M&A announcements by small and large firms during the crisis : evidence from the US

(1)

M&A announcements by small and large firms during the crisis: Evidence from the US

Abstract

This study investigates the announcement effect of M&As on the bidding firms’ stock returns. I find that US bidding firms experience a significant positive abnormal return of 0.45%, 0.44% and 0.49% around the announcement of M&As for three event windows. The announcement effect is positive during the crisis. Secondly, the size-effect on bidder returns of M&As is investigated during the crisis, to see whether there is a difference between small and large bidding firms. Results suggest that if firm size increases with 1%, the cumulative abnormal return of the announcement of M&As increases with 0.1%. However, these results are inconsistent measuring the size-effect with two other measures. The conclusion is that smaller firms may use M&As in the crisis to reduce the likelihood of bankruptcy leading to a less positive announcement effect than larger firms.

Author: D. Huijzer

Student number: 11063319

Thesis supervisor: Shivesh Changoer Finish date: June 2018

Credits: 12 EC

University: University of Amsterdam Bachelor: Economie & Bedrijfskunde Specialization: Finance & Organization

(2)

This document is written by David Huijzer who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document are 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)

I. Introduction

In recent years, mergers and acquisitions (hereinafter: “M&As”) have become an increasingly popular method for companies to diversify and acquire synergies (Thijssen 2008). In 2017, companies announced over 50.600 transactions with a total value of more than 3.5 trillion USD, according to the Institute of Mergers, Acquisitions, and Alliances (IMAA). In comparison, twenty years ago (1997) companies announced, according to the IMAA, 26.500 transactions with a total value of 1.8 trillion USD.

This increase in popularity triggered researchers to do research into the value creation of these M&A deals. Dodd & Ruback (1977), Rani et al. (2012) and Masulis et al. (2012) suggest that bidding firms experience significant positive returns around the announcement of M&As. On the other hand, Franks et. al (1991), Moeller et al. (2005) and Betton et al. (2009) conclude that bidding firms experience negative returns around the announcement of M&As.

In this study, I examine whether M&As create more value for smaller US bidding firms during the financial crisis by using an event study. The market reaction on the announcement of M&As is studied by calculating the cumulative average abnormal return (hereinafter: “CAAR”) over the event window of one day before and one day after the announcement date, three days before and three days after the announcement date and five days before and five days after the announcement date. I assume that M&As create value if the announcement effect is positive. The size-effect is studied by an ordinary least squares method. I use a sample of transactions taking place in the United States in 2005-2010 to study the bidding firms’ stock price market reaction to the announcements of M&As for small and large firms.

I find that the shareholders of US bidding firms earn a 0.45% abnormal return in the 10 days around the M&A announcement, suggesting that the announcement of M&As create value for bidding firms. During the crisis, I find that the bidding firms’ shareholders earn a 0.50% abnormal return in the 10 days around the M&A announcement, suggesting that the announcement of M&As create value for bidding firms during the crisis, consistent with Beltratti & Paladino (2013). Results suggest that if firm size increases with 1% during the crisis, the cumulative abnormal return of the announcement of a M&A increases with 0.1%. The conclusion is that the announcement effect of smaller firms is less positive during the crisis than the announcement effect of larger firms because small firms use M&As to diversify their bankruptcy risk.

My study builds on the paper of Beltratti and Paladino (2013). They examine the announcement effect of M&As in Europe during the crisis. They find that the crisis had a

(4)

positive significant impact on the value creation of M&As for bidding firms and conclude that the main cause for this is that stronger bidding firms have the opportunity to acquire weaker firms for distressed prices. My study is different from Beltratti and Paladino (2013) by not only focusing on the banking sector, but the market as a whole. Secondly, I use a more recent sample. Thirdly, and most importantly, I examine whether the effect of the crisis on M&As is different for large and small firms. Empirical findings show that the crisis had more severe impacts for small firms than for large firms (Chowdhury 2011). Business credit became scarce, leading to less M&As by smaller firms (Chowdhury 2011). The size-effect of firms participating in M&As during the crisis is not investigated before. The goal of this thesis is to explore whether the announcement effect of small bidding firms during the crisis is more positive than the announcement effect of large bidding firms.

The remainder of this paper is organized as follows. Section two discusses some relevant papers regarding M&As. In the third section, the hypotheses are derived from the literature. In the fourth section, the research design is presented and the sample selection is described. The fifth section presents the results of this study. The sixth section presents a sensitivity analysis. Afterward, in section six, the conclusion is provided and limitations are given.

II. Literature review

A. Meaning and motives of M&As

M&As means mergers and acquisitions. An acquisition is an “Absorption of one firm by another” (Ross 1998). A merger is a “Combination of two or more firms in which only one of the firms survives” (Hampton 1989). By combining these two definitions, M&As refer to as the combination of at least two businesses into one business.

Firms participate in M&As because both the bidding and the target firm can finance, aid or help each other out in their industry or between industries without spending a lot of capital and time in creating a new firm (Sherman and Hart 2006).

Trautwein (1990) discusses several theories of merger motives that can be classified into different groups. According to Trautwein (1990), the first motive for bidding firms to acquire target firms is to create synergies. There are three types of synergies: Financial synergies are achieved by lowering the systematic risk scale, investing in unrelated businesses, increasing companies size or establishing an internal capital market (Trautwein 1990). Operational synergies are realized by combining operations, transfers of knowledge or economies of scale

(5)

(Chatterjee 1986). Managerial synergies arise from managers abilities, such as superior monitoring and planning abilities, that benefit the target firm performance (Trautwein 1990).

The second reason why bidding firms engage in M&As is to increase market power (Kim & Singal 1993). By achieving more market power, firms can embark on the following activities. Firms can cross-subsidize their products, firms can focus on simultaneously limiting competition in more than one market, and firms can focus on driving away potential entrants in their markets (Trautwein 1990).

The third reason why bidding firms engage in M&As comes from the valuation theory (Steiner 1975). This theory argues that firms participate in M&As because managers of the bidding firms have better information about the value of the target firm than the market has (Ravenscraft & Scherer 1987). This occurs when managers have unique information about possible advantages that arise from combining the businesses of the bidding and target firm (Trautwein 1990).

There are two reasons why managers engage in M&As. The first reason, according to Avery et al. (1998), is building an empire with the aim of getting a higher salary, because firm size and salary are often linked to each other (Berger & Hannan 1988). The second reason is that managers are overconfident and overestimate synergy gains associated with M&As (Petzemas & Doukas 2007). The overestimating of synergies can be a result of political factors, organizational routines, agency issues and limited information accessibility (Herbert 1957).

