Acquirer financial constraints, takeover characteristics, and short‐term performance of distressed target takeovers

Acquirer Financial Constraints, Takeover Characteristics,

and Short‐term Performance of Distressed Target Takeovers

University of Amsterdam, Amsterdam Business School

Master Thesis

Statement of Originality: This document is written by Student F. L. Gelens 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.

Abstract

Many motives for corporate acquisitions have been investigated, however the takeovers of financially distressed targets are less well explored. The papers that investigate distressed target takeovers examine the takeover characteristics and their wealth effects for the target and acquirer shareholders. These papers however, do not investigate the possible influence of acquirer financial constraints on these factors. This study examines whether acquirer financial constraints are related to the premiums, payment method, and short‐term performance of distressed target takeovers. Despite inconsistent findings across the different financial distress measurements used, the results suggest that acquirers of financially distressed targets are more likely to pay with equity, and this likelihood is increased when the acquirer is financially constrained. Additionally, the results of this research suggest that mergers and acquisitions in which distressed targets are combined with a financially unconstrained acquirer, lead to the highest wealth effects for the target company.

    Table of Contents 1. Introduction ... 4 2. Literature Review ... 7 2.1 Motives for distressed target takeovers... 7 2.2 Fire sales ... 8 2.3 Distressed target takeovers ... 9 2.4 Method of payment ... 12 2.5 Determinants of short‐term performance in M&A ... 13 2.5.1 Determinants of short‐term wealth effects for acquiring companies... 14 2.5.2 Determinants of short‐term wealth effects for target companies ... 14 2.6 Acquirer financial constraints ... 15 2.7 Hypothesis ... 16 3. Research Design ... 18 3.1 Data and sample selection ... 18 3.1.1 Data ... 18 3.1.2 Sample selection ... 18 3.2 Methodology ... 19 3.2.1 Financial constraints and Financial distress ... 19 3.2.2 Financial distress and the takeover premium ... 20 3.2.3 Financial distress and the method of payment ... 22 3.2.4 Short‐term M&A performance and its relationship to financial distress and takeover characteristics ... 24 4. Results ... 28 4.1 Descriptive statistics ... 28 4.2 Regression results ... 31 4.2.1 Premium Regression ... 31 4.2.2 Method of payment Tobit regressions ... 32 4.2.3 CAR Regressions ... 34 4.3 CAAR t‐tests... 35 4.3.1 Full cash versus Full stock CAARs ... 35 4.3.2 CAARs differences by acquirer constraints and target distress ... 36 4.4 Robustness Checks ... 37 4.4.1 Premium robustness check ... 37 4.4.2 Method of payment robustness check ... 37 4.4.3 CAAR differences robustness check. ... 38 5. Discussion ... 40 6. Conclusion ... 45 References ... 46 Appendix ... 50    

1. Introduction

Many studies investigate the main motives for corporate takeovers. All the motives are based on creating value for the acquirer in one way. In case of acquiring a financially distressed company (a company that has difficulties paying off its financial obligations to its creditors) a number of attractive motives can be found. The acquisition of a distressed company is potentially attractive in the means of market share, efficiency enhancement, cost reduction, and diversification (Jensen, 1986; Bradley, Desai & Kim, 1988; Amit et al. 1989; Morck et al. 1990, Jensen, 1991).

Little existing literature on Mergers and Acquisitions (M&A) focusses on the acquisitions of distressed target companies. Companies facing financial distress or even bankruptcy might see an acquisition as the most preferable exit path (Jensen, 1991). The stock and assets of these distressed companies often trade at prices that reflect the difficulties they face and may therefore be under pressure to sell assets or securities quickly to raise capital or to pay down debt. Acquirers may have an opportunity to acquire attractive assets or securities at a discount (Watchell et al, 2013). Attractive targets usually have interesting assets or high growth and improvement opportunities. The bidder might redeploy the distressed assets and realize post‐acquisition synergies (Bruton et al. 1994; Hotchkiss and Mooradan, 1998). However, a distressed acquisition is a risky investment due to the distress itself. The bidding company is taking a risk by acquiring such a firm since it is not sure it will be able to turn the distressed position around. If the bidder is improperly prepared, it might not be able to eliminate the target’s distress and might be subjected to a risk transfer from the distressed target (Bruyland and de Maeseneire, 2011).

Bidding companies face a number of choices in their deal consideration. An acquiring company has to make choices regarding the takeover premium and the method of payment (a payment in cash, stock, or a mixture of both). The choices of these deal characteristics will have an influence on the post‐merger capital‐ and ownership structure. In case of a distressed target takeover, a number of different factors, like information asymmetry and the availability of funds, might influence these choices compared to non‐distressed target takeovers. Another, rather unexplored factor that might influence earlier mentioned takeover characteristics is the level of financial constraints faced by the acquiring company. The scarce existing literature on distressed

target takeovers mainly focuses on a distressed position of the target but does not investigate constraints of the acquiring company. Many of the studies focus on the post‐ merger performance of financially distressed target takeovers but do not take into account the possible financial constraints the acquiring company is facing. Alshwer et al. (2011) find that financially constrained bidders are more likely to use stock in acquisitions but do not examine financial distress for the target companies. Amit et al. (1989) find that distressed targets earn lower abnormal returns than non‐distressed targets. Hotchkiss and Mooradian (1998) focus on acquisitions of bankrupt and reorganizing firms and find positive abnormal returns for both the bidder and the target. Both these studies do not investigate the effect of acquirer constraints and target distress on the takeover premium and method of payment choice. Clark and Ofek (1994) examine distressed target takeovers and find that bidders are unsuccessful at restructuring targets. They show that the post‐merger performance of the bidding companies is negatively related to the takeover premium. Their study provides limited evidence on the takeover premium and payment choice due to testing in a non‐ multivariate way. The research of Ang and Mauck (2010) focuses on the existence of fire sales and their relation to the takeover premium. They find that distressed targets receive higher takeover premiums compared to non‐distressed targets. Furthermore they find that acquirers do not gain from these transactions both over the short‐ and the long run. Both the research of Clark and Ofek (1994) and that of Ang and Mauck (2010) do not examine the effect of acquirer financial constraints on the payment characteristics and the post‐merger performance.

This study will investigate the relationships between acquirer financial constraints, the takeover characteristics, and the post‐merger performance of both the acquirer and the target in distressed target takeovers. The research question of this study is: How do

acquirer financially constraints affect the takeover premium, payment choice, and post‐ merger performance in distressed target takeovers?

