is used to check the impact of the leverage deficit on decision-making in mergers and acquisitions. I adopt methodology of Uysal (2011), but I contribute 2 additional variables. The first one is a cross-border dummy (1 for cross-border acquisition, 0 for domestic acquisition). Cross-border dummy gets multiplied by a leverage deficit in order to account for joint effects. A second additional variable is a financial system dummy that should account for the influence of financial system of the country of the acquirer on the effect of the leverage deficit on mergers and acquisitions behavior. It is formed as a dummy variable becoming 1 for the country of origin of the bidder having a bank-based system and 0 for it having a market-based system. Interaction terms with leverage deficit variable and overleveraged/underleveraged dummies are also used.
The original methodology of Uysal (2011) will be restricted, since I would like to study only the effect of the leverage deficit (with the regard for cross-border/domestic and bank-based/market-based dualities) only on the frequency, relative size and payment method of the acquisitions. Hence 3 time-series regressions have been performed. The control variables are assumed, but not presented in the formulas below and will be discussed in details further.
P(Acquirer=1)=Φ
(
α10+α11LD+α12CB+α13FS +α14FS∗ LD+α15CB∗ LD+....α19Zi)
(2)P(AllCash=1)=Φ
(
α20+α21LD+α22CB+α23FS +α24FS∗ LD+α25CB∗ LD+.... α29Zi)
(3)DealValue=α30+α31LD+α32CB+α33FS +α34FS∗ LD+α35CB∗ LD+.... α39Zi (4)
where CB – cross-border dummy, FS – financial system dummy, LD – leverage deficit, the difference between the actual leverage and target leverage, calculated using Equation 1 and DealValue is the relative value of the deal, calculated as a ratio of the value of the acquisitions to total assets of the target.
following mostly Hovakimian, Opler, and Titman, (2001).
There is mixed evidence and opinion about how cross-border acquisition impacts decision-making in mergers and acquisitions. I specify a cross-border acquisition as an acquisition where the country of origin or the acquirer and a country of origin of the target (as provided by Zephyr database) are not the same.
The effect of financial system type on borrowing opportunities has been studied by numerous scholars with different results. The allocation of the countries to bank-based and market-based financial systems roots in the paper of Lee (2012) and in the defining paper of Kunt and Levine (1999) in World Bank Policy Working Papers. The data on the country of origin of the companies is taken from Datastream database.
The details of acquisition sample are provided below:
Table 3
The overview of acquisition sample
Total Data Points 11616 data firm-year data points where the firm either engaged into acquisitions or did not. It is vital to notice that an observation for probability model does not equal an acquisition – a year without an acquisition for a particular firm is equally important.
Total Acquisitions 2506 acquisitions1 with complete information about payment method and
deal value – they are the observations for size and payment method (since in order to observe both of them the Parameter Acquirer = 1 is a must.
All Cash acquisitions 886 acquisitions, where cash is the only payment method as provided by Zephyr data base
Cross-border acquisitions
1106 cross-border acquisitions, specified as when the country of the acquirer and the country of the target is not the same in the information provided by Zephyr database
1 For acquisitions I adopt the specification given by Zephyr data base, which is the main source of deal-level data. 100%
Certain statistics for the ratio of deal value to total assets which is used in equation 4:
Table 4
Descriptive statistics for the Deal Value dependent variable
Mean 0.874
Median 0.841
Maximum 2.680
Minimum 0.050
Standard Deviation 0.225
Table 5
Variables of the second stage of the analysis, their meaning and sources
Variable Name Proxy for Literature Source
Leverage Deficit (LD) (Difference between actual Market Leverage and Target Market Leverage (the fitted variable from stage 1))
Deviation from optimal capital structure
Uysal, 2011
Cross-border Dummy (CB)
Difference for cross-border and domestic acquisition
