1
University of Amsterdam
Master Finance, Corporate Finance
Master Thesis
How do Mergers and Acquisitions affect the
Firm’s Capital Structure?
A study on European M&A’s.
Author: Damon Nijland
Student Number: 10813705
Thesis Supervisor: dr. T. (Tolga) Caskurlu Date: 1st July 2018
2
Statement of originality
This document is written by Damon Nijland who declares to take full responsibility for the contents of this document.
I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.
The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.
3 Abstract
This paper seeks to investigate the effect that mergers and acquisitions have on the acquirers’ capital structure. This paper focusses on European M&A’s between 2000-2013 and seeks to find whether there is a difference between the results of this paper and those of previous research on American and Australian M&A’s. By using a partial adjustment framework, this paper finds significant evidence that firms more actively rebalance their capital structure in the year of the merger/acquisition. This evidence is in line with the trade-off theory of debt, that suggests that firms have an optimum level of debt that they strive towards. In the year of the acquisition, firms
rebalance their capital structure by 35.19%, in order to close the gap to the optimum level of debt. This thesis also looks at what firms characteristics play a role in the determination of the method of payment. The results of these tests are in line with the expectations of the pecking order theory and the trade-off theory, but there are no significant results that support the market timing theory.
4 Table of Contents Statement of Originality……….………2 Abstract………..3 Table of Contents………4 1. Introduction………..5 2. Literature review………7
2.1. Mergers and Acquisitions in Europe……….7
2.2. Capital Structure theories……….…8
2.2.1. Modigliani and Miller I and II………..9
2.2.2. The Static Trade-off Theory………..10
2.2.3. The Dynamic Trade-off Theory………11
2.2.4. The Pecking Order Theory………..11
2.2.5. The Market Timing Theory……….12
2.3. The effect of M&A’s on Capital Structure………..12
3. Methodology and Hypothesis………13
3.1. Financing of Mergers and Acquisitions….………..13
3.2. Methodology for Target Leverage………..14
3.3. Partial Adjustment Framework……..………15
3.4. Hypothesis Forming………16
4. Data and Descriptive Statistics………..17
4.1. Data………18
4.2. Descriptive Statistics……..……….………..…………..20
5. Results………22
5.1. Financing of M&A’s………...………22
5.2. Results of the Partial Adjustment Model……….…………..26
6. Conclusion……….…………..……….……….30
References……….……….……….……….………..33
5
1. Introduction
This paper seeks to find conclusive evidence on the effect that mergers and acquisitions have on the acquirers’ capital structure. When making an acquisition, acquirers face a certain shock to their capital structure. The trade-off theory of debt (Myers, 1984) suggests that firms move towards an optimum level of debt (further explained later). Thus, after firms face a shock to their capital
structure, their adjustment speed should increase, because they need to more firmly rebalance their capital structure in order to get to the optimum level of debt. This paper uses a partial adjustment model, which consists of two parts. In the first part, the target level of leverage is estimated and the second part consists of the partial adjustment process, in which the actual difference in capital structure is regressed on the estimated difference in capital structure.
This partial adjustment framework studies completed and confirmed European mergers and acquisitions between 2000-2013. This paper follows Hovakimian and Li (2011) their measurement for the predicted target leverage. In order to find the desired effect, a period of three years before up until three years after the acquisition is used. For each year t, the predicted change in leverage ratio is calculated as the predicted leverage ratio in year t+1 minus the actual debt ratio in year t. The actual change in the leverage ratio is found by subtracting the actual debt ratio in year t from the actual debt ratio in year t+1. By using a partial adjustment model, the speed of adjustment is calculated by regressing the dependent variable Y (actual change in leverage ratio) on the independent variable X (predicted change in leverage ratio).
In this partial adjustment framework, historic data is used in order to estimate the predicted level of debt. According to Chang and Dasgupta (2009), this eliminates the look-ahead bias and leads to better results. Chang and Dasgupta (2009) argue that previous literature regarding this topic provides results that are positively biased. Therefore, this thesis uses the methodology introduced by Hovakimian and Li (2011), to eliminate the biases. This should lead to improved results over previous research. The results of this paper support the trade-off theory, since they show that firms actively move towards a target level of debt in all years surrounding an acquisition. It is also shown that in the year of the acquisition, the adjustment speed is largest. This states that a shock to a firm’s capital structure, coming from a merger or acquisition, leads to more rebalancing of the capital structure.
Besides this partial adjustment framework, this paper also uses three logit models to investigate what firm characteristics determine the method of payment. Distinction is made between three methods of payment: completely through equity, completely through debt and a mixed method of payment. The logit models have a dummy variable as the dependent variable, which represents what type of method firms use. These logit models test the capital structure
6
theories. The results show that, for European acquirers, both the trade-off theory and the pecking order theory hold. There is no significant evidence that the market timing theory holds.
The topic for this thesis is interesting, since over the last years, mergers and acquisitions in Europe have occurred in large numbers. According to the institute for Mergers, Acquisitions and Alliances (IMAA), there were 422 European deals in 1985, compared to the 2385 deals in North America. 1993 was the first year that there were more European deals and in the last few years (2015, 2016 and 2017) there have even been more deals in Europe, compared to North America. Growth in the European mergers and acquisitions is due to multiple reasons. Faccio and Masulis (2005) state that the differences in rules and regulation between the US and Europe make for large differences in the results. Altunbaş and Marqués (2008) state that the introduction of the Euro has led to a large increase in mergers and acquisitions in Europe. This makes sense, since more countries started to use the same currency. This leads to easier and more rapid transactions, which also reduces the transaction costs of a merger/acquisition. They also state that technological innovation and financial globalization have had a large impact on the increase of mergers and acquisitions.
However, even though research on European mergers and acquisitions has grown, there are still less papers regarding European M&A’s when compared to U.S.’ M&A’s. Therefore, this paper focusses on the European M&A’s and, more specifically, on their capital structures. The aim of this study is to establish whether or not there is a significant change in the acquirers’ capital structure around the time of a merger or acquisition in Europe. Also, this paper investigates how certain firm characteristics determine the method of payment for firms.
The capital structure of a firm is one of the most important characteristics. On one hand, holding more leverage leads to an increase in the chance of bankruptcy (Myers, 1984). On the other hand, firms that hold less debt do not maximize their value, since debt can lead to certain tax savings, according to Kraus and Litzenberger (1973), who show that tax savings is the most
important benefit of holding debt. Over the years, multiple capital structure theories are developed. Since this paper also looks at how firms finance their acquisition, whether they choose to finance through equity, debt or a mixed method of payment, different capital structure theories (mainly the trade-off theory, pecking order theory and the market timing theory) are tested.
This paper contributes to multiple branches of literature. First off all, this paper contributes to previous literature on European mergers and acquisitions. Data on European M&A’s allows for more information on private firms. Previous literature shows that European firms differ from their U.S. counterparts. Differences between the U.S. and Europe arise because of multiple reasons (as stated before) and therefore it is interesting to investigate this matter for European firms.
7
capital structure theories date back to the two propositions by Modigliani and Miller (1958 & 1963). Modigliani and Miller formed the basis of the capital structure theories and almost every paper refers to their theories as the basis of capital structure theories. After Modigliani and Miller I and II, multiple theories have been developed. This paper contributes to existing literature on the trade-off theory, the pecking order theory and the market timing theory, because the logit models provide further insight into which capital structure theories are relevant for European firms that take part in a merger or acquisition.