M&As can be friendly and hostile (Bacon & Gersdorff 2009). Friendly takeovers are acquisitions approved by the Board of Directors of the target firm (Bacon & Gersdorff 2009). Hostile takeovers are acquisitions against the will of the target firm by purchasing the majority of the outstanding shares of a target company (Bacon & Gersdorff 2009).

M&As can be horizontal, vertical, or conglomerate (Herger & McCorriston 2014). Horizontal M&As are takeovers of companies in the same industry (Herger & McCorriston 2014). The main motives for horizontal M&As are advantages of economies of scale and increasing market share (Green & Cromley 1982). Vertical M&As are takeovers of target firms that are in the supply chain of the bidding firm (Herger & McCorriston 2014). The main motive for vertical M&As is to guarantee an adequate supply of materials at a reasonable price for the bidding firms (Motis 2007). Conglomerate M&As are takeovers of two different firms making different products (Herger & McCorriston 2014). The main motive for conglomerate M&As is diversification benefits (Motis 2007). According to Herger and McCorriston (2014), conglomerate M&As are rarely mentioned or lumped together in a “non-horizontal” group.

(6)

B. Efficient market theory

To see whether M&As create value, researchers look at the announcement effect of M&As on shareholder return (Asquith & Kim 1982; Dodd & Ruback 1977; Huang & Walking 1987; Asquith et al. 1983; Beltratti & Paladino 2013). The announcement effect is studied by an event study (Jong 2007). This method of research assumes that the market is efficient (Jong 2007). There are three forms of market efficiency, according to Fama (1970) and Ross (1998): the weak form, the semi-strong form, and the strong form.

The weak form, also known as the random walk theory (Fama 1970), states that it is not possible for an investor to achieve positive abnormal returns by using the information on historical prices (Malkiel 1989). If the market is in the weak form, then the past information is already incorporated in the share price (Ross 1998).

The next form of the efficient market theory is the semi-strong form (Fama, 1970). In this form share prices fully reflect all available public information, meaning neither technical nor fundamental analysis can be used to obtain abnormal returns (Malkiel 1989). However, abnormal returns can be achieved by having information that is not publicly available, for instance inside information (Ross 2008).

The last form, according to Fama (1970), of the efficient market theory is the strong form. The strong form states that all available, whether public or private, information is fully reflected in the share price, meaning that abnormal returns cannot be obtained (Malkiel 1989). C. Previous research on the announcement effect of M&A on shareholder return

In the past, a lot of research has been done on the effect of M&As on shareholder return. In this section, several studies that discuss the announcement effect are highlighted.

Asquith & Kim (1982) investigate the effect of M&A announcements on the wealth of the shareholders from the target firm. Their sample consists of public merging firms that engage in complete mergers between 1960 and 1978. They find a positive announcement effect and conclude that shareholders of the target firm gain from a merger bid due to the premiums paid to the target firms. This finding is consistent with Dodd & Ruback (1977), Kummer & Hoffmeister (1978), Bradley (1980), Jarrell & Bradley (1980), Asquith et al. (1983), Ruback (1983), and Huang & Walking (1987).

Dodd & Ruback (1977) find that only successful bidders earn significant positive abnormal returns from a M&A announcement in the United States between 1970 and 1977. This finding is consistent with Asquith et al. (1983), who show that United States bidding firms earn

(7)

significant positive abnormal returns between 1975 and 1983 and conclude that M&As create wealth for the bidding firms’ shareholders. Asquith & Kim (1982) investigate the effect of M&A announcements on the shareholder return of bidding firms. Their sample consists of public mergers between 1960 and 1978 that are completed. They find no evidence that M&As create or destroy value for the bidding firms’ shareholders. This finding is consistent with Eckbo (1983). Dodd (1980), on the other hand, investigates the effect of M&A announcements on the return of NYSE bidding firms’ shareholders and finds a significant abnormal return of -1.09% for 60 successful bidders. Dodd (1980) concludes that merger bids are, on average, negative net present value investments for shareholders of the bidding firm. This finding is consistent with Franks et al. (1992) who find that shareholders of NYSE bidding firms suffer a statistically significant loss of about 10% during a M&A announcement.

Different researchers find different explanatory variables that influence the abnormal returns coming from a M&A announcement. Asquith et al. (1983) conclude that the announcement effect of M&As is positively related to the relative size of the merger partners and that the gains for bidding firms are larger for mergers that are completed. Moeller et al. (2005), who examine a sample of 12.023 M&As by public firms in the US, find that larger bidding firms destroy value for the company during a M&A announcement due to the fact that managers of larger firms maximize their own utility instead of their shareholders’ utility. Moeller & Schlingemann (2004) investigated the difference in M&A deals between cross-border and domestic target firms and conclude that the announcement effect of M&A deals is negatively related with cross-border M&As. This finding is consistent with Andrade et al. (2001) and Very & Schweiger (2001). However, Bassen et al. (2010) show a positive effect of cross-border M&As for US bidding firms between 1990 and 2004. The finding of Bassen et al. (2010) is consistent with Seth et al. (2002) who also find a positive effect of cross-border M&As as a result of diversification benefits created by the bidding firms. Dutta et al. (2013) show in their study that acquisitions by Canadian bidding firms financed with stock are more positively related than acquisitions financed with cash. On the other hand, Myers & Majluf (1984) find that cash is chosen as payment method if the target firm is undervalued, which leads to positive abnormal returns for bidding firms that use cash as payment method. Ghosh and Jain (2000), who examine a sample of 239 completed M&As between 1978 and 1987, find a positive relation between the leverage of a bidding firm and the returns from an announcement of M&As. This is consistent with Bruner (1988). Both researchers conclude that firms are able to signal better M&As to the market because of the monitoring- and agency benefits of debt (Ghosh and Jain 2000; Bruner 1988). Beltratti & Paladino (2013), who focus on a sample of banking firms in

(8)

Europe during the financial crisis find that announcement effect of M&As is more positive during the crisis than before the crisis, because it is a signal of quality of the firm (Beltratti & Paladino 2013). Managers of healthy firms see possibilities to expand and strengthen their position in the market by using M&As (Beltratti & Paladino 2013).