Uysal, 2011, Shimizu et al., 2004, Erel et al., 2012, Chen, 2004, Lehto, 2006
Interaction term (Cross-border dummy times Lev. Deficit) (CBLD)
Difference for cross-border and domestic acquisition and its impact on leverage deficit influence
Uysal, 2011, Shimizu et al., 2004, Erel et al., 2012, Chen, 2004, Lehto, 2006
Interaction term (financial system dummy times Lev. Deficit) (FSLD)
Difference in market/bank based financial system and its impact on the influence of the deviation from optimal capital structure
Weinstein and Yafeh, 1998; Degryse and Van Cayseele, 2000; Boot, 1999; Chakraborty and Ray, 2006; Mehrotra et al., 2011; Heidhues and Patel, 2008; Lee, 2012
Financial System Dummy (of Acquirer) (FS)
Difference in market/bank based financial system
Weinstein and Yafeh, 1998; Degryse and Van Cayseele, 2000; Boot, 1999; Chakraborty and Ray, 2006; Mehrotra et al., 2011; Heidhues and Patel, 2008; Lee, 2012
Deal Value (relative deal value; Deal Value / TA of the target)
Relative Size of the Target Uysal, 2011; Lehto, 2006
Variable Name Proxy for Literature Source Overleveraged Dummy
(OVER)
Leverage Deviation Direction Uysal, 2011
Underleveraged Dummy (UNDER)
Leverage Deviation Direction Uysal, 2011
Underleveraged*Financial System (UNDERFS)
Leverage Deviation Direction & Interaction with the Financial System effects
Uysal, 2011, Weinstein and Yafeh, 1998; Degryse and Van Cayseele, 2000; Boot, 1999; Chakraborty and Ray, 2006; Mehrotra et al., 2011; Heidhues and Patel, 2008; Lee, 2012 Overleveraged*Financial
System (OVERFS)
Leverage Deviation Direction & Interaction with the Financial System effects
Uysal, 2011, Weinstein and Yafeh, 1998; Degryse and Van Cayseele, 2000; Boot, 1999; Chakraborty and Ray, 2006; Mehrotra et al., 2011; Heidhues and Patel, 2008; Lee, 2012 Underleveraged*Cross-
border Dummy (UNDERCB)
Leverage Deviation Direction & Interaction with the specificity of cross-border acquisitions
Uysal, 2011, Shimizu et al., 2004, Erel et al., 2012, Chen, 2004, Lehto, 2006
Overleveraged*Cross- border Dummy (OVERCB)
Leverage Deviation Direction & Interaction with the specificity of cross-border acquisitions
Uysal, 2011, Shimizu et al., 2004, Erel et al., 2012, Chen, 2004, Lehto, 2006
Control Variables SIZE (LN Sales)
(acquirer)
Size of the acquirer Almazan, de Motta, Titman and Uysal, 2010
EBITDA/TA (acquirer) Performance of the acquirer Harford, 1999 Industry M&A2 Liquidity
(LIQ)
M&A Wave Schlingemann, Stulz and Walkling, 2002
Herfindahl Index (HHI)3 Industry concentration Uysal, 2011
2 Total Deal Value of Acquisitions within an industry within a year divided by Total Assets of the industry
(cross-country-level data).
Variable Name Proxy for Literature Source Trailing 2-year MLEV
(AMLEV)
Leverage Level Uysal, 2011
MtB ratio (acquirer) (MTB)
4. Analysis and results
4.1. Stage 1
In order to estimate the target leverage for the specified companies in specified years numerous variables are used. The variables and their predicted signs are specified by the existing theory.
Size is represented by natural logarithm of Net Sales of the company (as specified in Datastream). Rajan and Zingales (1995) make a claim that bigger companies have less volatile cash flows and have easier access to capital market which, in turn, allows them to have higher leverage ratios.
Frank and Goyal (2009) make a point that higher growth opportunities of the company are supposed to decrease the target leverage of the company. The seminal paper of Uysal (2011) uses two proxies for growth opportunities – Market-to-Book ratio and Stock Returns which have the opposite signs in the results, while being highly significant. I follow Uysal (2011) in using Market-to-Book ratio, since in his results the influence of this variable was in coherence with existing theory. It is generally accepted (and confirmed by Titman and Wessels (1988) and numerous other studies) that higher growth opportunities are expected to have a negative effect on market leverage.
I follow Goyal and Frank (2003) by including a ratio of CAPEX to Total Assets. This ratio is a proxy for current growth via current investments. Since higher growth is associated with higher capital requirements, it is expected for the relationship to be positive.