Finally, this paper contributes to existing literature on the relationship between capital structure and mergers and acquisitions. Following previous research of Hovakimian and Li (2011) and Koh et al. (2011), this paper uses a partial adjustment model to answer the research question: ‘’What is the effect of Mergers and Acquisitions on the acquiring firm’s capital structure?’’. Where previous research regarding this topic study firms in the United States or firms in Australia, this paper contributes to them by studying the effect for European firms.
The rest of this thesis is structured as follows. In section 2, relevant previous literature is described, regarding the difference between European and U.S.’/Australian mergers and
acquisitions, different capital structure theories and previous research on the relationship between capital structure and mergers and acquisitions. Section 3 describes the methodology and the hypotheses for this paper. Section 4 provides an explanation about the data collection and shows the correlation matrix between the variables. This section also shows the descriptive statistics of the sample. Section 5 displays the results of the two different tests (logit models and the partial
adjustment framework). Finally, section 6 provides a conclusion and a discussion on the paper. 2. Literature Review
This section provides an overview of the relevant literature for this thesis. For each section, several papers are described and how this thesis contributes to them is explained as well. The first section provides an overview of literature on European mergers and acquisitions and shows the differences between studies in Europe and the United States. The second section describes the important capital structure theories and this thesis’ contribution to them. The final section looks at previous research on the relationship between M&A’s and capital structure.
2.1 Mergers and Acquisitions in Europe
First off, this paper contributes to existing literature on M&A’s in Europe. Studying the effect of M&A’s in Europe allows for more information on private firms plus European firms are more closely held. Faccio and Masulis (2005) investigate the payment choices of European bidders in M&A’s. They find that European firms choose to finance their bids through stock financing, when their financial
8
condition is weakened. These results differ from the results of Martin (1996) and Gadhoum et al. (2003), who investigate this issue for the United States. According to Faccio and Masulis (2005), the differences between the results in Europe and the results in the U.S. arise because of differences in rules and regulation. Martynova and Renneboog (2009) follow up on this topic and study the bidder’s choice of financing. They show that internally financed M&A’s perform worse and provide evidence supporting the pecking order theory (Section 2.2.24 describes the pecking order theory).
Altunbaş and Marqués (2008) study the impact of similarities between bidders and targets on the financial performance post-merger for European Union banks. They find evidence that mergers lead to improved bank performance. The increase in performance is greatest following a cross-border merger, however the increase is also significant for domestic mergers (1.2%). Their study followed recent studies on U.S. bank mergers1, where the most papers find no change in the stock returns post-merger. The biggest difference in the stock returns is in the case of cross-border mergers and acquisitions. Altunbaş and Marqués (2008) state that these difference tend to arise because of capital structures. For cross-border M&A’s, the differences between capital structure cause a lower performance.
Martynova and Renneboog (2006) provide an overview of mergers and acquisitions in the European market. They state that this is one of the few papers investigating the main features of European M&A’s. They find that cross-border M&A’s occur more often than domestic ones and that all M&A’s are mainly between firms in the same industry. They also find that, in the 1990s, the amount of hostile bids increased and that firms choose to finance their acquisitions more through a combination of debt and equity, instead of financing through cash. Other characteristics of takeovers that they show are that 70% of the acquisitions are domestic ones, 54% are all-cash offers and that 93% are friendly takeovers. The characteristics show that mergers happened for different reasons: Martynova and Renneboog (2006) state that cost cutting, expanding and acting on the mispriced premium are the main reasons. Finally, they state that the M&A’s increased the stock prices at the announcement date. This states that the public thought that M&A’s would increase the efficiency of firms.
2.2 Capital Structure Theories
The second part of this literature review is based on the capital structure theories. Over the last decades, the capital structure puzzle has been the center of a lot of research2. The capital structure
1 They refer to Amihud et al. (2002) and Houston and Ryngaert (1994), who both investigate the effect of
mergers on bank performance for U.S. firms.
2 Myers’ (1984) ‘’Capital structure puzzle’’ and Brealy and Myers (2003) their ‘’Principles of corporate finance’’
9
choice of firms is important because it can give explanations as to why firms finance their
merger/acquisition through cash or debt, which is what this paper seeks to investigate. Furthermore, this paper looks at the capital structure of firms before and after the acquisition. The capital
structure theories give further explanation as to why firms change, or don’t change, their capital structure before and after an acquisition. Fama and French (2005) provide evidence that for stock-financed mergers new equity issues are more frequent. Therefore, it is interesting to study the method of financing and how this affects the capital structure.
2.2.1 Modigliani and Miller I & II
Theories on capital structure date back to the first study on capital structure of Modigliani and Miller (1958). Their first proposition states that the value of a firm is not related to its capital structure. The main assumption in this statement is that there are perfect capital markets. In the case of perfect capital markets, there are no taxes, no transaction costs, no agency costs, no bankruptcy costs etc. Their first proposition is known as the irrelevance theory (Berk and DeMarzo, 2007) and is
formulated as VU = VL. In this case, VU is the value of the unlevered firm and VL is the value of a levered firm. This propositions claims that holding more debt does not increase the value of a firm and that firms are indifferent between debt financing and other forms of financing.
Modigliani and Miller, however, make a correction to their initial theory in 1963. In their second proposition, they show that a firm’s capital structure does have an effect on the value. Berk and DeMarzo (2007) formulate that in the second proposition, the cost of capital of levered equity is equal to the cost of capital of unlevered equity, plus a certain premium. This premium depends on the debt-to-equity ratio of a firm: rE = rU + DE(rU – rD).
• rE = Cost of capital of levered equity • rU = Cost of capital of unlevered equity • D
E = Debt-to-equity ratio • rD = Cost of debt
Where Modigliani and Miller (1958) make the assumption of perfect capital markets, Kraus and Litzenberger (1973) provide evidence that the main benefit of leverage is the tax savings. These tax savings are, as mentioned before, one of the imperfections of perfect capital markets. Kraus and Litzenberger show that holding more debt leads to less taxes paid and thus holding more debt benefits the firm. This benefit is called the ‘’Tax shield’’. Wrightsman (1978) shows how the tax shield should be valued. He demonstrates that in the case of riskless debt, the present value of the tax shield is continuously rising. However, when debt is not risk-free, the present value of the tax
10
shield is equal to the levered value minus the unlevered value. The tax shield should be discounted by the weighted average cost of capital after tax: rpost-tax = E+DE * rE + (1-t) * rD * E+DD . All variables are explained before except for t, which is the tax rate. In conclusion, Kraus and Litzenberger (1973) state that the value of a levered firm is equal to the value of an unlevered firm plus the present value of the tax shield:
VL = VU + PV(Tax Shield).
Figure 1. An illustration of the static trade-off theory by Myers (1984).
2.2.2 The Static Trade-off Theory
The trade-off theory of capital structure states that there is an optimal amount of leverage for the firms (Singh and Kumar, 2012). The goal of the firm is to be near this optimal point, or move towards this point. The trade-off theory can be split into two different theories, one static theory and one dynamic theory. Myers (1984) states that the static trade-off theory is an extension of Modigliani and Miller their second proposition (1963). Myers (1984) defines the static theory as ‘’setting a
target debt-to-value ratio and gradually moving towards it, in much the same way that a firm adjusts dividends to move towards a target payout ratio’’ and he formulates the static trade-off theory as
follows.