III. Hypotheses

The paper of Beltratti & Paladino (2013) notices that the crisis represents opportunities for stronger firms. They conclude that healthy firms, with strong capital and liquidity, have the opportunity to take over competitors at distressed prices and thereby increasing their market share and profitability during the crisis (Beltratti & Paladino 2013). Though the financial crisis had severe impacts on large firms, the smaller firms did not escape from the crisis (Chowdhury 2011). Chowdhury (2011) find that smaller firms are hit even harder by the financial crisis than larger firms, because increased payment delays on receivables led to a shortage of working capital which, in contrast to larger firms, smaller firms can hardly sustain because banks were not able to lend capital to the smaller firms. The financial crisis is characterized as a period where credit became difficult to obtain (Strahan 2012). Moreover, the large firms in the United States get 30 percent of their financing from the banks, whereas smaller firms in the United States get 90 percent of their financing from the banks (Chowdhury 2011). Because of these financial constraints, especially for small firms, smaller firms that engage in M&As during the crisis may signal the market that their firm is healthy. Therefore I expect a more positive announcement effect for the smaller firms. The hypotheses that I research in this study are stated as follows:

H0: The announcement effect of small bidding firms during the crisis is equal or less positive than the announcement effect of large bidding firms.

H1: The announcement effect of small bidding firms is more positive than the announcement effect of large bidding firms.

(9)

IV. Research design

A. Methodology

I study the announcement effect of M&As by means of an event study. This method is developed by Fama et. al (1969) and is widely used to measure the effect of events on the share price. The effect of the announcement of M&As on the share price is measured by the abnormal returns. The abnormal returns are defined as the realized return minus the expected return (Fama 1969). The formula of the abnormal return is stated as follows:

𝐴𝑅_𝑖𝑡 = 𝑅_𝑖𝑡− 𝐸𝑅_𝑖𝑡 (1)

where:

• ARit is the abnormal return of firm i on day t; • Rit is the realized return of firm i on day t; • ERit is the expected return of firm i on day t.

The expected return can be estimated using the mean-adjusted model, the market adjusted return model model or the market model (Jong 2007).

In the mean-adjusted model, the expected return is calculated by averaging a series of past returns (Jong 2007). According to Jong (2007), the disadvantage of this model is the omission of market-wide stock price movements from the benchmark return. This means that results are biased if the whole market moves in one direction, up or down, in the event period (Jong 2007).

The market adjusted return model corrects for this omission (Jong 2007). In this model, the expected return is estimated by the market index for that period (Dyckman et al. 1985). The market adjusted return model assumes that the ‘beta’ of each share is equal to one (Jong 2007). This assumption is not the case for every firm, therefore I use the market model to perform the event study.

In the market model, the abnormal returns are defined as residuals of the market model. The market model formula is stated as follows:

𝑅_𝑖𝑡 = 𝛼_𝑖𝑡+ 𝛽_𝑖 ∗ 𝑅_𝑚𝑖+ 𝜀 (2)

where:

• Rit is the realised return of firm i on day t; • αit is the constant factor of firm i;

(10)

• Rm is the market return for firm i; • ε is the error term of the model.

The realised return is not corrected by stock splits or dividend payments, consistent with the study of Beltratti & Paladino (2013). The estimation window is the period of trading days that is used to calculate the abnormal returns (Jong 2007). Consistent with Thompson (1995), I estimate the βi and αi using an estimation window of 200 days [-250,-50] to exclude the effects of rumours before the announcement date. The market index is calculated as the value-weighted return on all NSYE, AMEX and NASDAQ stocks, that are taken from the Fama-French database. After estimating the market model with OLS, I use the estimated coefficients to calculate the expected returns. The expected returns are defined in formula terms as follows:

𝐸𝑅_𝑖𝑡 = 𝛼̂_𝑖+ 𝛽̂ ∗ 𝑅_𝑖 _𝑚𝑡 (3)

where: all variables are defined as described above.

With formula (2) and (3) it is possible to calculate the abnormal returns (1). The abnormal returns are calculated over the 10 days around a M&A announcement to investigate the short-term effect of a M&A announcement on the bidding shareholders’ return. The event windows of this paper to calculate the abnormal returns are [-5,+5], [-3,+3] and [-1,+1] to investigate the short term effect of M&A announcement on shareholder return. After calculating the abnormal returns for each day in the event window, I calculate the cumulative abnormal returns. I do this as follows:

𝐶𝐴𝑅𝑖,𝑇 = ∑𝑁𝑖=1𝐴𝑅𝑖,𝑡 (4)

where:

• CARi,T is the cumulative abnormal return of firm i in period T; • ∑𝑁

𝑖=1 is the sum for the event window N for firm i [-5,+5], [-3,+3], [-1,+1]; • ARi,t is the abnormal return of firm i on day t.

After the CARs of each firm are calculated, the outliers and missing data are removed from the sample. Then, I calculate the cumulative average abnormal return (CAAR) and test whether the CAAR is significant different from zero using a t-test. The formula to calculate the CAAR is stated as follows: 𝐶𝐴𝐴𝑅_𝑇 = 1 𝑛∗ ∑ 𝐶𝐴𝑅𝑖,𝑇 𝑁 𝑖=1 (5) where:

• CAART is the cumulative average abnormal return over the event window T; • n is the number of firms;

(11)

• The other variables are defined as described above.

After the CAAR of each event window is calculated, a t-test is used to detect the significance of the abnormal returns. The t-value is calculated as follows:

𝑡_{𝐶𝐴𝐴𝑅}_𝑇 = √𝑁 ∗_𝑆𝐶𝐴𝐴𝑅

𝐶𝐴𝐴𝑅 (6)

where:

• tCAARt is the t value of the cumulative average abnormal return on event window T; • SCAAR is the standard deviation of the cumulative average abnormal return;

• the other variables are defined as described above.

If t-value > 1.96, then the announcement effect is positive and significant different from zero with a significance level of 0.05. To investigate whether the announcement effect of small bidding firms during the crisis is more positive than the announcement effect of large bidding firms, I estimate the next regression model by using OLS:

𝐶𝐴𝑅𝑖,[−1,+1]= 𝛼 + 𝛽1∗ 𝑆𝐼𝑍𝐸 + 𝛽2∗ 𝐶𝑅𝐼𝑆𝐼𝑆 + 𝛽3∗ 𝑆𝐼𝑍𝐸 ∗ 𝐶𝑅𝐼𝑆𝐼𝑆 + 𝛽4∗ 𝐶𝐴𝑆𝐻 + 𝛽₅∗ 𝐶𝑅𝑂𝑆𝑆𝐵𝑂𝑅𝐷𝐸𝑅 + 𝛽₆∗ 𝑀𝑇𝐵𝑅 + 𝛽₇∗ 𝐿𝐸𝑉𝐸𝑅𝐴𝐺𝐸𝐴𝐶𝑄

where:

• CARi,[-1,+1] is the cumulative abnormal return for firm i during the event window [-1,+1];

• SIZE is the log of the total assets of firm i at the end of the year of the announcement;

• CRISIS is a dummy variable equal to 1 if firm i engages in a M&A during the

financial crisis and 0 if not;

• SIZE * CRISIS is an interaction term between the SIZE of firm i and the dummy

variable CRISIS of firm i;

• CASH is a dummy variable equal to 1 if firm i pays for the M&A with cash and 0

otherwise;

• CROSSBORDER is a dummy variable equal to 1 if firm i engages in a M&A with

the target firm outside the US and 0 otherwise;

• MTBR is the market to book ratio of equity of firm i at the end of the year of the announcement;

• LEVERAGE is the ratio between the total liabilities and total assets of firm i at the end of the year of the announcement.