Asset tangibility is associated with higher target leverage: firms are more likely to borrow against these assets (Titman and Wessels, 1988). I follow Nunkoo and Boateng (2010) and use the ratio of Fixed Assets to Total Assets as a proxy for asset tangibility in a situation of data limitations.
According to the results of Frank and Goyal (2009), dividend-paying firms are likely to have lower target leverage. However, as they state in their earlier paper (Goyal and Frank, 2003), under the pecking order theory the dividends are a part of the financing deficit and thus are associated with more (debt) financing requirements. Thus I add a dummy that accounts for (not)payment of the dividends by the company in the relevant year.
potential financial distress and hence it is associated with increased market leverage. They conclude by stating that the predictions about this relationship are unclear. I follow Uysal (2011) and use the ratio of EBITDA to Total Assets as a proxy for profitability of the company, however no specific predictions about the sign of the variable are made.
There is also a need to account for firm fixed effects. I follow Uysal (2011) and Lemmon, Roberts and Zander (2010) by using a lagged Market Leverage ratio in order to account for such effects.
I also account for industry fixed effects by adding industry dummies based on Fama and French (1997) industry classification that has been widely used in the relevant literature. The companies are grouped into industries according to their 4-digit SIC codes, which have been obtained from Orbis Database. The descriptive statistics of the major control variables of stage one after winsorization are provided in the table below:
Table 6
Descriptive statistics of major variables of the first stage
CAPEX/TA EBITDA/TA FIXED/TA
MARKET LEVERAGE (t-1) MARKET to BOOK MARKET LEVERAGE (t) (Dependent) SIZE DIVIDEND DUMMY Mean 0.048997 0.093696 0.257124 0.171874 1.852514 0.173384 12.73651 0.709797 Median 0.036153 0.115790 0.210825 0.142276 1.284336 0.143250 12.77466 1.000000 Maximum 0.468266 0.715614 1.533408 0.774373 49.69236 0.797596 19.23041 1.000000 Minimum -0.058641 -3.151759 0.000000 0.000000 0.324255 0.000000 0.000000 0.000000 Std. Dev. 0.050099 0.213571 0.214157 0.153240 2.749622 0.153949 2.651530 0.453876 Skewness 2.905126 -7.492521 1.311352 0.936159 10.86617 0.953090 -0.822882 -0.924508 Kurtosis 16.94768 96.13706 5.310978 3.433381 159.1385 3.505896 6.372076 1.854716 Jarque-Bera 110495.7 4307147. 5914.088 1787.604 12028143 1882.495 6814.446 2289.581 Probability 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 0.000000 Sum 569.1438 1088.369 2986.757 1996.490 21518.80 2014.032 147947.3 8245.000 Sum Sq. Dev. 29.15276 529.7880 532.6996 272.7509 87814.32 275.2782 81660.55 2392.725 Observations 11616 11616 11616 11616 11616 11616 11616 11616
same variable with one year lag.
Due to the danger of heteroscedastisity White's robust standard errors are used on both stages of the research.
I follow Harford et al. (2009) and conduct yearly regressions. The results and the corresponding p-values are presented in Table 7. Industry dummies are omitted from the table. Due to abundance of them (one dummy variable for each of 48 industry categories) the yearly equations do not contain a constant, so that the equations do not fall into the dummy variable trap.
Table 7
The results of 11 yearly regressions. The p-values are presented in parentheses below the coefficients. The dependent variable is Market Leverage. All independent variables are at t-1. Accounting for industry effects did not allow the usage of constant in the model. White's robust standard errors are used. 11616 company-year
observations are used
Variable Coefficient
2001 Coefficient2002 Coefficient2003 Coefficient2004 Coefficient2005 Coefficient2006 Coefficient2007 Coefficient2008 Coefficient2009 Coefficient2010 Coefficient2011 Coefficient2012
4.2. Stage 2
On the second the stage the Market Leverage Deficit variables are calculated by subtracting the Target Leverage Ratios from Market Leverage Ratios. After that, leverage deficits are organized into 4 quartiles (on yearly basis) the lowest quartile represents underleveraged companies and the highest – overleveraged ones (where actual market leverage ratio is vastly higher than the optimal one). After that leverage deficits and overleveraged/underleveraged dummy variables are used in 6 separate estimations – 2 different model types (as suggested by Uysal (2011)) in order to estimate the effect of leverage deficit on probability of acquiring the company, the relative size of the company being acquired and the payment method combined with the effect of the cross-border type of the acquisition and of the financial system of acquirer's country of origin on the leverage deficit.