VL = VU + PV(Tax shield) – PV(Financial distress costs).
11
shield and the costs of financial distress. Figure 1, illustrated by Myers (1984), shows that when debt levels are too high, firms are more prone to going bankrupt and the present value of the financial distress costs is relatively high. On the other hand, if the debt levels are too low, firms do not maximize the benefits of holding debt. Figure 1 shows that there is an optimum level of debt that firms try to reach, but there are still variations in the debt ratios. Myers (1984) claims that this is due to adjustment costs, which makes it harder for firms to move towards the optimum when shocks occur.
2.2.3 The Dynamic Trade-off Theory
The dynamic trade-off theory is an adjustment of the static trade-off theory, in which adjustment costs are accounted for. Adjustment costs, which can be reflected in transaction costs, make it harder for firms to move towards the optimum level of debt (Basu, 2015). Hamada et al. (1984) are one of the first to introduce dynamic models, which is seen as an improvement on previous research regarding firm financing. Frank and Goyal (2007) state that firms look at market expectations when making decisions. More profitable companies have superior investment knowledge and are
therefore likely to be better off. This means that those firms have higher earnings and more equity and that they are able to take on more debt. Ahmad and Etudaiye-Muhtar (2017) provide evidence for the dynamic models. In their paper, the adjustment costs are accounted for by studying multiple periods. Their results show that firms attempt to maximize shareholders’ wealth, with the use of dynamic models. However, their results might be biased due to a small sample size.
2.2.4 The Pecking Order Theory
On the other hand of the trade-off theory, there is the pecking order theory, introduced by Myers and Majluf (1984). In the pecking order theory, there is no optimum level of debt that firms strive towards. The asymmetric information problem, in which the management of a firm has superior knowledge over the investors, is the basis of the pecking order theory. Myers and Majluf (1984) design a model that shows that firms prefer internal financing (to external). Firms prefer internal financing, because of the costs of issuing shares are higher when the asymmetric information is larger. If, however, internal financing is not possible, firms choose to finance through debt before financing through equity. According to Myers and Majluf (1984), issuing debt signals an
undervaluation of the current stock price. Therefore, firms with good information (that is not known to the public) prefer to finance through debt, to prevent a discount of new shares, that would come when issuing new equity.
12
2.2.5 The Market Timing Theory
A more recent theory on capital structure is the market timing theory (Baker and Wurgler, 2002). This theory states that when stock prices are overvalued, firms choose to finance their projects through debt. If the firms are undervalued, they use equity financing. However, research on this topic provides different results. On one hand, authors like Welch (2004) find that firms do little to rebalance their leverage ratios when the stock price fluctuates. On the other hand, Setyawan (2011) shows that the hypothesis of the market timing theory is accepted, using a small sample. If the hypothesis is accepted than this can explain why firms finance certain projects more through debt and some more through equity.
2.3 The effect of M&A’s on Capital Structure
The third part of this literature review describes previous literature on the relationship between M&A’s and capital structure. Harford et al. (2009) investigate the trade-off theory of capital structure with evidence from acquisitions. They study U.S. firms in two ways. First off, how do they finance their acquisitions and secondly how to they adjust the capital structure post acquisition. Their results are consistent with the trade-off model, since firms with low amounts of leverage choose to finance more through debt. Also, after an acquisition, firms reduce their debt holdings largely within a period of five years. They use a measurement of expected capital structure introduced by Kayhan and Titman (2007), who use a partial-regression model.
In a later paper on the same topic, Hovakimian and Li (2011) argue that the method used by Kayhan and Titman (2007) has serious flaws. Chang and Dasgupta (2009), as well as Shyam-Sunder and Myers (1999) previously, show that this partial-regression model is biased, since it gives a positive speed of adjustment, even for firms that do not adjust their capital structure significantly. Therefore, Hovakimian and Li (2011) address the issues introduced by Chang and Dasgupta (2009) and use a different methodology, in order to find how firms finance their acquisitions and how they change their capital structure post-acquisition. The complicated part about this is that the ‘’optimum amount of leverage’’, as explained in the trade-off theory, is hard to measure and a proxy has to be used. This proxy is explained in detail in the methodology, however it is a measurement based on firm characteristics that determine the debt ratio of a firm. Hovakimian and Li (2011) find similar results, compared to Harford et al. (2007). They show that firms have a target amount of leverage and that they try to move towards this point after a merger or acquisition. Thus, their results support the trade-off theory.
Koh et al. (2011) use the same methodology as Hovakimian and Li (2011) when studying the effect of acquisitions on capital structure for Australian firms between 1993 and 2005. They state
13
that when firms are more profitable, they are more likely to finance through debt and they claim that firm characteristics affect the target leverage. Koh et al. (2011) state that their findings of the Australian firms are consistent with the trade-off theory. However, they do imply that firms tend to move away from their optimum level of debt, when they are able to raise more debt.
Khoo et al. (2017) further discuss the paper of Koh et al. (2011) and the Australian firms. Khoo et al. (2017) their goal is to check to validness of the results. Their hypotheses state that firms should move towards their target capital structure post-acquisition and that they should do so faster than firms not incurring this shock (merger or acquisition). They study Australian acquirers for a longer period, namely 1990 up until 2013. In the end, their evidence shows that acquirers face significant changes in their capital structure during the year of acquisition. This makes sense, since acquirers need to finance their acquisitions and this payment changes their capital structure. Also, they find evidence supporting both Koh et al. (2011) and the trade-off theory. Acquirers have to readjust the capital structure after an acquisition, in order to get to the optimum amount of debt. However, all these papers investigating the effects of acquisitions are either on U.S. firms (Harford et al., 2007; Hovakmian and Li, 2011) or Australian firms (Koh et al., 2011; Khoo et al., 2017). This paper contributes to those papers by investigating the European firms and their acquisitions. As mentioned before, there can be differences between European and U.S.’ firms and it is important to find out if there are differences.
3. Methodology and Hypothesis Development
This section describes the methodology used to solve the research question. This thesis also further investigates how firm finance their acquisitions. First off in this section, the methodology that is used to investigate how firms finance their acquisitions is explained. Secondly, the partial adjustment model is explained, which is used to answer the main research question. The partial adjustment model consists of two steps. The first step explains how the target level of leverage is measured and the second step explains how the Speed of Adjustment is calculated. After the methodologies, the hypotheses are formed based off previous research on this topic, as well as previous research on the differences between Europe and The United States.
3.1 Financing of Mergers and Acquisitions
Following papers of Hovakimian and Li (2011) and Koh et al. (2011), research on this topic allows for an estimate of how firms finance their acquisitions. Besides, the determinants for firm financing can shed light on if (and why) firms change their capital structure post acquisition. Over the years, more evidence regarding this topic find significant variables affecting the financing of M&A’s. Following Koh et al. (2011), three different logit regressions are run, that are all formulated as follows.
14
Dt = a + β1M/Bt-1 + β2PROFITt-1 + β3ln(AT)t-1 + β4PPEt-1 + β5DEPt-1 + β6BLt-1 + εt (1) In section 3.2, all the variables are explained. Table A in the appendix provides an overview off all the variables used in this thesis.