CASH is included as control dummy variable because the choice of payment may impact the

(12)

given the access to private information, cash is chosen as payment method only if the target firm is undervalued. Travlos (1978) find that M&As paid with shares are associated with negative shareholder wealth effects at the announcement, while M&As paid with cash are associated with normal or positive announcement effects. Therefore is the control variable

CASH expected to have a positive relation with the CAR of the firms.

CROSSBORDER is included as control dummy variable because Beitel et al. (2003)

find in their study that M&As with the target firm outside the country of the bidding firm signal less cost savings, less understandability by the capital markets and are difficulty approved by the shareholders of the bidding firm, which results in a negative shareholder wealth effect at the announcement. Moeller & Schlingemann (2004) investigate the difference between cross-border and domestic firms and find that cross-cross-border M&As are associated with negative shareholder wealth effects at the announcement. Because of these findings, my expectation is that the control variable CROSSBORDER is negatively related to the CAR of the firms.

MTBR is included as control variable because Dong et al. (2006) find that firms with

higher market-to-book ratios make worse acquisitions and conclude that a higher MTBR has a negative effect on the announcement effect. On the other hand, Rajan and Zingales (1995) find that firms with lower market-to-book ratios make worse acquisitions and find a negative effect on the announcement of firms with lower market-to-book ratios. Therefore the expected relation with the CAR of the firms is ambiguous.

LEVERAGEACQ is included as control variable because Lang et al. (1991) find that

firms with higher leverage have a positive impact on the market reaction of a M&A because firms with higher leverage have less cash to spend and may lose their jobs if their firm gets into financial distress. My expectation is that the control variable LEVERAGEACQ is positively related to the CAR of the firms.

B. Data

In this part of the research design the data is described. The sample used in this paper is obtained from the database of Zephyr. Zephyr is a database consisting of information about M&As in the world. In Zephyr the following filters are used:

- Deal type: Acquisition, Merger - Listed acquiring company

(13)

- Time period: 01/01/2005 – 01/01/2011 - Country: United States

- Current deal status: completed

This first filter is applied to exclude non M&A deals. The second filter is applied to exclude non-listed firms. The reason for this is that there is more data available about the share prices for listed firms. The share prices are needed to get stock returns. The stock returns are needed to perform an event study (Jong 2007). The third filter is applied to exclude bidding firms that do not obtain full control in the target firm after the M&A. The bidding firms can obtain higher gains when bidding firms obtain full control in the target company (Kiymaz 2004). This is because bidding firms can set their expert knowledge and own management style in the target firm which results in higher gains for the bidding firms after a M&A (Kiymaz 2004). Without full control, shareholders may not see the value of a M&A (Kiymaz 2004), which may give a bias to my results. The fourth filter is applied to get M&A deals in the period between 01/01/2005 and 01/01/2011. This period starts from 01/01/2005 because Martynova and Renneboog (2006) stated in their research that the economic downturn started from 2005. Moreover, 2004 is excluded because it was a year of economic expansion (Martynova and Renneboog 2006), which would give my results an upward bias. In this paper, the crisis period starts in January 2007 and ends in November 2010, which is the same period as in the paper of Beltratti & Paladino (2013) on which my paper is built. The economic downturn ended around 2011 (Beltratti & Paladino 2013), which is the end date of the studied period. The fifth filter is applied to exclude M&A deals in which the bidding firm is not from the United States. This country is chosen because the crisis had a significant impact on firms in the United States and the data availability is high. The sixth filter is applied to exclude incomplete M&A deals that are not plausible for the shareholders of the bidding firm to react.

With this dataset, I find a sample consisting of 7734 M&A deals. By checking my data for the event study I still find 28 deals for which the bidding firm is unlisted. There were 9 deals with an invalid ISIN code. These deals are removed. To get the results from the event study in WRDS the ISIN numbers of each company is converted into CUSIP codes. 875 companies did not require the right CUSIP code in WRDS. There were 741 companies that had too little data available in the event window. 41 companies had too few estimation period days with the data available and there were 3 companies that had too many event period days with missing data. Therefore the sample consists of 6037 companies, which is still high enough to do the event study and a regression analysis.

(14)

V. Results

A. Event study

For the event study three different windows are analysed [-5,+5], [-3,+3] and [-1,+1]. First, the cumulative abnormal returns of the three different event windows are plotted in figure 1 to see the differences between the CAR of each event window. The event window [-5,+5] is plotted in figure 1A, the event window [-3,+3] is plotted in figure 1B and the event window

[-1,+1] is plotted in figure 1C. [-5,+5] CA R [-3,+3] CA R Figure 1A Figure 1B

Figure 1A. Graphical plot of the cumulative abnormal returns against the day relative to the event for 6,037 firms in the United States participating in M&A. The cumulative abnormal returns are calculated through the database of WRDS using the market model regression given by: 𝐴𝑅𝑖𝑡= 𝑅𝑖𝑡− 𝐸𝑅𝑖𝑡.

Figure 1B. Graphical plot of the cumulative abnormal returns against the day relative to the event for 6,037 firms in the United States participating in M&A. The cumulative abnormal returns are calculated through the database of WRDS using the market model regression given by: 𝐴𝑅𝑖𝑡= 𝑅𝑖𝑡− 𝐸𝑅𝑖𝑡.

(15)

The plots in figure 1 show that the largest change in cumulative abnormal return is during the event window of [-1,+1]. Another remarkable point that figure 1C shows is that there seems to be an information leakage before the event date by looking at the cumulative abnormal returns between [-1,+1]. However, this is not further investigated in this thesis. In table 1 the cumulative abnormal returns are given for the event window [-5,+5] to see at which event dates exist significant cumulative abnormal returns.