Besides the specified variables I follow Uysal (2011) in using a set of control variables.
Following Harford (1999) I include EBITDA/TA as a measure of accounting performance. It is accepted that better performing companies engage into more acquisitions.
I follow Uysal (2011) by using 2-year trailing average Market Leverage (not 3-year as in original paper due to data availability limitations) in order to account for leverage effects and Herfindahl Index in order to account for industry concentration – higher industry concentration is generally associated with less acquisition opportunities.
Bigger firms engage into more acquisitions and bigger acquisitions thus LN (Sales) is used as a proxy for size of the firms (Uysal, 2011). Another control variable is a construct of Industry M&A Liquidity which accounts for the liquidity in the corporate assets market (Schlingemann, Stulz and Walkling, 2002). Higher liquidity is naturally associated with more acquisitions. Finally, I follow Uysal (2011) in using Market-to-book Ratio of the acquirer in order to account for possible mis-valuation of the company performed by the management of the acquirer.
bank-based financial system), it will get the same value in all other years, without any regard of whether the acquisitions took place or not. Cross-border dummy, however, is a deal-level variable. Thus it has a value of 1 when an acquisition took place and it was a cross-border one according to previously stated definition. However, it has a value of 0 for both cases when the acquisition was domestic, and when acquisition did not happen for a particular firm in a particular year. Thus, it is impossible to include cross-border dummy as a separate variable in the model that predict acquisitions, because it has inflated prediction properties – every value of 1 for cross-border dummy is associated with a dependent variable also having the value of one. This is an important shortcoming of the model and therefore it was explained in such a detail.
I specify two major model types. One is using the absolute level of leverage deficit (variable LD) and the interaction terms of it with dummy variables for cross-border acquisition and for bank-based financial system. Another is using overleveraged/underleveraged dummies and the interaction terms of them and financial system dummies. Second model type cannot evaluate the impact of cross-border nature of the acquisition due to perfect correlation between cross-border dummy and a dependent variable (Acquirer = 1). Binary Logit is used to estimate both model types for researching the effects on the probability of conducting an acquisition and on paying cash. Tobit is used in order to estimate the effects of leverage deficit on the size of the target. Yearly dummies are included in every model as suggested by Uysal (2011).
Table 8
The dependent variable for probability is Acquirer=1, for payment method: All Cash = 1 and for Size – Deal Value, ratio of Acquisition value to acquirer's TA. Robust standard errors and yearly dummies are used in all models. . *,**,*** refer to 10%, 5% and 1% significance levels. Pseudo R squared is used for all the models,
Probability and Payment Method use logit as a model of estimation, while Size models use tobit and are censored at 0.
Probability (11616 observations) Size (2506 observations) Payment Method (2506 observations)
Model 1 Model 2 Model 1 Model 2 Model 1 Model 2
Variable Coefficient z-stat Coefficient z-stat Coefficient z-stat Coefficient z-stat Coefficient z-stat Coefficient z-stat
5. Discussion of the empirical results
The results obtained during the analysis are quite contradictory. While the signs and sig-nificance of control variables generally follow the existing theory and thus add to con-firming the correct specification of the model, the major variables gave results different from prior predictions and the results are sometimes inconsistent between two model types. It will be best to discuss the results in three separate sections.
5.1. Probability
The signs of leverage deficit (LD) and underleveraged/overleveraged dummies (UNDER/OVER) generally follow the results of the seminal paper of Uysal (2011), how-ever, while for Uysal (2011) these variables were highly significant, in these results it is not the case. This allows us to effectively reject Hypothesis 1A. It is important to notice the marginal significance of underleveraged dummy (UNDER), which was not significant in the original study. Moreover, Uysal's rejection (2011) of free cash flow hypothesis-based explanation (Jensen, 1986) for his results was hypothesis-based exactly on non-significance of the underleveraged dummy. Now, however, we can see that with different model specifi-cation and sample item UNDER can be significant, therefore alternative explanations to both the results of Uysal (2011), and the results of this paper are possible.