The dependent variable in Equation (1) is a dummy variable. This logit regression is run three times and Dt has a different meaning each time. First off all, the difference between full debt (D=1) and other issuers (D=0) is measured. Secondly, the difference between full debt (D=1) and full equity issuers (D=0) is measured and finally, the difference between other (D=1) and full equity issuers (D=0) is measured. In Equation (1), only acquisitions that have a known method of payment are used, therefore there are less M&A’s compared to the amount of M&A’s in the partial adjustment model, explained in Section 3.3 and Section 3.4.
3.2 Methodology for Target Leverage
In order to find the effect that mergers and acquisitions have on capital structure, a partial adjustment model is used. The core idea of this framework is to find the difference between the estimated increase/decrease in capital structure and the actual increase/decrease in capital structure. The difficult part is finding an estimated capital structure, which is an estimate of the target level of debt that firms strive towards, according to the trade-off theory (Myers, 1984). Fama and French (2002) state that firms change their debt ratios in order to reach the optimum level of debt. This theory is in line with the trade-off theory (Myers, 1984). Leary and Roberts (2005) use a dynamic rebalancing model, to estimate if and how firms rebalance their capital structure. Leary and Roberts (2005) investigate how the adjustment costs affect the restructuring of the capital structure. They use a framework that finds the motivation behind financing decisions. The Hazard Function that they use measures the probability that a firm will adjust their capital structure. Using a dynamic duration model is one way to measure whether firms change their capital structure. The results show that firms heavily rebalance their capital structure.
This thesis, however, uses a method similar to that of Hovakimian and Li (2011). As mentioned before, this methodology differs from traditional measurements of the speed of adjustment. This new measurement is introduced by Chang and Dasgupta (2009), who critique the previous literature on measurements of the speed of adjustment. The method of Hovakimian and Li (2011) is a two-stage process that eliminates the look-ahead bias and the mean reversion effect and thus, it provides a valid estimate of the speed of adjustment. The first stage of the regression determines the target level of debt that firms desire. This target leverage ratio is calculated using historical firm characteristics. Hovakimian and Li’s (2011) present a valid methodology that provides an estimate for the target leverage of debt, which is the method that is used in this thesis:
15
BL*i,t+1 = a + β1MTBi,t + β2ln(AT)I,t + β3PROFITi,t + β4PPEi,t + β5DEPi,t + vi + εi,t+1 (2) Following previous research on capital structure, for example Rajan and Zingales (1995), this thesis uses the total debt scaled by the total assets as a measurement of leverage/capital structure. The subscripts i and t represent the firm and the time respectively. The v refers to firm fixed effects and ε is the error term. The definitions of the other independent variables and the reason to include them are as follows.
• MTB is a measurement of growth opportunities and is calculated as total debt plus
shareholders’ equity, divided by total assets. This measurement is used by Hovakimian et al. (2001). Titman and Wessels (1988) claim that firms with a higher market to book ratio have higher book leverage. The market-to-book ratio is also used to test the market timing theory of debt.
• Ln(AT) is the natural logarithm of total assets and is a measurement of the size of the company. Firm size affects leverage, since larger firms are more diversified, they are less prone to bankruptcy and thus, they are able to have higher debt ratios (Warner, 1977). • PROFIT is a measurement of the profitability. Profitability, in this paper, is measured as the
earnings before interest, tax depreciation and amortization (EBITDA) divided by total assets. Firms with higher profitability have superior investment knowledge and are likely to be better off, according to Frank and Goyal (2007). Kayhan and Titman (2007) provide evidence that shows that firms with higher profits invest more and have higher debt ratios.
• PPE is the property, plant and equipment, divided by total assets. This is a measurement of the tangibility of a firm. Frank and Goyal (2009) state that firms with more tangible assets have higher leverage ratios.
• DEP is the depreciation, depletion and amortization over total assets. DEP is a measurement of the interest deductibility effect. As mentioned before, Kraus and Litzenberger (1973) show that the main advantage of holding debt is tax savings.
For further simplicity, the independent variables are named as one. The target level of debt is thus measured as follows, where X is the combination of the independent variables:
BL*i,t+1 = β1Xi,t (3)
3.3 Partial Adjustment Framework
In the second stage of the regression, the difference between the predicted difference in leverage ratios and the actual difference in leverage ratios is estimated. This is done using Equation (4). Following Hovakimian and Li (2011), observations with a book leverage ratio above 80% are
16
excluded. This is done in order to reduce the bias in favor of the trade-off theory. In order to eliminate the look-ahead bias, as mentioned by Chang and Dasgupta (2009), this thesis uses historic data for Equation (2). For each year, the previous year’s characteristics are used to measure the target value of leverage. After that, the speed of adjustment is measured in Equation (4). Equation (2), (3) and (4) combined are the partial adjustment model of this paper.
BLi,t+1 – BLi,t = a + λ1(BL*i,t+1 - BLi,t) + εi,t+1 (4)
Equation (4) finds the measurement for the speed of adjustment (SOA), λ1, Following the research of Hovakimian and Li (2011), this is the correct measurement, since the measurements used in other research provide inaccurate results due to mean reversion. From now on, λ1 is referred to as the ‘’Speed of Adjustment’’.
The partial adjustment framework has a lot in common with an event study. The framework allows for an unbiased test for a causal effect of the M&A’s on the capital structure. It eliminates possible endogeneity issues, that can arise when investigating this topic. This partial adjustment framework should, therefore, give more accurate results.
3.4 Hypothesis forming
This paper has one main hypothesis on the research question ‘’What is the effect of Mergers & Acquisitions on capital structure?’’. However, there is also a hypothesis for the logit models on firm financing. This hypothesis is explained first and the main hypothesis can be found at the bottom of this section.
Three different logit models are run in order to investigate the determinants of firm financing. These tests test the pecking order theory (Myers and Majluf, 1984), the market timing theory (Baker and Wurgler, 2002) and the trade-off theory (Myers, 1984).
H1.0: The independent variables do not affect the method of payment (βn = 0). H1.1: Some (or all) independent variables affect the method of payment (βn ≠ 0).
According to the pecking order theory, firms prefer internal financing over external and debt financing over equity (Myers and Majluf, 1984). Thus, firms with a higher profitability have lower debt. Also, firms with a higher book leverage before the payment are more likely to finance their acquisition with more debt, since they prefer debt financing over equity. Following the pecking order theory, it is expected that the effect of profitability before the payment (PROFITPRE-M&A) is negative and a higher book leverage ratio (BLPRE-M&A) leads to more debt financing.
17
larger have higher debt ratios3. Thus, it is expected that firms with high ratios for these variables prefer debt financing over the others methods. In the case of either mixed financing or equity financing, the expectation is that firms with high ratios for these variables prefer equity financing. The other theory that is tested is the market timing theory (Baker and Wurgler, 2002). This theory is tested with the book variable. The market timing theory states that a higher market-to-book ratio pre-M&A leads to an increase in equity financing. Since the expectation is that firms face some form of the market timing theory, it is expected that a higher market-to-book ratio leads to more equity financing, compared to mixed financing.
The final hypothesis regards the main study of this paper, the partial adjustment model. This model looks at how firms change their capital structure around an acquisition. The hypothesis that is tested is whether or not firms have a target level of debt. In the model, λ1 is the speed of
adjustment, which tests if (and ho w) firms change their capital structure surrounding an acquisition. H2.0: There is no target level of debt that firms strive towards (λ1 = 0).