Table 1. In this table the short-term effect of a M&A announcement on the share price return is presented. The event window is [-5,+5] with the event day 0. The n is the number of US bidding firms that are analysed. The AR(mean) is the cumulative abnormal return in %. The AR (median) is the median of abnormal returns. The t-value tests whether the CAR is significant different from zero with a t-test. The p-value is the smallest significance level at which the null hypothesis can be rejected.

Event date N AR (mean) AR (median) t-value p-value

[ -5,+5 ] -5 6,037 -0.01% -0.04% 0.470 0.342 -4 6,037 -0.01% -0.05% -0.161 0.436 -3 6,037 0.02% -0.03% 0.678 0.249 -2 6,037 0.00% -0.06% -0.050 0.480 -1 6,037 -0.03% -0.09% -0.934 0.175 0 6,037 0.30% 0.16% 8.691*** 0.000 1 6,037 0.22% 0.14% 6.379*** 0.000 2 6,037 -0.01% -0.05% -0.218 0.414 3 6,036 -0.01% -0.04% -0.326 0.372 [-1,+1] CA R

Figure 1C. Graphical plot of the cumulative abnormal returns against the day relative to the event for 6,037 firms in the United States participating in M&A. The cumulative abnormal returns are calculated through the database of WRDS using the market model regression given by: 𝐴𝑅𝑖𝑡= 𝑅𝑖𝑡− 𝐸𝑅𝑖𝑡.

(16)

4 6,036 0.04% -0.06% 1.208 0.114

5 6,036 -0.01% -0.06% -0.386 0.350

The symbols *,**,*** denote the statistical significance at the 0.10, 0.05 and 0.01 levels, respectively, using a one-tail t- test.

Table 1 confirms the idea that the event window [-1,+1] is the most significant. The results show that at the event date and one day after the event date the cumulative abnormal returns are positive and significantly different from zero. Table 1 shows that the cumulative abnormal return at the event date is 0.30%, which is statistically different from zero with a 0.01

significance level. The result one day after the event (t=1) show a cumulative abnormal return of 0.22%, which is also statistically different from zero with a 0.01 significance level.

Because the results are significant and positive, I assume that the announcement of a M&A create value. In the next table, the summary statistics for the cumulative abnormal returns are given for the three different event windows.

Table 1. This table presents the cumulative average abnormal returns for three different event

windows. The event window shows the days before and after the event date of zero which are analysed. The mean is the average of the cumulative abnormal returns over the event window; also called cumulative average abnormal returns (CAAR). The n is the number of US bidding firms that are

analysed. The St. Dev. Is the standard deviation of the CAR. Q1 and Q3 are the 1st_{and 3}rd_quartiles

that show the lowest 25% and the highest 25% of the observations. The median is the observation that separates the lowest and the highest 50% of each other.

Event window Mean N St. dev. Q1 Median Q3

[-5,+5] 0.45%*** _6,036 _8.23% _-3.10% _0.06% _3.80%

[-3,+3] 0.44%*** 6,037 6.37% -2.40% 0.07% 2.99%

[-1,+1] 0.49%*** 6,037 4.80% -1.50% 0.08% 1.90%

The three event windows [-5,+5], [-3,+3] and [-1,+1] show that the cumulative average abnormal returns are 0.45%, 0.44% and 0.49%, respectively, which are all significant different from zero. The CAAR is positive and significantly different from zero, which suggest that the announcement effect of a M&A is positive. This finding is consistent with Asquith et al. (1983) and Dodd & Ruback (1977), who find evidence of wealth creation for the bidding firms at the announcement. Beltratti & Paladino (2013) find a positive announcement effect in the crisis. In table 3 are the summary statistics presented for the firms participating in M&As during the crisis.

(17)

Table 3. This table presents the cumulative average abnormal returns for three different event windows during the crisis. The event window shows the days before and after the event date of zero which are analysed. The mean is the average of the cumulative abnormal returns over the event window; also called cumulative average abnormal returns (CAAR). The n is the number of US bidding firms that are

analysed. The St. Dev. Is the standard deviation of the CAR. Q1 and Q3 are the 1st_{and 3}rd_quartiles

that show the lowest 25% and the highest 25% of the observations. The median is the observation that separates the lowest and the highest 50% of each other.

Event window Mean N St. dev. Q1 Median Q3

[-5,+5] 0.50%*** 2,718 7,95% -3.10% 0.19% 4,10%

[-3,+3] 0.44%*** 2,721 6.50% -2.50% 0.12% 3.20%

[-1,+1] 0.48%*** 2,721 4.83% -1.54% 0.13% 2,16%

The three event windows show that the cumulative average abnormal returns are positive and significantly different from zero, suggesting that the announcement effect of a M&A is positive in the crisis. This finding is consistent with the finding of Beltratti & Paladino (2013).

B. Regression model

Next, I examine the correlation to see which variables correlate with each other. This is done to see if some variables are correlated with each other, which may give a bias to my results.

In table 3 the correlation matrix is presented at which all the variables of the regression model are presented. The correlation measures the strength and direction of a linear relation between two variables.

(18)

Table 2. Correlation matrix of the model variables. The correlation measures the strength and

direction of a linear relation between two variables. The sign. (2-tailed) is the p-value of the t-test. The p-value is the lowest possible significance level at which the null hypothesis is rejected.

Table 3 shows that the Pearson correlation between SIZE and CAR is about -0.078, which indicates that there is a moderate negative relation between these variables. This suggests that as the size of a firm increases, the cumulative abnormal returns of the firm decreases. This finding corresponds with the findings of Moeller, Schlingemann and Stulz (2004), who find that large firms that participate in M&As will result in large losses for the company. The Pearson correlation between CRISIS and CAR is about 0.020, but not significant. This finding does not correspond with the finding of Beltratti & Paladino (2013) who find that the crisis had a positive relation on the cumulative abnormal returns of a firm. The Pearson correlation between the CAR and SIZE*CRISIS is 0.013, but also not significant. I do not want to include variables that causes multicollinearity in my estimation, because it undermines the statistical significance of an independent variable (Allen 1997). Variance Inflation Factor (VIF) values above 4 indicate multicollinearity in my estimation (O’Brien 2007). The VIF is calculated as follows: 𝑉𝐼𝐹 =

1

1−𝑅2 (O’Brien 2007). Because none of my variables have a VIF above 4, there is no

multicollinearity in my data. Therefore, I continue with estimating the following model: 𝐶𝐴𝑅𝑖,−1,+1= 𝛼 + 𝛽1∗ 𝑆𝐼𝑍𝐸 + 𝛽2∗ 𝐶𝑅𝐼𝑆𝐼𝑆 + 𝛽3∗ 𝑆𝐼𝑍𝐸 ∗ 𝐶𝑅𝐼𝑆𝐼𝑆 + 𝛽4∗ 𝐶𝐴𝑆𝐻 +

𝛽₅∗ 𝐶𝑅𝑂𝑆𝑆𝐵𝑂𝑅𝐷𝐸𝑅 + 𝛽₆∗ 𝑀𝑇𝐵𝑅 + 𝛽₇ ∗ 𝐿𝐸𝑉𝐸𝑅𝐴𝐺𝐸𝐴𝐶𝑄

The results of the regression model are presented in table 4 below. In this table, the dependent variable is the cumulative abnormal returns and the coefficient of interest is 𝛽3.