This research will use a sample of 3,889 U.S. mergers and acquisitions between 1985 and 2014. The level of financial constraints for the acquirer is measured by the HP‐ index that was constructed by Hadlock and Pierce (2010). To check for robustness the Whited and Wu (2006) index of financial constraints will be used. In order to measure target financial distress a number of proxies, based on financials for the one or two years prior to the takeover announcement, will be examined. These proxies include

negative income, negative equity, the Altman‐, and Ohlson bankruptcy models (Ang and Mauck, 2011). An OLS‐regression will be used to find a relationship between the financial constraints, target distress, and the takeover premium. Following Faccio and Masulis (2005), Tobit and Ordered Probit models will be used to find evidence on the relation between financial constraints, target distress, and the method of payment. Abnormal returns to the bidding and target firm around the announcement will be used to measure short‐term performance. OLS regressions and multiple t‐tests on these abnormal returns will be used to find relationships between financial constraints, target distress, transaction characteristics, and the short‐term performance.

The remainder of this study will look as follows. The next section will discuss the existing literature on financial distressed takeovers, takeover premium, method of payment, and short‐term post‐merger performance. In this section the hypothesis, based on the existing literature, will be formulated. The third section will discuss the sample selection and the methodology that is used. The fourth section describes and presents the results of the regression models that were performed including the robustness checks. The fifth section discusses the found results and compares these findings to the related existing literature and analyzes its consequences for the constructed hypothesis. Finally, the sixth section presents the conclusion and possible implications for further research.

2. Literature Review

This section discusses the existing literature on distressed target M&As. First the motives for distressed target takeovers are stated. Second, the existing literature on fire sales and financially distressed takeovers is reviewed. Next the existing literature on takeover premium and payment choice is discussed. Furthermore, the short‐term wealth effects to acquirers and its determinants are reviewed. Finally, the hypotheses are constructed. 2.1 Motives for distressed target takeovers   Many studies investigate the main motives for corporate takeovers. In case of acquiring a distressed company, a number of attractive motives can be found. The achievement of economies of scale might be attractive to an acquirer, regardless of the targets past performance. When combining business that shares resources, economies of scope can be achieved (Seth, 1990). Some firms are distressed not because they lack the resource combinations but because of inefficient management. The change of management might improve the companies’ strategy (Jensen, 1988; Schleifer & Vishny, 1989; Hayward & Hambrick, 1997). Other motives for a distressed takeover could be the enhancements of market power, by increasing market share through the acquisition. Some takeovers can create value by facilitating tax savings (Devos et al., 2009). In summary, the acquisition of a company in distress is potentially attractive in the means of increasing market share, the achievement of synergies (Bradley, Desai & Kim, 1988; Morck et al. 1990), the reduction of costs (Jensen, 1986), diversification (Amit et al. 1989; Jensen, 1991), and more efficient use of existing assets (Jensen 1991; DePamphils, 2010). Jensen (1991) states that mergers and acquisitions are an effective way of resolving financial distress. Seth (1990) states that in general M&As increase the wealth of the stockholders that are involved in the takeover. Most of this wealth is transferred to the target firm stockholders, while the effect of M&As on the wealth of the acquiring firm vary (Martynova & RenneBoog, 2008).

2.2 Fire sales

Optimal capital structure theories suggest that firms choose their debt levels such that tax and agency benefits of debt are balanced with expected costs of financial distress. One of the indirect financial distress costs is the sale of assets at deep price discounts, the so‐called fire sales (Pulvino, 1998). In his paper Pulvino (1998) investigates if financial constraints cause firms to liquidate assets at discounts to fundamental values. He uses a sample including commercial aircraft transactions between 1978 and 1991. He shows that financially distressed airlines sell their planes at a significant discount. They are more likely to sell assets to industry out‐siders at low prices, especially during market recessions when the capital constraints limit inside buyers from paying full value for distressed firm’s assets. Pulvino does not provide evidence of fire sales from the bidders’ viewpoint.

Eckbo and Thorburn (2008) test for fire‐sale tendencies in case of automatic bankruptcy auctions. Their sample exists of 258 mandatory auctions of entire Swedish firms in bankruptcy. They find evidence that fire‐sale discounts exist when the auction lead to piecemeal liquidation, but not when the bankrupt firm is acquired as a going concern. They find that neither industry‐wide distress nor the industry affiliation of the buyer affect prices in going‐concern sales.

Ang and Mauck (2011) provide empirical evidence on the conjecture that in economic crises, firms could be forced to sell at deep discounts, or fire sale prices. They use a sample of 5,794 transactions between 1977 and 2008 of which 2,012 transactions are identified as distressed. They identify financial distress by four proxy variables: two years of negative income, negative equity, the Altman Z‐score and the Ohlson O‐score. Furthermore they use proxies for crisis periods. They state that fire‐sales occur in several conditions: non‐efficient market pricing or in crisis periods, the distressed firm is in a weakened bargaining position, there are inadequate numbers of bidders once the target receives the first bid, bidders do not commit behavioral errors, such as overconfidence and overpay, bidders do not overestimate the target’s value, or target managers who want to retain control may be willing to sacrifice interest of claimholders to sell at a discount in order to raise cash to continue operations. They state there are three types of fire sales: A transfer of wealth from the sellers to the buyers (in case of positive synergy gains for the bidder), a lose‐lose situation, and a win‐win situation in

which target firms are able to obtain funds and bidders gain from improving target performance. They find that distressed target firms receive a 30% higher offer premium in crisis period than distressed firms in normal periods. They also receive higher premium than non‐distressed firms in crisis periods. Furthermore their evidence suggests that acquirers do not gain over both the short‐ and long‐run.

2.3 Distressed target takeovers

In their research Amit et al. (1989) use a sample of 151 targets and 141 bidder firms on US mergers and acquisitions. They divide the mergers and acquisitions into three target subgroups. The first group is financially distressed targets that choose M&A as an alternative to bankruptcy. The second group consists of highly liquid target firms. The third group consists of the remainder of M&As. They expect that highly liquid targets are more able to obtain bids with high premium above market prices. They expect that the targets that choose M&A as an alternative to bankruptcy are less likely to receive a high premium since they are unlikely to have large free cash flows. They compare the target abnormal returns around the takeover announcement between the three subgroups. To identify the three subgroups they use the Altman bankruptcy score or Z‐ score (1968) that is based on business ratios, weighted by coefficients. They find statistically significant abnormal gains for target stockholders and negative abnormal returns for the bidding stockholders for the same period. Consistent with their expectations, stockholders of distressed target firms earned the lowest abnormal returns and stockholders with highly liquid positions earned the highest returns. The abnormal returns for the bidding firms are different. Both the bidders acquiring financially distressed or highly liquid companies did not demonstrate abnormal returns that were statistically different from zero. The remainder of the M&As however showed negative abnormal returns. In their research they control for premium, size, method of payment, and tender offers but they do not investigate these relationships in further detail.