H2.1: There is a target level of debt that firms strive towards (λ1 > 0) or strive away from (λ1 < 0). The expectation is that firms do have a target level of debt and that the null hypothesis will be rejected. This is because of the trade-off theory of debt (Myers, 1984), in which firms face certain costs and benefits of holding debt and choose their capital structure following these costs. The speed of adjustment is expected to be largest in the case of debt acquirers, since these firms face bigger shocks to their capital structure. In order to get near the optimum level of debt, they need to adjust their capital structure more than firms that use either equity or a mixed method of payment to finance their acquisition. It is also expected that the speed of adjustment is largest in the year of the merger or acquisition, since this is the year that firms face the biggest shock to their capital structure, compared to the years surrounding the M&A. Previous research from Koh et al. (2011) shows that Australian firms have a target level of debt. This paper seeks to find whether or not this is the case for European firms.
4. Data and Descriptive Statistics
This section provides an overview of the data and the descriptive statistics. In the first part, the source of the data and the sample period are defined and the correlation matrix is presented. The second part presents the descriptive statistics of the data. The descriptive statistics are presented in
3 As mentioned before, Warner (1977) states that firm size positively affects debt ratios. Frank and Goyal
(2009) state that there is a positive effect between property, plant, equipment and the debt ratio of a firm, whereas Kraus and Litzenberger (1973) show a positive relationship between depreciation, depletion, amortization and total assets.
18
two ways; one for the entire sample and one that separates the acquiring firms by their method of payment.
4.1 Data
In order to answer the research question, data is extracted from two different databases. First off all, data on European mergers and acquisitions is needed. The Euro-area is chosen, because of two reasons. First off all, data on European firms allow for more data on private firms and secondly, investigating European M&A’s can provide new findings in the research on the relationship between M&A’s and capital structure. In order to obtain this data (on European firms), the Zephyr database is used. Also, firm characteristics for the acquiring firms are needed. These are extracted from the OSIRIS database.
The sample period is 2000-2013. This provides the most recent results (since acquirers after 2013 do not have available data for the three year period after an M&A) and this gives an overview of the current situation. In order to find the speed of adjustment, a period of three years before up until three years after the acquisition is used, following Khoo et al. (2017). This way, it is shown how the speed of adjustment is different in the year of an M&A. Data on firm characteristics ranges from 1997-2013. Firms who have either negative total assets, or non-reported total assets are excluded from the sample.
For the mergers and acquisition deals, certain criteria are used to determine which deals to include, besides from the time period (2000-2013) and the region (Euro-area) mentioned above:
1. The M&A deal has to be completed and confirmed. Deals that are rumored to be completed or deals that are announced do not provide the effect that this paper attempts to
investigate. The data necessary is only on the acquirers’ change in capital structure before and post-acquisition.
2. There must be sufficient information available on the acquirer firm’s characteristics (Debt, equity, etc.)
3. The M&A deal has to have a value of at least one million Euro’s.
4. Only the first M&A deal for the acquirer each year is included. Some acquirers make multiple acquisitions in one year, but only their first one is included.
5. If the acquiring firm is from the financial sector (SIC between 6000-6999), the deal is excluded. Financial firms include firms such as banks, real estate firms, property trusts etc. These firms are likely to have very different capital structures (Koh et al, 2011), compared to the other firms. According to Khoo et al. (2017), these firms face greater supervision and behave different.
19
6. Acquiring firms with a book leverage (BL) greater than 0.8 are excluded. Chang and Dasgupta (2009) introduce this, in order to eliminate the mean reversion problem. This should lead to less biased results.
7. According to Baker and Wurgler (2002), who do research on capital structure, firms with high market-to-book ratios should be excluded (higher than 10). They also exclude firms whose profitability is either significantly high (greater than one), or significantly low (less than minus one). This paper follows their methodology.
After merging the database of the M&A’s with the database of the firm characteristics, more than 2000 mergers and acquisitions are present in the findings. After eliminating firms their second, third or even fourth acquisition in one year, 1476 acquisitions are reported. Following points 6 and 7 of the previous criteria, four observations are deleted because of a book leverage above 0.8 and twelve are dropped because of either a significantly high or significantly low profitability. This leaves 1460 observations, which is enough to provide accurate results for the regressions. In comparison, Khoo et al. (2017) have 936 observations.
Table 1. Correlation Matrix of The Variables
This table reports the correlation between the variables (dependent and independent). The variables can be found in Section 3.2 and are defined as follows. BL is the total debt scaled by the total assets and is a
measurement of the book leverage. Ln(AT) is the natural logarithm of total assets and measures the size of the firm. MTB (Market to book ratio) is total debt plus shareholders’ equity, scaled by total assets. PROFIT is the profitability of a firm and is measured by dividing the earnings before interest, tax depreciation and
amortization by total assets. PPE is the property plant and equipment scaled by total assets. DEP is the
depreciation, depletion and amortization scaled by total assets. The coefficients are marked with *, ** or *** if they are significantly different from zero at the 10%, 5% or 1% respectively.
BL Ln(AT) MTB PROFIT PPE DEP
BL 1 Ln(AT) 0.2804*** 1 MTB 0.6750*** 0.0269 1 PROFIT 0.0274 0.1006*** 0.0062 1 PPE 0.4097*** 0.2743*** 0.3087*** 0.1502*** 1 DEP 0.0148 0.0682*** -0.0188** 0.0040 -0.0745*** 1
Table 1 provides an overview of the correlation mix of the variables. This shows that the relationship between book leverage and all independent variables is positive. For the variables Ln(AT), MTB and
20
significant. These findings are in line with previous research on the relationship between book leverage and these variables4. In the case of multicollinearity, the correlation between two variables is higher than 0.8. Table 1 shows that there is no multicollinearity in this paper, since the highest value is 0.6750.
4.2 Descriptive Statistics
Table 2A and Table 2B present the descriptive statistics of this research. Table 2A presents the statistics for the entire sample of acquirers. These statistics show that, on average, firms that make an acquisition hold 13.58% debt, compared to total assets. The same table reports a coefficient of 13.8288 for the variable Ln(AT), which represents the size of the firm. This converts to an average amount of ± 1015 million total assets. Acquirers’ total assets range from ± 3.9million to ± 14.41 trillion. The market to book ratio (MTB) shows that, for acquiring firms, the ratio between debt plus shareholders’ equity and total assets is 52.84% The average ratio of EBITDA divided by total assets
Table 2A. Descriptive Statistics of the Full Sample
This table reports the descriptive statistics for the variables, covering the entire sample of acquirers (1460 firms). The variables can be found in Section 3.2 and are defined as follows. BL is the total debt scaled by the total assets and is a measurement of the book leverage. Ln(AT) is the natural logarithm of total assets and measures the size of the firm. MTB (Market to book ratio) is total debt plus shareholders’ equity, scaled by total assets. PROFIT is the profitability of a firm and is measured by dividing the earnings before interest, tax depreciation and amortization by total assets. PPE is the property plant and equipment scaled by total assets.
DEP is the depreciation, depletion and amortization scaled by total assets. This table reports the amount of
observations (N), the mean, the median, the standard deviation (Std. Dev.), the minimum (Min.) and the maximum (Max.) of the observations.
N Mean Median Std. Dev. Min. Max.