CAR [ -1,+1 ] SIZE CRISIS SIZE*CRISIS CASH CROSSBORDER MTBR LEVERAGE CAR [-1,+1] 1 SIZE -.78*** 1 (.000) CRISIS .020 .019 1 (.152) (.168) SIZE*CRISIS .013 .182*** .977*** 1 (.329) (.000) (.000) CASH .001 -.066*** -.052*** -.073*** 1 (.920) (.000) (.000) (.000) CROSSBORDER -.005 .089*** .051*** .070*** -.104*** 1 (.696) (.000) (.000) (.000) (.000) MTBR .001 .003 -.016 -.014 -.003 -.020 1 (.955) (.816) (.259) (.295) (.853) (.157) LEVERAGE .006 .044*** .012 .014 .013 -.012 -.015 1 (.642) (.001) (.381) (.300) (.351) (.365) (.265)

(19)

Table 3. This table presents the results from the regression model: CARi,-1,+1 = α + β1 * SIZE + β2 *

CRISIS + β3 * SIZE * CRISIS + β4 * CASH + β5 * CROSSBORDER + β6 * MTBR + β7 * LEVERAGE.

The dependent variable is the cumulative abnormal return for firm i for the event window [-1,+1]. The beta coefficient is the degree of change in the dependent variable for every 1 unit of change in the independent variable. The std. error is the standard error of the beta coefficient that measures the precision of the estimate of the beta coefficient. The t-value tests whether the beta is significant different from zero with a t-test. The p-value is the smallest significance level at which the null hypothesis can be rejected.

Coefficients

β Std. Error t-value p-value

.042 .007 5.612*** .000 SIZE -.003 .000 -5.312*** .000 CRISIS -.019 .010 -1.945* _.052 SIZE*CRISIS .001 .001 2.219*** _.027 CASH .000 .003 .070 .944 CROSSBORDER -.001 .002 -.605 .545 MTBR .000 .000 .102 .919 LEVERAGE .000 .000 .815 .415

The symbols *,**,*** denote the statistical significance at the 0.10, 0.05 and 0.01 levels, respectively, using a one-tail t- test

The coefficient on SIZE is -0.003, suggesting a significant negative effect with respect to the cumulative abnormal returns in the event window [-1,+1] for firm i. The interpretation of this value is that when I increase the size of the firm with 1%, the expectation is that the cumulative abnormal return will decrease by 0,3%. This confirms my expectations based on Moeller et al. (2005) who find that shareholders of large bidding firms experience significant wealth losses when they announce a M&A due to managerial hubris. Another reason for this finding is that managers of larger firms participate in M&A to maximize their own utility instead of their shareholders’ value (Trautwein 1990), which result in a negative announcement effect for larger firms.

The coefficient on CRISIS is -0.019, which suggests a significant negative effect, at the 0.10 significance level, with respect to the cumulative abnormal returns in the event window [-1,+1] for firm i. The interpretation of this value is that when firms engage in M&A during the

(20)

crisis, the expectation is that the cumulative abnormal return will decrease by 1.9%. This finding is inconsistent with the findings of Beltratti & Paladino, who find a positive effect of the crisis on the cumulative abnormal returns around the announcement.

The coefficient of interest SIZE * CRISIS is 0.001, which suggests a significant positive effect on the cumulative abnormal returns. When I increase the size of the company during the crisis with 1%, the expectation is that the cumulative abnormal return coming from a M&A announcement will increase by 0.1%. This is inconsistent with my expectations that the announcement effect of smaller firms is more positive than the announcement effect of larger firms. This study shows that the announcement effect of larger firms is more positive during the crisis than for smaller firms. Therefore, the null-hypothesis is not rejected. The reason for this can be that larger firms have more capabilities than smaller firms buying target firms for distressed prices during a crisis (Beltratti & Paladino 2013), because larger firms depend less on credit than smaller firms. The possible value creation from these M&As lead to the positive announcement effect for the larger firms. Another reason that my study does not show the effect for smaller firms can be that smaller firms use M&As during the crisis to diversify their bankruptcy risk. Singhal & Zhu (2013) find that firms in financial distress use M&As to reduce the likelihood of bankruptcy and liquidation. Therefore small firms may use M&As as last chance to survive in the market. This will not lead to a positive effect for smaller firms, which is consistent with my results. The control variables show no significant effects on the cumulative abnormal returns.

VII. Sensitivity analysis

In this section two different measures for the coefficient SIZE are examined, to see whether these measures are consistent with the results above. Smyth et al. (1975) and Shalit & Sankar (1977) find that alternative measures of firm size are not interchangeable and can give other results examining the announcement effect of a M&A. Other measures to estimate the size of a firm are the total sales or the market value of equity (Smyth et al. 1975).

Estimating the coefficient SIZE by the logarithm of the total sales of firm i during the event window [-1,+1], gives me the results presented in table 5.

(21)

Coefficients

Constant .034 .007 4.502*** _.000 SIZE -.002 .001 -4.189*** .000 CRISIS -.010 .010 -1.030 .303 SIZE*CRISIS .001 .001 1.300 .194 CASH -.001 .003 -.471 .638 CROSSBORDER -.001 .002 -.400 .689 MTBR .000 .000 .116 .908 LEVERAGE .000 .000 .517 .605

The coefficient on SIZE is -0.002, suggesting a significant negative effect with respect to the cumulative abnormal returns in the event window [-1,+1] for firm i. The interpretation of this value is that when I increase the size of the firm with 1%, the expectation is that the

cumulative abnormal returns will decrease by 0.2%. The positive, and significantly different from zero, effect of SIZE measured by the total sales is consistent with the earlier finding in which SIZE is measured by the total assets. However, the findings for the coefficients on

CRISIS and SIZE*CRISIS are not consistent with earlier findings. When SIZE is measured by

the total sales, the coefficients on CRISIS and SIZE*CRISIS are still in the same direction as before, but are not significantly differ from zero. This means that there is no, statistical

significant different from zero, difference in the announcement effect between small and large firms. This finding is important to be aware of, because different measures on the variable

SIZE give me different results. This is consistent with Smyth et al. (1975) who find that

researchers should use all three measures of firm size proxies as robustness checks, because the measures are not interchangeable.