Cross-border dummy (CBLD) was found to significantly diminish the influence of lever-age deficit (see opposite signs of leverlever-age deficit (LD) and interaction term (CBLD)), thus effectively rejecting Hypothesis 3A for probability. However, as I mentioned in the literature review, alternative explanations and theories exist that could potentially explain such behavior, such as possible value-creation in the cross-border acquisitions, higher rel-ative valuation of the firms that engage in it and other factors. This ambiguity in the re-sults asks for additional studies on the impact of leverage deficit on cross-border acquisi-tions, maybe, for a separate study with a purely cross-border sample.
sign – they decrease the effect of underleveraged/overleveraged position of the company on its acquisition behavior. However, the coefficients of these variables are not signifi-cant, so we have to reject Hypothesis 2A. The non-significance of these results could have been stemming from data selection. On the other hand, the effects, described during hypothesis formulation could be existent, but too weak to somehow significantly influ-ence the relationship between the deficit and acquisition behavior.
5.2. Size
The sign of leverage deficit variable (LD) also follows the predictions of Uysal (2011), however its insignificance makes us reject the Hypothesis 1B. Both cross-border dummy (CB) and financial system dummy (FS) are significant, which has a clear economic meaning: firms perform smaller cross-border acquisitions due to their higher risk (Lehto et al., 2006) and companies from the countries with bank-based financial system perform bigger acquisitions due to the possibility of relationship banking and easier access to debt capital – without the regard for leverage deficit.
Cross-border interaction (CBLD) term is positive and significant, which allows us to re-ject the Hypothesis 3 for size (once again emphasizing the potential validity of above-mentioned alternative effects of leverage deficit on cross-border acquisitions), while the insignificance of financial system interaction (FSLD) term makes us reject the Hypothe-sis 1 for size as well.
which implies that effects of overleveraged position are of lower magnitude in bank-based financial systems. This allows us to partially confirm Hypothesis 2 for size, how-ever, caution must be used since the actual influence of overleveraged dummy (OVER) deviates from both the previous predictions made by us, and the actual existing results of Uysal (2011). It is also important to notice, that HHI, one of the control variables, devi-ates from predictions here. This might be caused by the fact that now this is a cross-bor-der, within-industry construct, while originally, in the study of Uysal (2011) it was a do-mestic, within-industry concept.
5.3. Payment Method Choice
The results of payment method choice estimation also differ from existing findings of Uysal (2011). The leverage deficit variable (LD) is not significant, even though the sign of it reflects the expectations. Same situation is with overleveraged dummy (OVER), while the underleveraged dummy (UNDER) is positive and significant. This allows us to reject the Hypothesis 1C and makes us suspect the viability of alternative explanations to Uysal's (2011) results, especially, the free cash flow hypothesis. Another notable point is the positive sign and significance of cross-border dummy (CB) and partially financial system dummy (FS) – cross-border acquisitions are paid for in cash more often, while in countries with bank-based financial system cash is used more frequently to fund the complete acquisition.
The insignificance of the interaction terms for financial system dummy (OVERFS and UNDERFS) allows us to reject the Hypothesis 2 for payment method. The interaction term between cross-border dummy and underleveraged dummy (UNDERCB) is negative and significant, this allows us to reject Hypothesis 3 for payment method as well, since opposite signs for underleveraged dummy and the interaction term mean that cross-border acquisitions are less positively influenced by being underleveraged.
6. Conclusion
This study contributes to the literature by questioning the applicability of the findings of Uysal (2011) on a European sample that includes both domestic and cross-border acquisi-tions, while checking whether financial system of the acquierer and the cross-border component of the acquisition influence the impact of leverage deficit on the acquisition probability, size and payment method. The paper provides a contribution to capital struc-ture literastruc-ture, target capital strucstruc-ture (and deviations from it) studies and to acquisition literature as well.