Clark and Ofek (1994) study the effectiveness of mergers in restructuring distressed firms and examine determinants of the success of these restructurings by analyzing post‐merger performance using abnormal returns. They investigate 38 takeovers of financially distressed firms that occurred between 1981 and 1988. Financially distressed firms are identified by negative share returns prior to the announcement of

the acquisition. They find that the market shows some ability to forecast the restructuring’s success. Abnormal announcement period returns on the bidder’s shares are positively and significantly related to the post‐merger returns earned by the bidders on their investments in the targets. They find a positive relationship between the post‐ acquisition performance and the announcement abnormal returns. This indicates that the market identifies restructuring attempts that are not going to be successful the moment they are initiated. Furthermore, Clark and Ofek (1994) find that distressed target acquisitions involve fewer hostile takeovers, more bidders, and more targets in the same industry than acquisitions in general. They find a similar proportion of contested bids to the general population of acquisitions. They expected the premiums paid to the target shareholders of distressed targets to be different from the average target premiums. They state that the premiums will be higher if there is more to gain from combining the operations of a distressed target with the bidding company. On the other hand they state that premiums could be lower if distressed target companies have less bargaining power due to a weak condition or in case of low bidding competition. By using several factors that determine the success of the restructuring, they find a negative relationship between the size of the premium that was paid and post‐merger success of the combined firms. This implies that overpaying reduces the potential success of the restructuring. The findings for the premium, method of payment and the level of contested mergers are similar for distressed targets to those find in prior research examining non‐distressed targets. Since all tests on takeover premium, method of payment and the level of contested bidders are all tested separately, the conclusions that can be drawn from the research of Clark and Ofek (1994) are limited. Furthermore, their research does not compare regular acquisitions with financially distressed target acquisitions.

Hotchkiss and Mooradian (1998) investigate a sample of 55 acquisitions involving targets that are subjected to a Chapter 11 bankruptcy. They compare firms acquired in Chapter 11 to firms that are reorganized as independent companies. They use several ex‐merger measures of success of the transaction and the abnormal returns for both the bidder and target shareholders. Their analysis shows that there is no evidence that the post‐bankruptcy performance of firms reorganizing as independent companies differs from those acquired in Chapter 11. They find that firms in the same industry most often acquire bankrupt targets. Furthermore they find that bankrupt targets are on average

purchased at a 45% discount relative to prices paid for non‐bankrupt targets in the same industry. Consistent with the idea that acquisitions of bankrupt firms create value, they find that firms merged with bankrupt targets show significant improvements in operating performance, while matching non‐bankrupt transaction show no significant improvement. They find positive significant abnormal returns for the bidder and bankrupt target at the announcement of the acquisition. For the reorganizing matching transactions, they find positive abnormal returns to the target but not to the bidding firms. One of their possible explanations for this difference in bidder’s stock price reaction is that empire‐building managers find acquiring bankrupt firms less desirables because they require more complex negotiations with creditors and the courts. Since both the bankrupt and the non‐bankrupt companies are financially distressed, their research does not compare distressed takeovers with non‐distressed takeovers.

In relatively efficient markets, competing bids for a target firm often reflect the future value generated by an acquisition (Barney, 1988). However, sometimes the acquirer pays even more than the future value for a target because of the underestimation of the costs of exploiting potential synergies (Jemison & Sitkin, 1986, Roll, 1986; Salter & Weinhold, 1979). These valuation errors might lead to one overestimating bidder placing the highest offer. This phenomenon is the so‐called the winner’s curse (Oster, 1990). The chance of avoiding the winner’s curse may increase in case of distressed target acquisition in a business related to that of a potential acquirer. In case of financially distressed targets, the potential bidders will be taking more time to estimate the true value of firm. The distress of a target discourages the interest from some potentially competing bidders (Varian, 1988). In this case the acquirers who are interested in the distressed firm have more time to study the target. In case the industry of the acquirer and the target company are in the same industry, the bidding company can more easily uncover targets’ hidden problems and can more easily uncover opportunities for synergy and thus are more likely to discover undervalued assets (Dundas & Richardson, 1982). Thus, in case of a related distressed target takeover, the acquiring firm is less likely to suffer from the winner’s curse (Barney, 1988; Harrisson et al., 1991).

2.4 Method of payment

Many existing literature has been written about the determinants of the method of payment choice in M&A. Amihud et al. (1990), Chaney et al. (1991) and Martin (1996) study the determinants of payment choice in the US. Later, Faccio and Masulis (2005) and Schwieringa and Schauten (2008) focus on European countries. The determinants that are found include firms‐, deal‐ and target‐ specific factors. The choice of payment method has important implications for both the acquirer and target, including post‐ merger ownership structure, risk profile, and allocation of gains from the transaction (Faccio and Masulis, 2005; Golubov Petmezas and Travlos, 2012).

According to Hansen (1987) there is an asymmetric information problem in the real market, and this problem will cause misvaluation between the acquirer and target firms. Shleifer and Vishny (1992) argue that the purchase of illiquid assets is more likely to be financed with debt in case there is more uncertainty about the value of the acquired assets. Myers and Majluf (1984) and Hansen (1987) predict that if acquirer’s stock price experiences a run‐up for a certain period before the M&A transaction, stock payment will be offered because of overvaluation. They find significant evidence that there is a positive relationship between acquirer’s stock run‐up 1 year before the announcement date and the likelihood of stock payment. Martin (1996), and Faccio and Masulis (2005) find consistent evidence for European countries.

Due to limited cash and liquid assets, the payment in cash is usually financed by the use of debt. Hansen (1987) who takes into account the importance of firm size when attracting debt finds a negative relationship between the size of the acquiring firm and the probability of financing the takeover with stock. Martin (1996) uses financial leverage as a proxy for the ability to attract debt and finds a similar relationship. Faccio and Masulis (2005) and Swieringa and Schauten (2008) also find a negative relationship between the debt capacity of the bidding company and equity transaction.

In their research Jensen (1986) and Fishman (1989) find a positive relationship between the liquidity, measured by the available free cash flow, of an acquirer and a payment in cash. In contrast, Faccio and Masulis (2005) find a negative relationship between the availability of cash and cash payments in European M&As.

Jung, Kim and Stulz (1996) find that an acquirer with higher investment opportunities prefers to pay with stock. They define the investment opportunity

available as the market‐to‐book ratio and find a positive relationship with the likelihood of payment in stock. A higher market‐to‐book ratio is related to a higher level of tax‐ deductible expenditures and lower cash dividends, which make a cash payment less attractive. This evidence is consistent with the findings of Martin (1996) and Faccio and Masulis (2005).