(22)

Estimating the coefficient SIZE by the logarithm of the market value of equity of firm i during the event window [-1,+1] gives me the results presented in table 6.

Coefficients

Constant .021 .004 5.190*** .000 SIZE -.002 .001 -4.711*** .000 CRISIS -.003 .005 -.582 .561 SIZE*CRISIS .001 .001 1.042 .297 CASH -.001 .003 -.293 .770 CROSSBORDER -.001 .002 -.330 .741 MTBR .000 .000 .149 .882 LEVERAGE .000 .000 .407 .684

The coefficient on SIZE is -0.002 and is statistically different from zero. This finding is consistent with earlier findings in which SIZE is measured by total assets and total sales. However, the findings on the variables SIZE and SIZE*CRISIS are not consistent with the findings in which SIZE is measured by the total assets. When I measure the size-effect by the market value of equity, I find that the coefficients on CRISIS and SIZE*CRISIS are in the same direction, but not significantly different from zero. This means that there is no, statistical different from zero, difference in the announcement effect between small and large firms.

Because the findings in table 5 and 6 give me different results, it is important to be aware of the measurement of firm size and the results I present in my thesis based on the measurement by total assets. The findings contribute to earlier evidence that different size

(23)

measures are not interchangeable and may give other results (Smyth et al. 1975; Shalit & Sankar 1975).

VIII. Conclusion

Beltratti & Paladino (2013) find that the announcement effect is positive in the crisis. In this research, I study whether the crisis effect is more positive for small bidding firms than for large bidding firms in the crisis in the United States. Results point out that during the crisis, the size-effect is significantly positively related to the cumulative abnormal returns around the announcement of a M&A. When the firm size is increased by 1% during the crisis, results suggest that the abnormal return from a M&A announcement increase significantly by 0.1%. However, I show that the results are inconsistent measuring the size-effect with two other measures. The more positive announcement effect for large firms can be the result of large firms having more capabilities to engage in M&As to acquire target firms for distressed prices, which result in value creation and a positive announcement effect of a M&A. The less positive announcement effect for small firms can be the result of using M&As to reduce the likelihood of bankruptcy and liquidation.

Limitations of this research, and for further research, is that there may be an information leakage around the announcement of M&As in the sample analysed. The consequence is that the abnormal returns are not find at the event date, but before the event date, which may bias my abnormal returns at the event date downwards. Another downside of the sample is that only public US firms are analysed, because of the choice of research method. Therefore the conclusion is only based on the public firms, but it may give other results including private firms. In further research, the results in the sensitivity analyses can be investigated, because I show that different measures give different results.

(24)

VIII. References

Allen, M. (1997). The problem of multicollinearity. Understanding Regression Analysis. New York: Springer, p. 176-180

Andrade, G. M. M., Mitchell, M. L., & Stafford, E. (2001). New Evidence and Perspectives on Mergers. Journal of Economic Perspectives, 15(2), p. 103–120.

Asquith, P., Bruner, R., & Mullins, D. (1983). The gains to bidding firms from merger.

Journal of Financial Economics, 11, p. 121-139

Asquith, P., & Kim, E. (1982). The impact of merger bids on the participating firms' security holders. Journal of Finance, 37, p. 1209-1228

Avery, C., Chevalier J. A., & Schaefer S. (1998) Why do managers undertake acquisitions? An analysis of internal and external rewards for acquisitiveness. Journal of Law,

Economics, & Organization, 14(1), p. 24-43

Bassen, A., Schiereck, D. & Wübben, B. (2010). M&A success of German acquisitions in the US – evidence from capital market and survey data. Applied Financial Economics, 20(7), p. 543-559.

Beitel, P., Schiereck, D., & Wahrenburg, M. (2004). Explaining M&A success in European banks. European Financial Management, 10, p. 109-139

Beltratti, A., & Paladino, G. (2013). Is M&A different during a crisis? Evidence from the European banking sector. Journal of Banking and Finance, 37, p. 5394-5405 Berger, A. N., & Hannan, T. H. (1998). The efficiency cost of market power in the banking

industry: A test of the `quiet life’ and related hypotheses. Review of Economics and

Statistics, 80, p. 454–465

Betton, S., Eckbo, B. E., & Thorburn, K. S. (2009). Merger Negotiations and the Toehold Puzzle. Journal of Financial Economics, 91, p. 158-178.

Bradley, M. (1980). Interfirm Tender Offers and the Market for Corporate Control.

Journal of Business, 53, p. 345-376

Bruner, R. F. (1988). The use of excess cash and debt capacity as a motive for Merger.

Journal of Financial and Quantitative Analysis, 23, p. 199-217.

Chatterjee, S. (1986). Types of synergy and economic value: The impact of acquisitions on merging and rival firms. Strategic Management Journal, 7(2), p. 119-139

Chowdhury, S. R. (2011). Impact of Global Crisis on Small and Medium Enterprises. Global

Business Review, 12(3), p. 377-399

Dodd, P. (1980). Merger proposals, management discretion, and stockholder wealth. Journal

of Financial Economics, 8, p. 105-138

Dodd, P., & Ruback, R. (1977). Tender offers and stockholder returns: An empirical analysis.

Journal of Financial Economics, 5, p. 351-373.

Dong, M., Hirshleifer, D., Richardson, S., & Teoh, S. (2006). Does investor misvaluation drive the takeover market? Journal of Finance, 61, p. 725-762

Dutta, S., Saadi, S., & Zhu, P. (2013). Does payment method matter in cross-border acquisitions? International Review of Economics and Finance, 25, p. 91-107 Dyckman, T., Philbrick, D., & Stephan, J. (1985). A comparison of Event Study

(25)

Methodologies Using Daily Stock Returns: A Simulation Approach. Journal of

Accounting Research, 22, p. 1-30

Eckbo, B. E. (1983). Horizontal Mergers, Collusion, and Stockholder Wealth. Journal of

Financial Economics, 11, p. 241-273

Fama, E. (1970). Efficient capital markets: a review of theory and empirical work.