BL 1460 0.1358 0.1637 0.2536 -0.9926 0.7677 Ln(AT) 1460 13.8288 13.7467 2.2677 8.2738 23.3912 MTB 1460 0.5284 0.5599 0.2110 -0.8872 0.9983 PROFIT 1460 0.1042 0.1062 0.0961 -0.7729 0.7561 PPE 1460 0.2223 0.1712 0.1912 0 0.9723 DEP 1460 -0.0485 -0.0383 0.0842 -2.4506 0
4 There has been research on each variable and its effect on the book leverage. Titman and Wessels (1988)
claim that higher growth opportunities lead to higher leverage ratios. Warner (1977) claims the same relation for size. Kayhan and Titman (2007) provide evidence for a positive relation between profitability and debt ratios. Frank and Goyal (2009) state that firms with higher tangibility have higher leverage ratios. The DEP variable measures the interest deductibility effect and Kraus and Litzenberger (1973) show that the main benefit of a firm is tax holding. Therefore, this positive relation between depreciation and book leverage is expected.
21
equals 10.42%. PPE measures the tangibility of a firm and this coefficient shows that the Property, plant and equipment scaled with total assets is 22.23% on average. The final measurement, on the depreciation scaled to assets is the only negative coefficient. This coefficient shows a -4.85% ratio between the depreciation, amortization and depletion and the total assets.
Table 2B. Descriptive Statistics separated by Method of Payment
This table reports the descriptive statistics, separated by method of payment. The variables can be found in Section 3.2 and are defined as follows. BL is the total debt scaled by the total assets and is a measurement of the book leverage. Ln(AT) is the natural logarithm of total assets and measures the size of the firm. MTB (Market to book ratio) is total debt plus shareholders’ equity, scaled by total assets. PROFIT is the profitability of a firm and is measured by dividing the earnings before interest, tax depreciation and amortization by total assets. PPE is the property plant and equipment scaled by total assets. DEP is the depreciation, depletion and amortization scaled by total assets. This table separates the results based on the method of payment. The first Column represents firms financing completely through equity, the second one completely through debt and the third one represents the other forms of payment (mixed etc.). This table reports the amount of firms that choose a certain method of payment (N), the mean of the variables and the standard deviation of the variables.
Full Equity Full Debt Other Forms
N 180 (12.33%) 62 (4.25%) 1218 (83.42%) BL Mean Std. Dev 0.0642 (0.3183) 0.2257 (0.1658) 0.1418 (0.2442) Ln(AT) Mean Std. Dev 12.9362 (2.2972) 14.4397 (1.7838) 13.9296 (2.2544) MTB Mean Std. Dev 0.5063 (0.2498) 0.5765 (0.1696) 0.5292 (0.2063) PROFIT Mean Std. Dev 0.0638 (0.1655) 0.1098 (0.0642) 0.1099 (0.0810) PPE Mean Std. Dev 0.2105 (0.2148) 0.2321 (0.1620) 0.2236 (0.1889) DEP Mean Std. Dev -0.0725 (0.2141) -0.0380 (0.0249) -0.0455 (0.0403)
Table 2B presents the descriptive statistics for the acquiring firms, separated by the way they finance their acquisitions. Distinctions are made between firms financing their acquisitions completely through equity (Column 1), firms financing their acquisition completely through debt (Column 2) and firms using a mixture of payments (Column 3). It has to be noted that the amount of firms is less
22
than the amount in the full sample, since for some acquisitions the method of payment is not known/not reported. The first thing that jumps out in Table 2B is the fact that most of the reported acquisitions are financed through mixed payments of debt and equity and others. More than 80% of the sample financed the acquisitions with a mixed payment. Of the rest, only 12.33% are financed completely through equity and even less (4.25%) are financed completely through debt.
When looking at the descriptive statistics for the separate acquisitions, it shows that book leverage is highest when firms choose to finance completely through debt. This makes sense, since firms with higher debt levels are more likely to finance their acquisitions through debt and when they are low, firms are more likely to finance through equity. Firms who finance their acquisition with equity have, on average, a leverage ratio that is 16.15% (difference between the coefficients 0.2257 and 0.0642) lower than firms who finance completely through debt. As for the other variables, firms who finance completely through debt are reported to be more profitable, which states that firms with higher debt ratios might be more profitable. Also, these types of firms are found to be larger in size, have higher market-to-book ratios, higher property, plant and equipment and higher depreciation, amortization and depletion compared to acquirers who finance through equity.
Firms who choose to finance using a mixed method of payment are, on average, the ones with the highest profitability. For the other variables (BL, Ln(AT), MTB, PPE and DEP), the values of firms using a mixed method of payment lay in between the values of the firms financing completely through debt and the firms financing completely through equity. The firms that choose to finance completely through debt have the highest values for these variables, whereas the firms financing completely through equity have the lowest. This signals that equity acquirers, on average, are the smallest firms with the lowest book leverage, lowest market-to-book ratio, lowest PPE and the lowest DEP.
5. Results
This section reports the results from the research. There are two different sections, since this thesis has two different tests. The first test looks at what firm characteristics play a role in the
determination of firm financing. The second test determines if/how firms change their capital structure surrounding the time of an acquisition.
5.1 Financing of M&A’s
Table 3 shows the results for the first set of tests, regarding the way firms choose to finance their mergers and acquisitions. In this set, three logit regressions are run, with a different dummy variable as the dependent variable each time. In Column 1, the dummy variable is equal to 1 if the firm
23
finances through debt and 0 if the firm finances through a mixed form of payment. For Column 2, the dummy variable equals 1 if the acquiring firm finances through debt and 0 if it finances through equity. In the 3rd Column, the dummy variable is equal to 1 if the acquirer finances with a mixed form of payment and 0 if the firm finances through equity. The independent variables are all mentioned before: Book leverage, Size, Market-To-Book ratio, Profitability, Property, plant and equipment (PPE) and Depreciation, amortization and depletion (DEP). In this case, the independent variables that are used are those before making the merger/acquisition. This way the effect of the independent variables on the method of payment is investigated. Appendix B, C and D show the complete results for the logit models, whereas Table 3 provides the most important statistics (the coefficient, standard error and the significance level), as well as the Chi² of the logit regressions.
Column 1 of Table 3 shows that firms with a higher book leverage ratio before the M&A are more likely to finance their payment through debt. This coefficient (3.1557) is significant at the 1% level. This follows the pecking order theory (Myers and Majluf, 1984) that states that firms prefer to finance their acquisitions with debt. Since they have a lot of debt before the issue (high leverage ratio), they are capable of using debt and they will use debt. Column 1 also shows that an increase in the market-to-book ratio leads to more mixed financing (coefficient of -2.1771, significant at 1%). This means that firms with a higher market-to-book ratio pre-issue rather use a mixed method of financing, instead of using only debt.
This Column also shows that the coefficient of size (-0.1455) is significant at the 1% level. This means that when firms are larger, they are more likely to use a mixed form of payment, instead of using debt. The final significant coefficient shows that firms with higher depreciation, depletion and amortization (5% significance) are more likely to finance their acquisitions through debt. Most of these results in Column 1 are in line with the expectations in Section 3.4, where it is stated that firms with higher depreciation and a higher book leverage use debt as their main method of payment. The significant negative effect that size has on debt financing is in contrast with the expectations in Section 3.4, since it is expected that larger firms have more debt in general and thus are more likely to finance through debt. The coefficient of profitability (1.7900) and PPE (0.5103) are not significant. It can not be said with certainty that a higher profitability and/or PPE leads more debt financing.