Faccio and Masulis (2005) conclude that when the acquirer and target firm are in the same industry, they can get more up‐to‐date information and reduce the risk of asymmetric information. Therefore, the target firms are willing to accept a stock payment from the acquirer in the same industry. They find a positive relationship between cross‐industry M&As and a payment in cash.

Another widely studied determinant is the relative deal size. The existing literature is not in line with each other but significant relationships have been found. Grullon et al. (1997) find evidence of a positive relationship between the deal size and a payment in stock. Faccio and Masulis (2005) and Schwieringa and Shouten (2008) support this conclusion. The bigger the target firm’s assets relative to the acquirer’s assets, the more likely is a payment in equity.

2.5 Determinants of short‐term performance in M&A

In general, mergers and acquisitions create positive, statistically significant short‐term abnormal returns for targets and slightly negative abnormal returns for the acquirer. Andrade et al. (2001) and Georgen and Renneboog (2004) find large, statistically significant positive abnormal returns for the targets using an event study around the announcement. Jensen and Ruback (1983) and Jarrel et al. (1988) find slightly negative abnormal returns for the acquirers. A number of studies have sought to identify the underlying determinants of abnormal returns, including the method of payment, the relative size of the target company to the bidding company, the type of the acquisition, the growth prospects of the companies, market‐to‐book value ratios, and the number of bidding companies contesting the bid for the target company. Dennis et al. (2002) and Freund et al. (2007, 2008) argue that abnormal returns are negative for acquirers in diversifying acquisitions. Martynova and Renneboog (2006) show evidence of higher abnormal returns for targets in diversifying deals but negative abnormal returns for the acquirers. They find higher abnormal returns for targets involved in hostile deals compared to friendly deals. The opposite is found for the acquiring companies.

Furthermore Martynova and Renneboog (2006) find that abnormal returns are higher for targets in tender offers (negotiated deals) compared to mergers. By analyzing of the effect of method of payment on short‐term performance Faccio et al. (2006) find that abnormal returns are higher for cash offers compared to stock offers. Servaes (1991), Schwert (2000), and Moeller et al. (2005) report a positive relation between the acquirers’ abnormal return and their Tobin’s Q. However, Freund et al. (2008) argue that this relationship should be negative because overvalued firms are poor acquirers.

2.5.1 Determinants of short‐term wealth effects for acquiring companies

Pettyway and Yamada (1986) and Scanlon Trifts and Pettyway (1989) found that acquiring firm shareholders lost significantly when relatively large firms were acquired. Asquith, Brunner and Mullins (1983), and Jarrel and Poulsen (1989) provided contrary evidence by finding a significant positive relationship between relative sizes and returns to acquiring firms. They explained this relationship by arguing that small acquisitions would move with the performance of the acquirer whereas the acquisition of a relatively larger target would not. Walker (2000) found significant associations between abnormal returns and the method of payment, the relative size of the transaction, and contesting bidders. Travlos (1987) showed that, in pure share exchanges, bidding companies experienced significant negative abnormal returns during takeover announcements, while for cash acquisitions, bidding companies returns were normal. Honert, Barr, Affleck‐Graves & Smale 1988 found that for cash acquisitions the bidding firm’s cumulative abnormal returns (CARs) would decline rapidly after the announcement. For share‐based acquisitions, bidding companies exhibited random behavior. Travlos and Papaioannou (1991) examined impacts of method of payment on bidding firms’ stock returns at the initial announcement of the takeover bids. They found that the abnormal return of bidding firms on the announcement day were negative, and 0.5% higher for cash offers.

2.5.2 Determinants of short‐term wealth effects for target companies

Wansley, Lane and Yang (1983) indicated that target company shareholders earned around 16% more on average in case of cash payment acquisitions compared to stock payments. Huang and Walkling (1987) confirmed that abnormal returns to target companies associated with cash offers were significantly higher than those associated with share offers. Davidson and Cheng (1997) explained the difference in target’s

abnormal returns between cash and stock financed takeovers by the bid premium. They argued that large bid premiums are positively related to the abnormal returns, and showed that after controlling for bid premiums, there was no significant difference between cash and stock acquisitions. Suk and Sung (1997) looked at the effects of methods of payment, form of acquisition and type of offer on target firm’s abnormal returns around the takeover announcement. They showed that there is no difference in premiums between a stock and a cash offer. Georgen and Renneboog (2004) investigate European takeover bids an find that cash offers produce an abnormal announcement return to targets that is 3% higher than those offered to abnormal returns to targets in stock offers.

In their research Wansley et al. (1983) investigate the relationship between abnormal returns by looking at the acquisition type and payment method. They find that the shareholder of poorly performing targets earn larger abnormal returns than shareholder of well performing targets in takeover over announcements but do not investigate what the factors are that determining this abnormal return. Furthermore they find that because of tax effects, premium should be larger for cash mergers than for stock mergers. These findings are consistent with earlier literature and can be explained by the fact that, in case of a stock transaction, the capital gains taxes may be deferred until the new securities are sold while in cash mergers, any capital gains are taxed in the year of the acquisition (Gorden and Yagil, 1981; Amihud et al. 1990; Brown and Ryngaert, 1991).

2.6 Acquirer financial constraints

Financial constraints affect the investment and cash policies of some firms. The constraints increase the costs of attracting external capital and thus make the use of internally generated funds, like cash and cash holdings, more attractive (Myers and Majluf, 1984; Greenwald et al., 1984). The financial constraints might lead to underinvestment, which reduces future growth and destroys firm value. To prevent these negative impacts and to fund these expenditures, some companies use the available internal resources (Denis and Sibilkov, 2010; Alshwer, 2011). For financially constrained companies, the changes in capital expenditures are mainly determined by the changes in cash flow. Companies facing financial constraints tend to accumulate and save more cash (Almeida et al., 2004; Denis and Sibilkov, 2010). A company facing

financial constraints has to restrict its investments to the most valuable ones in order to preserve its liquidity and ability to finance future growth opportunities (Almeida et al., 2004). The companies will limit their investment to the most profitable and less risky investments, and have to pass up profitable opportunities (Stein, 2003). This could suggest that financially constrained firms are more careful in making acquisitions and are therefore less likely to overvalue a company.