Journal of Finance, 25, p. 383-417

Franks, J., Harris, F., & Titman, S. (1991). The postmerger share-price performance of acquiring firms. Journal of Financial Economics, p. 81-96

Gersdorff, N. V., & Bacon, D. F. (2009). U.S. Mergers and Acquisitions: A Test of Market Efficiency. Journal of Finance and Accountancy, 1(8), p. 1-8

Ghosh, A. & Jain, P. C. (2000) Financial leverage changes associated with corporate mergers.

Journal of Corporate Finance, 6, p. 377-402.

Green, M. B., & Cromley, R. G. (1982). The Horizontal Merger: Its Motives and Spatial Employment Impacts. Economic Geography, 58(4), p. 358-370

Hampton, J. J. (1989). Financial Decision Making: Concepts, Problems, and Cases, 4th edition. New Jersey: Prentice-Hall

Herbert, S. (1957). Models of Man, Social and Rational. New York: Wiley. p. 287

Herger, N., & McCorriston, S. (2014). Horizontal, Vertical, and Conglomerate FDI: Evidence from Cross Border Acquisitions. Working Papers 14.02, Study Center Gerzensee Huang, Y., & Walkling, R. A. (1987). Target Abnormal Returns Associated with Acquisition

Announcements: Payment, Acquisition Form, and Managerial Resistance. Journal of

Financial Economics, 19, p. 329–349

Jarrell, G., & Bradley, M. (1980). The Economic Effects of Federal and State Regulations of Cash Tender Offers. Journal of Law and Economics, 23, p. 371-407

Jong, F. (2007). Event study methodology. Working paper, Tilburg University.

Kim, E. H., & Singal, V. (1993). Mergers and Market Power: Evidence from the Airline Industry. The American Economic Review, 83(3), p. 549-569

Kiymaz, H. (2004). Cross-border acquisitions of US financial institutions: Impact of macroeconomic factors. Journal of Banking & Finance, 28, p. 1413-1439 Kummer, D., & Hoffmeister, R. (1978). Valuation Consequences of Cash Tender Offers.

Journal of Finance, 33, p. 505-516

Lang, L. H. P., Stulz, R., & Walkling, R. A. (1991). A test of the free cash flow hypothesis: the case of bidder returns. Journal of Financial Economics, 29(2), p. 315-335

Malkiel, B. G. (1989). Efficient Market Hypothesis. London: Palgrave Macmillan, p. 127-129 Martynova, M., & Renneboog, L. (2006). Mergers and acquisitions in Europe. ECGI Working

Paper Series in Finance, p. 1-83

Masulis, R., Swan, P. L., & Tobiansky, B. (2012). Do Wealth Creating Mergers and Acquisitions Really Hurt Acquirer Shareholders? School of Banking and Finance, 24, p. 1-62

Moeller, S., & Schlingemann, F. (2004). Are Cross-Border Acquisitions Different from Domestic Acquisitions? Evidence on Stock and Operating Performance for U.S. Acquirers. Journal of Banking and Finance, 29, p. 533-564

(26)

Moeller, S., Schlingemann, F., & Stulz, R. (2005). Firm size and the gains from acquisitions.

Journal of Financial Economics, 73(2), p. 201-228

Motis, J. (2007). Mergers and Acquisitions Motives. Toulouse School of Economics and

University of Crete

Myers, S., & Majluf, S. N. (1984). Corporate financing and investment decisions when firms have information that investors do not have. Journal of Financial Economics, 13, p. 187-221

O’Brien R. M. (2007). A Caution Regarding Rules of Thumb for Variance Inflation Factors.

Quality and Quantity, 41, p. 673-690

Petmezas, D., & Doukas, J. (2007). Acquisitions, overconfident managers and self-attribution bias. European financial management, 13(3). p. 531-577

Rajan, G. R. & Zingales, L. (1995). What Do We Know about Capital Structure? Some Evidence from International Data. Journal of Finance, 50(5), p. 1421-1460 Rani, M., Surendra, S. Y., & Jain, P. K. (2012). Impact of Mergers and Acquisitions on

Returns to Shareholders of Acquiring firms: Indian Economy in Perspective. Journal

of Financial Management and Analysis, 25(1), p. 1-26

Ravenscraft, D. J., & Scherer F. M. (1987). Sell-Offs, and Economic Efficiency. Washington: The Brookings Institution

Ross, S. A., Westerfield, R. W., & Jordan, B. D. (1998). Fundamentals of Corporate Finance. United States of America: McGraw-Hill

Ruback, R. S. (1983). Assessing Competition in the Market for Corporate Acquisitions.

Journal of Financial Economics, 11, p. 141-153

Seth, A., Song, K. P., & Pettit, R. R. (2002). Value creation and destruction in cross-border acquisitions: An empirical analysis of foreign acquisitions of U.S. firms. Strategic

Management Journal, 23(10), p. 921–940.

Shalit, S. S., & Sankar, U. (1977). The measurement of firm size. The Review of Economics

and Statistics, 59(3), p. 290-298

Sherman, A. J., & Hart M. E. (2006). Mergers & Acquisitions: from A to Z. The Netherlands: Amacom, p. 147

Singhal, R., & Zhu, Y. (2013). Bankruptcy risk, cost and corporate diversification. Journal of

Banking & Finance, 37(5), p. 1475-1489

Smyth, D. J., William, B. J., & Peseau, E. D. (1975). The measurement of firm size: theory and evidence for the United States and the United Kingdom. The Review of Economics

and Statistics, 57(1), p. 111-114

Steiner, P. (1975). Mergers: Motives, Effects, Policies. University of Michigan Press Strahan, P. E. (2012). Liquidity Risk and Credit in the Financial Crisis. FRBSF Economic

Letter, 15, p. 1-5

Thijssen, J. J. J. (2008). Optimal and strategic timing of mergers and acquisitions motivated by synergies and risk diversification. Journal of Economic Dynamics and Control, 32(5), p. 1701-1720

Thompson, R. (1995). Empirical methods of event studies in corporate finance. Handbooks in

(27)

Trautwein, F. (1990). Merger motives and merger prescriptions. Journal of Strategic

Management, 11, p. 283–295

Travlos, N. G. (1987). Corporate takeover bids, methods of payment, and bidding firms' stock returns. Journal of Finance, 42, p. 943-963

Very, P., & Schweiger, D. M. (2001). The Acquisition Process as a Learning Process: Evidence from a Study of Critical Problems and Solutions in Domestic and Cross-Border Deals. Journal of World Business, 36(1), p. 11–31.