The second Column of Table 3 has only one insignificant coefficient, the property, plant and equipment, scaled with total assets (-0.8139). Nothing can be said with certainty about this variable. It must be said that in Column 2, only 242 observations are included, 62 debt payments and 180 equity payments. This small sample can lead to biased results.
Column 2 shows, once again, that a higher book leverage before the acquisition leads to more debt financing. This is line with the expectation in Section 3.4. Also, a higher market-to-book
24
ratio leads to a lower amount of debt financing. Both of these results are significant at the 1% level. The coefficient of DEP (10.4249) is once again positive, however this time only at the 10% level. This shows again that firms with higher depreciation prefer debt financing (over equity in this case). Coefficients in the logit model can be interpreted as follows. For DEP (10.4249), a one unit increase in DEP leads to a 10.4249 increase in the log-odds of the dependent variable, thus stating that a higher depreciation leads to more debt financing.
In contrast with the first column, Column 2 shows that larger firms do prefer more debt financing, compared to equity financing (significant at 1%). This is the results that is expected, since firms that are greater in size can take on more debt and prefer debt over equity, according to the pecking order theory (Myers and Majluf, 1984). The positive coefficient of profitability (2.8768) is significant at the 10% level and shows that more profitable firms prefer to finance their acquisitions via debt.
The third Column in Table 3 shows how the firm characteristics determine whether firms finance through equity or through a mixed method of payment. This Column shows that firms with higher leverage ratios pre-issue choose to finance their acquisition with a mixed method of
payment. This effect is shown by the positive coefficient of 0.9854, which is significant at the 1% level. The coefficient of 0.1435, regarding firm size (Ln(AT)) shows that larger firms prefer a mixed method of payment. This is significant at the 1% level and is in line with the coefficient of -0.1455 in Column 1, which also shows that larger firms prefer to finance with a mixed method of payment.
The positive coefficient for profitability (3.4072) and the negative coefficient for PPE (-0.9693) show that more profitable firms with less PPE are more likely to finance their acquisitions with a mixed method of payment. These two coefficients are significant at the 1% and 5% level respectively. The last two variables, MTB and DEP are not significant and therefore do not provide further insight in the determinants of the method of financing.
In conclusion, the BL, PROFIT and DEP variables are positive in Column 1, 2 and 3. This shows that firms with high leverage ratios, profitable firms and firms with higher depreciation, amortization and depletion prefer to finance their acquisition through debt. If they have to choose between mixed and equity payments, they choose mixed payments. These results are in line with the pecking order theory, that states that firms prefer debt over equity financing (Myers and Majluf, 1984). Where the book leverage variable is significant in all three columns, the PROFIT and DEP variable are only significant in two of the three regressions. The MTB variable shows that firms with a higher market-to-book ratio pre-issue prefer both mixed or equity financing over debt financing. The Ln(AT) variable shows that firms that are greater in size prefer equity financing over debt and mixed
25
payment. The last variable, the PPE is only significant in Column 3, where it shows that a higher PPE
leads to more equity financing.
Table 3. Results from Logit Models
This table reports the results for the first set of tests. Multiple logit regressions are run in order to find out what determines firm financing. This logit model has a dummy variable as the dependent variable and is formulated as follows.
Dt = a + β1M/Bt-1 + β2PROFITt-1 + β3ln(AT)t-1 + β4PPEt-1 + β5DEPt-1 + β6BLt-1 + εt.
In Column 1, the dummy variable is 1 if the firm finances through debt and 0 if the firm uses a mixed method of payment. In Column 2, the dummy variable is 1 if the firm finances through debt and 0 if the firm finances through equity. In Column 3, the dummy variable is 1 if the firm uses a mixed method of payment and 0 if the firm uses equity financing. The independent variables can be found in Section 3.2 and are defined as follows.
BL is the total debt scaled by the total assets and is a measurement of the book leverage. Ln(AT) is the natural
logarithm of total assets and measures the size of the firm. MTB (Market to book ratio) is total debt plus shareholders’ equity, scaled by total assets. PROFIT is the profitability of a firm and is measured by dividing the earnings before interest, tax depreciation and amortization by total assets. PPE is the property plant and equipment scaled by total assets. DEP is the depreciation, depletion and amortization scaled by total assets. For all variables, the value before the merger/acquisition is used. This table provides the coefficient of the variables, as well as the standard error in between the parenthesis. The coefficients are marked with *, ** or *** if they are significantly different from zero at the 10%, 5% or 1% level respectively.
Debt (1) vs. Mixed (0) payments Debt (1) vs. Equity (0) payments Mixed (1) vs. Equity (0) payments BLPRE-M&A 3.1557*** (0.8232) 4.1608*** (1.0007) 0.9854*** (0.3839) Ln(AT)PRE-M&A -0.1455*** (0.0281) 0.2509*** (0.0822) 0.1435*** (0.0182) MTBPRE-M&A -2.1771*** (0.7875) -2.3696*** (0.9087) -0.2510 (0.4422) PROFITPRE-M&A 1.7900 (1.9769) 2.8768* (1.4974) 3.4072*** (0.7696) PPEPRE-M&A 0.0826 (0.7600) -0.8139 (0.9123) -0.9693** (0.4820) DEPPRE-M&A 13.4589** (5.6411) 10.4249* (5.6127) 1.9014 (1.1695) Chi² 483.58*** 47.94*** 549.37*** N 1280 242 1398 Dummy = 1 62 62 180 Dummy = 0 1218 180 1218
26
According to Hovakimian et al. (2004), if the market timing theory holds, Column 3 should provide a significant coefficient for the MTB variable. Since this coefficient is insignificant, the market timing theory does not hold. Hovakimian et al. (2004) also state that, in case of the trade-off theory, the market-to-book ratio and profitability are negative and positive respectively, when comparing debt payments with mixed payments. Firms that face good market performance (High MTB) are more likely to finance their acquisitions with a mixed payment to reduce their leverage ratios. The trade-off theory is shown to hold, since the MTB variable is negative and the BL variable is positive in Column 1.
The most important variable for the next set of tests is that of the book leverage ratio. This coefficient is positive and significant in all three columns at the 1% level. These findings are the same of that of Koh et al. (2011), who do research on Australian firms. The significant coefficients show that firms have a target level of optimum debt (trade-off theory) that they strive towards. From this, it is expected that, in the partial adjustment model, there is a significant coefficient, showing that firms change their capital structure around the acquisition.
5.2. Results from the Partial Adjustment Model
Table 4 shows the speed of adjustment of the capital structure between the years -3 (three years before the acquisition) and +3 (three years after the acquisition). The Speed of Adjustment is calculated as the average λ1 from Regression (4). The main finding is that, in the year of the acquisition, the speed of adjustment is much higher than the years surrounding the acquisition. In year 0, the acquisition year, the speed of adjustment is 0.3519 and significant at the 1% level. This implies that acquisitions do have an effect on the capital structure. This coefficient shows that acquirers adjust 35.19% of their leverage ratio, in order to close the gap to the target amount of leverage. This result is in line with the expectation of the (dynamic) trade-off theory of capital structure (Myers, 1984), which expects that firms strive towards an optimal level of debt. Shocks to the capital structure, coming from an M&A in this case, means that firms have to adjust their capital structures. The findings in Table 4 are in line with the expectation in Section 3.4, stating that firms do rebalance their capital structure around the time of an acquisition. This means that the second null hypothesis is rejected since λ1 is significantly different from zero.