Existing literature states that the choice between stock and cash payment in M&As depends on the relative benefits and costs of issuing stock and using cash. In case the acquiring company is facing financial constraints this has several consequences for the payment choice. As earlier mentioned, the financial constraints increase the costs of issuing equity and result into a preference for, and higher value of, internal resources including cash and cash holdings. These two distinctions likely generate the main difference in the method of payment for constrained versus unconstrained acquirers. The cost of a payment in stock, and thus the issuance of equity, is higher for constrained firms than for unconstrained. This would suggest that the use of stock payment by a financially constrained acquirer is less likely. However, when the financially constrained acquirer has other available growth opportunities, he will be more likely to save cash to reduce the uncertainty about their future financing of investments. This would indicate that the availability of growth opportunities is positively related to the likelihood of a payment in cash. In case of an unconstrained acquirer, the availability of internal resources does not determine capital investments and thus, growth opportunities should not affect the method of payment (Denis and Sibilkov, 2010; Alshwer et al. 2011). The financially constrained companies are more likely to have many unexploited investment opportunities and growth options. An acquisition can reduce financial constraints if the acquirer’s financial condition and ability to attract capital allow the target firm to engage in more profitable investments (Erel et al. 2014). 2.7 Hypothesis Based on the findings of Ang and Mauck (2011), the takeover premium is expected to be higher in case of a distressed target takeover. When the acquiring company is financially constrained, the company will be more careful in their investments and will only invest in the most profitable and less risky investments (Stein, 2003). The acquirer is therefore

less likely to overpay. The takeover premium is expected to be lower in case the acquiring company is financially constrained. The following hypothesis is constructed: H1: There is a negative relationship between acquirer financial constraints and the

takeover premium in distressed target takeovers.

In case of a financially distressed target, the existing literature suggests that the acquirer is less likely to offer cash due to the information asymmetry problem (Hansen, 1987; Shleifer and Vishny 1992; Martin, 1996; Faccio and Masulis, 2005). The evidence provided by Alschwer (2011) suggests that financially constrained acquirers are more likely to offer cash due to the increase in the cost of issuing equity and their preference to save valuable cash. The existing literature leads to believe that financially constrained acquirers are more likely to pay with stock in a merger or acquisition, and this probability is increased in case the financially constrained acquirer is involved in a distressed target takeover. The following hypothesis is constructed:

H2: There is a positive relationship between acquirer financial constraints and the

likelihood of a stock payment in distressed target takeovers.

In case a company is in a distressed position, it is likely that has unexploited investments and thus has large growth potential. These opportunities might lead to high synergies, if the acquirer is able to turn the distressed position of the target and its assets around. In order to do so, the availability of capital might be beneficial (Bruton et al. 1994; Hotchkiss and Mooradan, 1998). Financially constrained acquirers have less access to capital, and therefore might be less able to exploit the investment and growth

opportunities of the distressed target (Myers and Majluf, 1984; Greenwald et al., 1984).

Therefore, acquisitions in which a non‐distressed acquirer is combined with a distressed target are expected to yield the highest wealth effects for both the target and the acquiring company. The following hypothesis is constructed:

H3: Mergers and acquisitions in which a non‐distressed acquirer is combined with a

distressed target lead to the highest wealth effects for both the target and acquirer shareholders.

3. Research Design

In this paragraph, the conducted research is described. First, the data and sample selection method used in this study are described. Then, the research methodology is discussed and explained in detail.

3.1 Data and sample selection

3.1.1 Data

For this thesis, a sample of U.S. mergers and acquisitions is collected from Thomson One Banker. This database contains an M&A section in which mergers and acquisitions occurring from 1979 and on can be found. This database includes the M&A announcements, deal characteristics including method of payment, takeover premium, deal value, and several financials about the acquiring and target companies. In order to complete the data, the financials found in Thomson One Banker are further completed with data from DataStream and Compustat. In gathering the necessary daily stock returns to calculate the abnormal returns, the CRSP database is used. In this study, the CRSP value‐weighted market index is used as the market proxy and its daily returns are collected from the CRSP database.

3.1.2 Sample selection

A number of data filters are applied: (1) Completed acquisitions are selected and must be announced between 1985 and 2014. The year 1985 is chosen as the starting year because only limited data can be found on the years before. (2) Acquisitions that do not involve U.S. corporate bidders and targets are removed. In addition, acquisitions that involve government owned companies, joint‐ventures and mutual funds are removed from the sample. (3) For complete acquisitions, the bidder needs to own more than 50% of the target after the transaction and has to acquire at least 50%. This way partial or remaining interest in the target is excluded. (4) Self‐tenders, repurchases, recapitalizations, and buybacks are removed from the sample and carve‐outs, spin‐offs, split‐offs, and transactions that are announced to the public after they became effective are dropped. (5) Financial firms (SIC codes 6000 to 6999) and regulated firms (SIC codes 4900 to 4999) are excluded. The final sample consists of 3,889 acquisitions.

3.2 Methodology

3.2.1 Financial constraints and financial distress proxies

In order to measure the financial constraints for the acquiring company the HP‐(or SA‐) index proposed by Hadlock and Pierce (2010) is used. They find that firm size and age are particularly useful predictors of financial constraint levels and they construct a measure of financial constraints that is based on these firm characteristics. The HP‐ index is calculated as follows:

0.737 0.043 0.040

Where Size is defined as the natural logarithm of the acquirer’s total asset capped at $4.5 billion. Firms are sorted into terciles based on their index values. Firms in the top tercile are coded as constrained. In order to test for robustness of the findings the Whited Wu (2006) index of financial constraints is used, which is calculated as follows:

0.091 0.062 0.021 0.044 0.102

0.035

Where is the acquirer’s available cash flow divided by total assets, is a

dummy variable indicating if the company paid dividends to its shareholders, is the

long‐term debt divided by total assets, is the natural logarithm of the total assets,

is the industry sales growth, and is the company’s sales growth. Firms are sorted

into terciles based on their index values. Firms in the top tercile are coded as constrained.

Since there is no universally accepted measure of financial distress, several measures of financial distress are used. Based on work of Ang and Mauck (2011), in order to identify if a target company is financially distressed the following proxies are used: (1) negative income in the last twelve months; (2) two successive years of negative income; (3) negative equity in the last 12 months; (4) the Altman Z‐score (Altman, 1968); and the Ohlson O‐score (Ohlson, 1980). Both the Altman‐score and the Ohslon‐score are designed to estimate the likelihood of bankruptcy. The Altman Z‐score is calculated as follows:

1.2 1.4 3.3 0.6 0.99

Where is the target’s working capital scaled by its total assets, are the target’s

and taxes scaled by its total assets, is the target’s market value of equity scaled by the book value of total liabilities, and are the target’s sales scaled by its total assets. Altman coded a company as distressed when the Z‐score was below a value of 1.81.

The Ohlson O‐score is similar to the Altman Z‐score in the sense that it is meant to predict corporate failure by the event of bankruptcy. Ohlson criticized the model of Altman and used a logit approach that was based on a larger sample than the sample used by Altman. The Ohlson O‐score is calculated as follows:

1.32 0.407 6.03 1.43 0.0757 1.72 2.37 1.83

0.285 0.521

Where are the target’s total assets adjusted for inflation, are the target’s total

liabilities scaled by its total assets, is the target’s working capital scaled by its total assets, is the current ratio, is a dummy variable taking the value of 1 if the total liabilities exceed total assets and 0 otherwise, is the targets net income scaled by the total assets, are the funds from operations scaled by the total assets, is a dummy variable taking the value of 1 if the net income was negative for the last two years and 0

otherwise, and is the ( )/ | | | |), where is the net income.