The findings of this study are relatively inconclusive. I could not fully replicate the find-ings of Uysal (2011) on a different sample: while the signs of the coefficients (and thus the nature of the relationship) were generally consistent with previous research, the sig-nificance of them was not. Moreover, certain findings, like positive and significant im-pact of underleveraged dummy on probability and on preference of cash as a payment method raise questions on the possible validity of alternative explanations to findings of Uysal. This is a strong call for another study replicating Uysal's (2011) result on a differ-ent, vast sample with a similar methodology and also for a simultaneous test of the alter-native explanations to such results like free cash flow hypothesis. Such extensive study would definitely be beneficial for our understanding of the existence of leverage deficit and its influence on acquisitions.
on cross-border acquisitions, for example, taking the Uysal's (2011) sample of US com-panies, but using their cross-border acquisitions instead.
This study has certain limitations and it, as mentioned before, uncovers multiple opportu-nities for future research. Data availability limitations could potentially induce a certain bias on the results. The model specification for finding the influence of cross-border ac-quisitions on acquisition probability is also relatively biased. Certain controls, like HHI and Industry M&A Liquidity are constructed on cross-country and within-industry basis, which is not an absolutely conventional construction of them. Finally, while the yearly differences were taken into account, possible time-series outliers were not.
The implications for managers are not that clear, since the paper did not provide a final solution to the questions of the influence of the leverage deficit on acquisition behavior. One can state, however, that companies should definitely consciously take their leverage position into account when making decisions on mergers and acquisitions, especially the cross-border ones. Knowing what your target (and optimal) capital structure is and how much is acceptable to deviate from it is always a good practice.
7. References
1. Agliardi, E., Koussis, N., & Amel-Zadeh, A. (2012). Optimal Capital Structure and Growth Options in Mergers and Acquisitions. Available at SSRN 2116335. 2. Almazan, A., De Motta, A., Titman, S., & Uysal, V. (2010). Financial structure,
acquisition opportunities, and firm locations. The Journal of Finance, 65(2), 529-563.
3. Barclay, M. J., & Smith, C. W. (1999). The capital structure puzzle: another look at the evidence. Journal of Applied Corporate Finance, 12(1), 8-20.
4. Billett, M.T., King, T.D., Mauer, D., 2004. Bondholder wealth effects in mergers and acquisitions: new evidence from the 1980s and 1990s. Journal of Finance 59, 107–135.
5. Boot, A. W. (2000). Relationship banking: What do we know?. Journal of financial intermediation, 9(1), 7-25.
6. Cakici, N., Hessel, C., & Tandon, K. (2000). Foreign acquisitions in the United States: Effect on shareholder wealth of foreign acquiring firms. Journal of Banking & Finance, 20(2), 307-329.
7. Chakraborty, S., & Ray, T. (2006). Bank-based versus market-based financial systems: A growth-theoretic analysis. Journal of Monetary Economics, 53(2), 329-350.
8. Chen, J. J. (2004). Determinants of capital structure of Chinese-listed companies. Journal of Business research, 57(12), 1341-1351.
9. Datta, D. K., & Puia, G. (1995). Cross-border acquisitions: An examination of the influence of relatedness and cultural fit on shareholder value creation in US acquiring firms. MIR: Management International Review, 337-359.
Intermediation, 9(1), 90-109.
11. Demirgüç-Kunt, A., & Levine, R. (1999). Bank-based and market-based financial systems: Cross-country comparisons. World Bank Policy Working Paper, (2143). 12. Erel, I., Liao, R. C., & Weisbach, M. S. (2012). Determinants of Cross Border ‐
Mergers and Acquisitions. The Journal of Finance, 67(3), 1045-1082.
13. Fama, E. F., & French, K. R. (1997). Industry costs of equity. Journal of financial economics, 43(2), 153-193.
14. Frank, M. Z., & Goyal, V. K. (2009). Capital structure decisions: which factors are reliably important?. Financial Management, 38(1), 1-37.
15. Goyal, V. K., Lehn, K., & Racic, S. (2002). Growth opportunities and corporate debt policy: the case of the US defense industry. Journal of Financial
Economics, 64(1), 35-59.
16. Graham, J. R. (2000). How big are the tax benefits of debt?. The Journal of Finance, 55(5), 1901-1941.
17. Harford, J. (1999). Corporate cash reserves and acquisitions. The Journal of Finance, 54(6), 1969-1997.
18. Harford, J., Klasa, S., & Walcott, N. (2009). Do firms have leverage targets? Evidence from acquisitions. Journal of Financial Economics, 93(1), 1-14. 19. Harris, M., & Raviv, A. (1991). The theory of capital structure. the Journal of
Finance, 46(1), 297-355.