Even though the speed of adjustment is largest in year zero, it is also positive and significant in all other years. Besides from the years -2 and +3 (significant at the 5% level), the coefficients are significant at the 1% level. This shows that firms are always adjusting their capital structure, in order to reach the target amount of leverage. However, as mentioned before, the coefficient is largest in the year of the acquisition. The coefficients shows that firms close the gap between the target level of debt and the actual amount of debt each year (for example, 2.67% in year -1 and 1.02% in year
27
+2). The fact that the coefficient is positive signals that acquirers move towards the target level of debt, whereas a negative coefficient would mean that firms move away from the target level of debt. However, these results only show how all firms change their capital structure on average. Since this thesis also looks at the method of payment and in order to compare the results to previous research (Koh et al., 2011; Khoo et al., 2017) it is also important to look at how different types of acquirers change their capital structure. Different types of acquirers such as acquirers who finance the acquisition through equity, debt or through some form of mixed payment, all three of which are described in previous sections.
Table 4. Adjustment of Capital Structure
This table reports the results of the partial adjustment model, which studies the effect that mergers and
acquisitions have on the acquirer’s capital structure. Speed of Adjustment (SOA) is calculated as λ1 in the
following formula.
BLi,t+1 – BLi,t = a + λ1(BL*i,t+1 - BLi,t)+ εi,t+1
The dependent variable is the actual leverage ratio (Total debt scaled by total assets) in year t+1 minus the actual leverage ratio in year t. The independent variable is the expected leverage ratio in year t+1 (as
calculated before) minus the actual leverage ratio in year t. Besides the SOA, the Standard Deviation is shown in the parenthesis. Also, the t-statistic and the R-squared are given for each regression. Results range from three years before the acquisition (-3) up until three years after the acquisition (+3). Coefficients of the Speed of Adjustment are marked with *, ** or *** if they are significantly different from zero at the 10%, 5% or 1% level respectively.
Year relative to Acquisition Speed of Adjustment (Std. Dev.) T-Statistic R-Squared -3 0.0444*** (0.0083) 5.31 0.0227 -2 0.0154** (0.0066) 2.34 0.0042 -1 0.0267*** (0.0056) 4.74 0.0162 0 0.3519*** (0.0195) 18.02 0.1860 +1 0.0350*** (0.0038) 9.18 0.0563 +2 0.0102*** (0.0031) 3.32 0.0080 +3 0.0071** (0.0029) 2.46 0.0048
28
Table 5 shows the final results of this paper, in which the acquirers are separated by method of payment. This shows how certain types of acquirers (either full equity financed, full debt financed or mixed financed) rebalance their capital structure, surrounding the time of an M&A. This table presents the coefficient of the speed of adjustment (λ1), the T-statistic and the R-squared of the regression. Also, coefficients are marked with *, ** or *** if they are significant at the 10%, 5% or 1% level respectively.
The results in Table 5 show, once again, that for each acquirer the speed of adjustment is largest in the year of the acquisition, stating that firms choose to more firmly rebalance their capital structures after an acquisition. The coefficient in year 0 is highest for firms using a mixed method of payment (0.3653). This shows that these types of firms rebalance 36.53% of their capital structure, in order to close the gap to the target level of capital structure. The coefficient of mixed financing, as well as the coefficient of debt financing (0.3235) are significant at the 1% level, whereas the
coefficient of equity financing is not significant (0.1323). These coefficients show that debt issuers rebalance their capital structure by 32.35% on average and equity issuers rebalance by 13.23% on average. Since the coefficient of equity issuers is insignificant, the adjustment that they make is not certain. The difference between debt and equity issuers is very large, stating that debt financing leads to more rebalancing, compared to equity financing. This is in line with the expectations, stating that firms that use debt as their only source of financing need to readjust their capital structure more. This is because these types of firms face bigger shocks to their capital structure (compared to either equity issuers or mixed issuers) and thus need more rebalancing to get to the optimum amount of leverage. Besides this, all acquirers move, at different speeds, towards the target level of debt. The reason for this is that all coefficients are positive, whereas a negative coefficient would indicate that firms move away from the target level of debt.
From the years surrounding the M&A (-3 up until +3), Table 5 shows that at the extreme points (-3, -2, +2 and +3), the difference between the three methods of payment is very slim. Small differences arise in year -1 and +1, but the big difference, as explained before, occurs in the year of the acquisition. One year before the merger/acquisition, equity issuers are already restructuring their debt by 6.56%, compared to 2.8% and 2.55% for debt and mixed issuers respectively. This coefficient (0.0656) is significant at the 5% level. For debt and mixed issuers, this year does not show such an increase in the rebalancing of the capital structure. Acquiring firms that use debt or a mixed form of payment only really start rebalancing their capital structure in year 0, whereas firms using an equity payment look to start a year in advance. Only the coefficients for equity issuers and mixed issuers are significant at the 5% and 1% level respectively. In the year after the acquisition (+1), the opposite effect is shown. Acquirers using only equity adjust their capital structure by only 0.8%,
29
Table 5. Speed of Adjustment by Method of Payment
This table reports the results of the partial adjustment model, which studies the effect that mergers and acquisitions have on the acquirer’s capital structure. This table is separated by the method of payment. Column 1 shows acquirers that finance the acquisition through equity, Column 2 shows acquirers that financed
through debt and the 3rd Column shows acquirers that financed through a mixed method of payment. The
speed of adjustment (SOA) is calculated as λ1 in the following formula.
BLi,t+1 – BLi,t = a + λ1(BL*i,t+1 - BLi,t)+ εi,t+1
The dependent variable is the difference between the actual leverage ratio (Total debt scaled by total assets) in year t+1 minus the actual leverage ratio in year t. The independent variable is the expected leverage ratio (as calculated before) in year t+1 minus the actual leverage ratio in year t. Besides the SOA, the T-Statistic and the R-Squared are shown. The results range from three years before the acquisition (-3) up until three years after the acquisition (+3). Coefficients of the Speed of Adjustment are marked with *, ** or *** if they are significantly different from zero at the 10%, 5% or 1% level respectively.
Method of Payment
Year relative to Acquisition
Equity Debt Mixed
-3 Speed of Adjustment T-Statistic R-Squared 0.0521* 1.80 0.0548 0.0442 1.34 0.0135 0.0472*** 5.38 0.0275 -2 Speed of Adjustment T-Statistic R-Squared 0.0255 1.00 0.0169 0.0150 0.63 0.0027 0.0168** 2.39 0.0052 -1 Speed of Adjustment T-Statistic R-Squared 0.0656** 2.48 0.0945 0.0280 1.38 0.0121 0.0255*** 4.23 0.0153 0 Speed of Adjustment T-Statistic R-Squared 0.1323 1.18 0.0230 0.3235*** 5.93 0.1699 0.3653*** 16.98 0.1955 1 Speed of Adjustment T-Statistic R-Squared 0.0080 0.33 0.0019 0.0545*** 3.38 0.0633 0.0309*** 8.23 0.0542 2 Speed of Adjustment T-Statistic R-Squared 0.0179 0.77 0.0102 0.0169 1.23 0.0096 0.0088*** 3.00 0.0078 3 Speed of Adjustment T-Statistic R-Squared 0.0172 1.14 0.0250 0.0200** 2.14 0.0328 0.0051 1.62 0.0024