The value of the Ohslon O‐score is than transformed into a probability using a logistic

transformation. If the probability is exceeding 50%, the target is coded as constrained.  

3.2.2 Financial distress and the takeover premium

In order to test for the relationship between the financial distress and the takeover premium, an OLS regression is used including the different dummy variables for target distress (DISTRESS). In order to calculate the premium the definition of Ang and Mauck (2011) is used. The premium is defined as:

∗ 100

The first independent variable of interest is the dummy variable indicating target financial distress (DISTRESS) which takes on the value of 1 in case the target company is in financially distress. The earlier mentioned financial distress measurements are used in separate regressions. The other variable of interest included is the variable indicating the HP‐index (HPINDEX). In a separate regression the dummy variable for the top tercile of the HP‐index is included (HPINDEXTOP). In order to examine the effect of a takeover

involving a financially constrained acquirer and a distressed target on the takeover premium, interaction terms between the HP‐index Top variable and the financial distress dummies are included. Besides the variables of interest, a number of control variables can be derived to control for other factors affecting the level of takeover premium. Baker et al. (2009), Walkling and Edmister (1985) state that the premium depends on the bargaining power of the parties. For this reason a higher competition, in terms of multiple bidders, leads to a higher takeover premium. For this reason a dummy control variable for multiple bidders is included (COMP). Furthermore, an acquisition involving a hostile takeover is expected to lead to a higher takeover premium. In order to control for this effect dummy variable indicating a hostile takeover (HOSTILE) is included in the regression. The size of a company is expected to influence its negotiation power. Although previous studies were unable to find a relationship between the size of the target firm and the takeover premium, the regression controls for this effect by including the logarithm of the total value of the market capitalization (TSIZE) in the regression. The same is done for the acquiring company (ASIZE). The theory of Grondhalekar et al. (2004) suggests that firms with a higher internal cash generation and low market‐to‐book ratios tend to pay higher premiums. The Free Cash Flow theory of Jensen (1986) supports suggests the findings because these companies generally face fewer available investment opportunities and therefore look for external growth through acquisitions. For the free cash flow available (AFCF), a control variable is included. Prior literature states that premiums should be higher when the transactions is a financed with cash (Huang and Walkling, 1987; Da Silva et al. 2000). A control variable for full cash acquisitions (CASHOFFER) is therefore included. Israel (1991) argues that more leverage leads to more concentrated ownership and therefore it will lead to a higher premium. A control variable for Acquirer firm leverage (ALEVERAGE) is included which is the debt‐to‐equity ratio of the acquirer. The same is done for the target company (TLEVERAGE). Offenberg (2012) finds higher takeover premiums for tender offers. For this reason the control variable indicating a tender offer (TEND) is included in the resgression. In order to control for entity‐fixed effects in premiums, dummy variables for the different industries (INDUSTRY) are included. The following regression model is tested using OLS:

Where indicates the HP‐Index in the first regressions, HP‐index Top in the second regressions, and the HP‐index Top and the interaction terms in the third regressions. In order to check for robustness of the results, the same regression is performed but

instead the WW‐index is used to identify financially constrained acquirers.

3.2.3 Financial distress and the method of payment

In order to test for the relationships between financial distress and the method of payment, two models similar to Faccio and Masulis (2005) are used. These models are the Tobit and Ordered Probit models. Both the models will include a dummy variable indicating if the target company is financially distressed. Furthermore, both the models will include a variables for the HP‐index by which the effect of acquirer financial constraints on the method of payment can be measured. The setup of the Tobit and Ordered Probit will now be described. A. Tobit regression In the Tobit regression the dependent variable is the proportion of cash that is used in the M&A transaction. The proportion is a value between 0 and 100. For this reason a two‐boundary Tobit estimator is used. The following general model is used: ∗ _,

Where indicates the variables of interest, is an independently distributed error term assumed to be normal with zero mean and variance . The dependent variable has both left and right censoring so that:

∗ ∗

0 ∗ _0,

0 ∗ ₁₀₀

100 ∗ ₁₀₀

Where 0 and 100 are the censoring points. The parameters are estimated by the maximization of the log likelihood function.

The first independent variable of interest is the dummy variable indicating target financial distress (DISTRESS) which takes on the value of 1 in case the target company is

in financially distress. The earlier mentioned financial distress measurements are used in separate regressions. The other variable of interest included is the variable indicating the HP‐index (HPINDEX). In a separate regression the dummy variable for the top tercile of the HP‐index is included (HPINDEXTOP). This dummy variable takes on the value of 1 in case the company is in the top tercile of the HP‐index distribution. In order to examine the effect of a takeover involving a financially constrained acquirer and a distressed target on the percentage of cash, interaction terms between the HP‐index Top variable and the financial distress dummies are included. The remaining independent variables in the regression are used to control for deal characteristics and target and bidder firm financial characteristics based on existing literature.

The first control variable is the takeover premium (PREMIUM). Existing literature states that in case of a cash payment in M&As, the premium should be larger due to tax effects (Wansley et al., 1983; Gorden and Yagil, 2001; Amihud et al. 1990; Brown and Ryngaert, 1991). Hansen (1987),Myers and Majluf (1984), Schleifer and Vishny (1992), Martin (1996), and Faccio and Masulis (2005) all find because of asymmetric information the acquiring firm is less likely to offer cash. A high market‐to‐book ratio usually indicates that a company is overvalued. For this reason a control for the market‐

to‐book value (TMTB) of the target company is included. Hansen (1987), Martin (1996),