20. Hovakimian, A., Opler, T., & Titman, S. (2001). The debt-equity choice. Journal of Financial and Quantitative analysis, 36(1), 1-24.
22. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of financial economics, 3(4), 305-360.
23. Lee, B. S. (2012). Bank-based and market-based financial systems: Time-series evidence. Pacific-Basin Finance Journal, 20(2), 173-197.
24. Lehto, E. (2006). Motives to restructure industries: Finnish evidence of cross border and domestic mergers and acquisitions.‐ Papers in Regional Science,85(1), 1-22.
25. Lemmon, M. L., Roberts, M. R., & Zender, J. F. (2008). Back to the beginning: persistence and the cross section of corporate capital structure.‐ The Journal of Finance, 63(4), 1575-1608.
26. Markides, C. C., & Ittner, C. D. (1994). Shareholder benefits from corporate international diversification: Evidence from US international acquisitions.Journal of International Business Studies, 343-366.
27. Memmel, C., Schmieder, C., & Stein, I. (2008). Relationship lending-Empirical evidence for Germany (No. 2008/01). Economic and financial reports/European Investment Bank.
28. Mehrotra, V., Van Schaik, D., Spronk, J., & Steenbeek, O. (2011). Creditor-Focused Corporate Governance: Evidence from Mergers and Acquisitions in Japan. Journal of Financial and Quantitative Analysis, 46(4), 1051.
29. Modigliani, F., & Miller, M. H. (1958). The cost of capital, corporation finance and the theory of investment. The American economic review, 48(3), 261-297. 30. Moeller, S. B., & Schlingemann, F. P. (2005). Global diversification and bidder
31. Morck, R., & Yeung, B. (1992). Internalization: an event study test. Journal of International Economics, 33(1), 41-56.
32. Morellec, E., & Zhdanov, A. (2008). Financing and takeovers. Journal of Financial Economics, 87(3), 556-581.
33. Myers, S. C. (1984). The capital structure puzzle. The journal of finance, 39(3), 574-592.
34. Nunkoo, P. K., & Boateng, A. (2010). The empirical determinants of target capital structure and adjustment to long-run target: evidence from Canadian
firms. Applied Economics Letters, 17(10), 983-990.
35. Rajan, R. G., & Zingales, L. (1995). What do we know about capital structure? Some evidence from international data. The journal of Finance, 50(5), 1421-1460. 36. Rhodes Kropf, M., & Viswanathan, S. (2004). Market valuation and merger ‐
waves. The Journal of Finance, 59(6), 2685-2
37. Seth, A., Song, K. P., & Pettit, R. R. (2002). Value creation and destruction in cross border acquisitions: an empirical analysis of foreign acquisitions of US ‐ firms. Strategic Management Journal, 23(10), 921-940.
38. Schlingemann, F. P., Stulz, R. M., & Walkling, R. A. (2002). Divestitures and the liquidity of the market for corporate assets. Journal of Financial Economics,64(1), 117-144.
39. Shimizu, K., Hitt, M. A., Vaidyanath, D., & Pisano, V. (2004). Theoretical foundations of cross-border mergers and acquisitions: A review of current research and recommendations for the future. Journal of International Management, 10(3), 307-353.
41. Titman, S., & Wessels, R. (1988). The determinants of capital structure choice.The Journal of finance, 43(1), 1-19.
42. Uysal, V. (2010, November). Deviation from the Target Capital Structure and Acquisition Choices. In AFA 2007 Chicago Meetings Paper.
43. Uysal, V. B. (2011). Deviation from the target capital structure and acquisition choices. Journal of Financial Economics, 102(3), 602-620.
44. Vasiliou, Dimitrios, Nikolaos Eriotis, and Nikolaos Daskalakis. 2008. "Target Capital Structure and Acquisition Choices: Evidence from the Greek Market." «SPOUDAI», Vol. 58, No 1-2, University of Piraeus pp. 274-284
45. Shleifer, A., & Vishny, R. W. (2003). Stock market driven acquisitions. Journal of financial Economics, 70(3), 295-311.