Faccio and Masulis (2005), and Swieringa and Schauten (2008) all find that debt capacity is positively related to a payment in cash. Since studies use different indicators of debt capacity, for both acquirer firm size (ATA) and leverage (ADEBT) a control variable is included in the regression. Jensen (1986) and Fishman (1989) both find a positive relationship between the available cash to the acquiring company and a payment in cash. In contrast, Faccio and Masulis (2005) find a negative relationship. Since they both find a relationship, to control for the free cash flow to the company a variable indicating the acquirer free cash flow (AFCF) is included. Jung, Kim and Stulz (1996), Martin (1996), and Faccio and Masulis (2005) all find that an acquirer with higher investment opportunities is preferring a payment stock. This indicates a negative relationship between investment opportunities and a payment in cash. Investment opportunity available can be measured by the Tobin’s Q. For this reason the acquirer Tobin’s Q (ATOBINQ) is included as a control variable in the regression. A dummy variable (DIVERSIFY) indicating if both the target and acquirer are in the same industry is included as Faccio and Masulis 2005 found a positive relationship between cross‐

industry M&As and a payment in cash. Grullon et al. (1997), Faccio and Masulis (2005), and Schwieringa and Schouten (2008) all find that there is a positive relationship between the relative deal size and stock transaction. For this reason a control variable for the relative size of the deal (DEALSIZE) is included. Similar to Amit et al. (1989) to control for tender offers and the characteristics of the takeover, dummy variables (TENDER, HOSTILE) are included in the regression. Industry dummies are used to control for entity‐fixed effects. B. Ordered Probit regression In line with Faccio and Masulis (2005), additional to the Tobit Model, an Ordered Probit estimation is used to check for robustness. The benefits of this model are that it allows us to focus on the qualitative decision in payment method. In the Ordered Probit model the dependent variable is 0 for pure cash deals, 1 for mixed stock and cash, and 2 for all stock deals. The same control variables as for the Tobit regression will be used. Although the magnitude of the coefficients in the Ordered Probit model cannot be interpreted directly due to scaling, the sings on the coefficients indicate the relationship with the method of payment. If the coefficient shows a positive sign, a payment in stock is more likely, if the coefficient shows a negative sign, a payment in cash is more likely.

3.2.4 Short‐term M&A performance and its relationship to financial distress and takeover characteristics

First the measurement of the short‐term takeover performance will be explained. Secondly, the regression model testing for the determinants of this performance will be constructed.

A. CAR model

In order to measure the short‐term performance of an acquisition the effects of the M&A on the announcement returns for both the target and the acquirer are examined using an event study. The cumulative abnormal returns (CARs) earned by both parties are measured over different event windows. The model to calculate the CARs looks as follows:

Where , denotes the cumulative abnormal return over the event

window , , where 0 indicates the day of the announcement. ∑ , denotes

the sum of the abnormal returns over the event window. These abnormal returns

( , ) capture the effect of the announcement on the returns earned on the company’s

stock. The following model is used to calculate the abnormal returns:

, , ,

Where , indicates the abnormal returns for stock “i” on day “t”. , denotes the

return for stock “i” on day “t”. The , is the return of the chosen

benchmark for stock “i” on day “t”.

In order to calculate this benchmark return the market model is used. The market model looks as follows:

,

Where denotes the estimated constant, denotes the estimated sensitivity of the

return on stock “i” on day “t”, denotes the return on the market index on day “t”, and denotes the error term of stock “i” on day “t”. In order to get the most accurate estimation of this benchmark model an estimation window is set starting 250 days prior to the takeover announcement until 50 days prior to the announcement. This estimation window will be used to determine the “normal” return the stock would have generated if the takeover announcement would not have taken place. Using OLS the coefficients of the market model can now be determined and the benchmark returns can be calculated. Performing separate regressions for each firm using data within the estimation window and save the intercepts and coefficients of the independent variables and implementing these in the market model, the “normal” return can now be predicted. After calculating these normal returns, the cumulative abnormal returns are calculated over different

event windows: 5, 5 , and 1, 1 . These CARs are then used in an OLS

regression to test for relationships with the variables of interest. The regression model will be discussed in the next subsection.

Additionally, the cumulative average abnormal returns (CAAR) for the total of companies are calculated as follows:

Where , is the CAAR over the event window , , N denotes the number of companies, and ∑ , denotes the sum of the CARs for each stock “i” over the event window , . These CAARs are calculated separately for constrained acquirers versus non‐constrained acquirers; and for financially distressed versus non‐distressed takeovers. Furthermore separate CAARs are calculated for cash offers and stock offers. By calculating these CAARs separately, a t‐test can be performed to test for a significant difference between constrained acquirers versus non‐constrained acquirers and distressed versus non‐distressed takeovers. In order to check for robustness of the results, similar t‐tests on the differences are performed using the WW‐index to identify financially constrained acquirers.

B. CAR Regressions

In order to test for the relationships between the CARs, financial constraints, financial distress, and the takeover characteristics, an OLS regression is used. The independent variable in this model is the CAR for either the acquiring or target company. The variables of interest are the variable for the HP‐index (HPINDEX), the dummy variable indicating if the company is in the top tercile of the HP‐index distribution (HPINDEXTOP), the dummy variable indicating a financially distressed takeover (DISTRESS). In order to examine the effect of a takeover involving a financially constrained acquirer and a distressed target on the takeover CARs, interaction terms between the HP‐index Top variable and the financial distress dummies are included. A number of variables affecting the short‐term performance can be found in the existing literature. These variables are included in the regression to control for other factors that might relate to the CARs. Dennis et al. (2002), Martynova and Renneboog ( 2006), and Freund et al. (2007, 2008) all find a relationship between a diversifying deal and the abnormal returns for both the acquirer and target company. Martynova and Renneboog (2006) further find relationships between the CARs for both parties involved in hostile deals and tender offers. For these reasons a control variables indicating a diversifying deal (DIVERSIFY), hostile takeover (HOSTILE), and tender offer (TEND) are included in the regression. Faccio et al. (2006) find that abnormal returns are higher for cash offers compared to stock offers. For this reason a dummy control variable indicating the percentage of cash is included in the regression (PERCCASH). Servaes (1991), Schwert (2000), and Moeller et al. (2005) Freund et al. (2008) find a relationship between the

CAR and the Tobin’s Q of the acquirer. For this reason this variable (ATOBINQ) is added to the regression. In addition controls for the transaction size relative to the size (RELSIZE) of the bidder and competing bidders (COMP) are included. The following regression model is tested using OLS:

4. Results

In this paragraph, the results from the conducted tests are stated. First, the summary statistics for the takeover characteristics, takeover premium, method of payment, and abnormal returns for the sample will be presented. Then, the regression results for the determinants of the takeover premium and method of payment are stated. Hereafter, the regression results of the abnormal returns for both the target and acquirer will be presented. Next, the results of the t‐tests performed to assess whether there is a difference in the abnormal returns between financially constrained and financially unconstrained acquirers in distressed target takeovers are listed. Finally, the results of the tests using the WW‐index, in order to check for robustness of the results, will be presented.

4.1 Descriptive statistics

Table 1a in the Appendix presents the summary statistics for the different proxies for the different financial constraints (NII1, NNI2, NEQ, ALTMAN, and OHLSON) and financial distress proxies (HPTOP, WWTOP). Looking at the different distress proxies, the portion of financially constrained target companies in the sample varies between 4.84% and 42.27%. The proportion of financially constrained companies indicated by negative equity seems low compared to the other proxies while the proportion indicated by one year of negative income seems rather high. Leaving out these proportions, the sample contains between 21.25% and 39.17% constrained target takeovers. Looking at transactions for which both the acquirer and the target are in a constrained or distressed position the sample contains between 1.60% and 18.06% distressed target takeovers by financially constrained acquirers. Leaving out the proportions of the negative equity and Altman Z‐score, which are relatively low compared to the other proxies, these proportions vary between 11.00% and 18.06% in our sample.

Table 1b in the Appendix presents the correlation coefficient between the variables in the sample. Looking at the proxies for the acquirer financial constraints, the HP‐index and the WW‐index, a high positive correlation is found that is significant at a 1% significance level. These findings are consistent with the findings of Farre‐Mensa and Ljungqvist (2013). All proxies for financial distress are positively correlated with

coefficients varying between 0.14 and 0.95. The correlations coefficients are statistically significant a 1% level. There is a strong, positive correlation between the variable indicating negative income in the last year and the variable indicating negative income for the last two years. This indicates that companies that are experiencing negative income two years prior to the takeover announcement, are likely to experience this negative income the year after. Both the correlation coefficients between the, negative income proxies and the variable indicating financial distress according to the Ohlson O‐ score are positive and around 0.62. The correlation coefficients between the variable indicating a negative equity value and the other proxies are all below 0.33.

Table 2a in the Appendix shows the takeover characteristics by the financial constraints and financial distress proxies. The sample contains 5.22% competing bids, 1.77% hostile takeovers, and 21.37% tender offers. Comparing the proportions by the different proxies of financial distress for the target companies. It is found that all distressed target takeovers in our sample include less competitive bids, hostile takeovers, tender offers, and more diversifying deals compared to non‐distressed target takeovers. The findings on the proportions of hostile takeovers are consistent with the findings of Clark and Ofek (1994). Looking at these proportions for the acquirer financial constraints, it is found that bidding companies that are identified as financially constrained involve fewer competitive bids, hostile takeovers, tender offers, and diversifying takeovers. These findings for the HP‐index are consistent with the findings for the WW‐index.

Looking at the distribution of the method of payment choice in the sample, the proportion of takeovers involving a full stock transaction is the largest with around 45.11%, followed by 41.75% full cash transactions. Comparing the method of payment proportions between distressed and non‐distressed targets it is found that, for all proxies, the percentage of full cash payments is lower in financially distressed target takeovers. These findings are the same for the proportion of takeovers involving a mixture of cash and equity payment, except for the negative equity distress indicator. The percentage of stock payments increases as the target is coded as financially distressed by all proxies, except for the negative equity proxy. When looking at acquirer financial constraints, the acquirers labeled as constrained, on average use more cash and less stock. The percentage of cash payments is lower for both acquirers in the top tercile of the HP‐index and the WW‐index while the percentage of stock is higher. Table

2b in the Appendix takes a more detailed look at the distribution of the method of payment choice by examining differences in method of payment by both target distress and acquirer financial constraints. Panel A looks at the distribution of the method of payment for the whole sample. Panel B looks at the distribution of the method of payment comparing financially constrained with financially unconstrained acquirers. In case the acquirer is financially unconstrained, the proportion of cash used in a distressed target takeover is typically larger in the sample while the proportion of stock and mixed transactions is lower. In case the acquirer is financially constrained, the proportion of stock is typically larger in the sample while the proportion of cash and mixed transactions is lower.

Table 2c in the Appendix shows the summary statistics for both the bidder and target divided grouped according to the different proxies for target distress and acquirer financial constraints. Column 1 reports the summary statistics for the full sample. The following columns show the summary statistics by the proxies for distress and constraints. The highest number of observations is 3,628. Comparing the statistics for the targets labeled as constrained to full sample average it is found that the distressed targets are smaller than their non‐distressed counterparts, in terms of, market value, total assets, net income, and sales. Looking at the statistics for acquirer it appears that the average Tobin’s Q is lower for acquirers facing financial constraints. This could indicate that these acquirers have less investment opportunities than unconstrained acquirers (Jung, Kim, and Stulz 1995; Martin 1996). Furthermore the Debt‐to‐Equity ratio is lower for financially constrained acquirers, this could indicate less concentrated ownership (Israel, 1991). On average the financially constrained acquirers in the sample are smaller in terms of market value, total assets, net income, and sales.

On average, the takeover premium is higher when the target company is financially distressed. This finding is consistent between all proxies for financial distress. This finding is consistent with the evidence found by Ang and Mauck (2010). Furthermore, the average takeover premiums paid by financially constrained acquirers are higher than those paid by unconstrained acquirers in the sample. Table 2d takes a more detailed look at the differences in takeover premiums by examining the differences in the takeover premium for both target distress and acquirer financial constraints. No significant differences between the takeover premiums of constrained and unconstrained acquirers are found. For distressed target companies the takeover

premium on average is higher than for those that are non‐distressed. Three of the five distress proxies show positive significant differences. This indicates that the distressed targets in our sample receive higher premiums than the non‐distressed targets. The differences in premiums for the negative equity and Altman proxies are insignificant.

Table 2e

Average abnormal returns around the takeover announcement Panel A:

CAARs Acquirer

Event Window Mean Q1 Median Q3

[‐5, 5] ‐0.93%(‐3.82)*** ‐5.98% ‐0.84% 3.89% [‐1, 1] ‐0.68%(‐3.57)*** ‐4.06% ‐0.47% 2.60% Panel B: CAARs Target Event Window Q1 Median Q3 [‐5, 5] 25.36%(42.53)*** 7.60% 21.00% 38.05% [‐1, 1] 22.23%(40.02)*** 4.98% 17.39% 33.70% *, **, and *** denote significance at the 1%, 5%,and 10% levels, respectively.

Table 2e shows the average abnormal returns around the takeover announcement earned by the acquiring and target firm. Panel A shows the average abnormal announcement returns for the acquirer for the different event windows. The found abnormal returns for the acquirer are slightly negative and significant at a 1% level. These findings are consistent with the work of Ruback (1983), Jarrel et al. (1988) and Amit et al. (1989). Panel B shows the average abnormal announcement returns for the target companies for the different event windows. Large, statistically significant abnormal returns are found for the target companies. These findings are consistent with the evidence found by Andrade et al. (2001), Georgen and Renneboog (2004), Amit et al. (1989), and Hotchkiss and Mooradian (1998). Table 3b in the Appendix shows the acquirer and target average abnormal announcement returns by the target distress and acquirer financial constraints.

4.2 Regression results

4.2.1 Premium Regression

Table 3a in the Appendix shows the OLS regression results of the takeover premium on the different financial constraints and financial distress proxies and the control variables as described in the methodology paragraph. Due to high correlations with the HP‐index and HP‐index Top variables, the ASIZE variable is dropped from the model.