FINANCIAL CONSTRAINTS IN EUROPEAN
ACQUISTIONS:
THE EFFECT ON BIDDER RETURNS
Gijsbert Veerman University of Groningen
ABSTRACT
This thesis examines the effect of financing constraints on shareholder returns from acquisitions by European bidders over the period 2002-2013. Three measures of financial constraints are used to study its effect on shareholder returns: The size-age index of Hadlock and Pierce (2010), the age-size-cash flow-leverage index from Mulier, Schoors and Merlevede (2016), and dividend payout-ratio (Fazarri, Hubbard and Peterson, 1988). Moreover, the interaction effect of a country’s investor protection and financial constraints on bidder returns is analysed. Results indicate significant positive returns to financially constrained bidders for the size-age index of Hadlock and Pierce (2010). No evidence is found for an interaction effect of financial constraints and strong investor protection on bidder returns.
JEL classifications: G11, G14, G34
Keywords: Mergers & acquisitions, financial constraints, investor protection, event studies
Author: Gijsbert Veerman Mail: g.veerman.1@student.rug.nl Phone: +31627072917
Student number: S2243687
1. INTRODUCTION 3
2. RELATED LITERATURE AND HYPOTHESES 5
2.1 CORPORATE ACQUISITIONS 5
2.2 FINANCIAL CONSTRAINTS 7
2.3 INVESTOR PROTECTION & FINANCIAL CONSTRAINTS 11
3. DATA AND METHODOLOGY 13
3.1 SAMPLE DESCRIPTION 13
3.2 SAMPLE DISTRIBUTION 13
3.3 VARIABLE CONSTRUCTION 15
3.3.1 BIDDER RETURNS 15
3.3.2 MEASURES OF FINANCIAL CONSTRAINTS 16
3.3.3 INVESTOR PROTECTION 17
3.3.4 CONTROL VARIABLES 19
3.4 METHDOLOGY 22
4. EMPIRICAL RESULTS 23
4.1 SUMMARY STATISTICS 23
4.2 FINANCIAL CONSTRAINTS AND BIDDER RETURNS 26
4.3 FINANCIAL CONSTRAINTS, BIDDER RETURNS AND INVESTOR PROTECTION 26 4.4 ROBUSTNESS TESTS 30
4.4.1 NON-PARAMETRIC TEST 30
4.4.2 ALTERNATIVE CLASSIFICATION OF FINANCIAL CONSTAINTS 30
1. INTRODUCTION
“Global M&A value climbs to highest level on record on 2015” (Zephyr global annual M&A activity report 2015)1
In 2015, a record breaking 89,440 deals, worth over 6 trillion USD were announced, according to the Zephyr global annual M&A activity report (2015). Recently, the European Central Bank (ECB) has also announced to extend its ‘quantitative easing program’, with the main goal being to stimulate investments, including corporate investments. As a result, European firms, and more specifically, managers in these firms are even more inclined to increase investments. In order to justify this value of capital circulating in the takeover market, one might expect that that an acquisition is highly beneficial. Otherwise, why would firms make acquisitions in the first place?
By documenting that acquisitions create value for target firms (Bruner, 2002) and for bidder- and target- firms combined (Campa and Hernando, 2004), empirical literature provides this justification. However, evidence on persistent gains to the bidder firm, who make the decision to acquire another firm in the first place, is more ambiguous. Martynova and Renneboog (2008) show that bidder gains from acquisitions are around zero, while Moeller, Schlingemann and Stulz (2005) find that acquisitions resulted in a total loss of 240 billion USD between 1998-2001. Several arguments exist as to why bidding firms incur losses in an acquisition, with one of the most prevalent being that managers in firms with excess cash make acquisitions to realize personal gains, instead of creating value for its shareholders (Jensen, 1986). Based on the assumption that acquisitions do indeed not (always) result in gains to bidding firms, this thesis elaborates on a specific firm-characteristic, for which it argues that it can be a determinant in realizing positive gains from an acquisition: Bidders’ financial constraints.
Financial constraints occur when firms “face a wedge between the internal and external costs of funds” (Kaplan and Zingales, 1997 p.172), in other words, when its cost of capital is higher. Existing literature has related the concept of financial constraints to a variety of attributes in investments. For instance, Fazarri, Hubard and Petersen (1998), show that constrained firms increase their level of investment with the availability of more internal funds, while Almeida, Campello and Weisbach (2004) find that financially constrained firms save more cash out of available cash flows in order to finance future opportunities.
Within the international dimension, another aspect is also closely related to the concept of financial constraints: The quality of the legal- and regulatory protection of a firm’s shareholders, (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 1997) or a country’s investor protection. Investor protection has been documented to improve a firm’s access to capital, reducing the presence of financial constraints (La Porta et al., 1997;1998), to lower a firm’s dependence on cash flows (McLean, Zhang and Zhao, 2012) and to make firms more efficient (Burkart, Gromb, Mueller and Panunzi (2014). This thesis combines these theories on bidders’ gains from acquisitions, financial constraints and investor protection, and examines it in the international context. On first sight, the relation between financial constraints and acquisitions seems contradictory. How (and why-) is it possible that financially constrained firms make acquisitions in the first place?
As the aforementioned numbers indicate, it is a fact that acquisitions are a phenomenon that account for a substantial share in the financial market. Moreover, even financially constrained bidders invest, and make acquisitions. However, they are faced with a higher cost of capital, as a result of the ‘wedge’. To illustrate this wedge, Almeida et al. (2004), show that financially constrained firms forego profitable opportunities in order to preserve its liquidity to be able to fund future growth options. Based on these findings, this thesis develops several arguments for a relation between a bidder’s degree of financial constraints and its returns from acquisitions. Most importantly, it posits that financial constraints have a ‘disciplinary effect’ on investment decisions of bidders, as they are forced to focus only on the most value-enhancing investments. In an attempt to discover the potential disciplinary effect of financing constraints on bidder returns, this thesis aims to give an answer to the following research question:
Do bidders’ financial constraints result in higher returns from acquisitions, and how does investor protection interact in this relationship?
The sample in this study is recent, and includes (listed-) European bidders during the period 2002-2013. Europe is characterized by a variation in legal systems, resulting in different levels of investor protection, and making it possible to examine financial constraints and bidder returns in a dynamic, international setting. Also, it is believed to be one of the first to utilize the recently proposed age-size- cash flow leverage (ASCL) index of Mulier, Schoors and Merlevede (2016) as measure of financial constraints.
Besides this index, it also employs two other measures of financial constraints: The size-age index (SA-index) of Hadlock and Pierce (2010), and a firm’s dividend payout-ratio. (Fazarri et al.,1988). These three measures show different results. The size-age index of Hadlock and Pierce (2010), where constrained firms are smaller and younger, finds a significant positive effect on bidder returns, ranging from 1.2% to 3.2%. The other two however, do not show any significant effect. Hence, the intuition that constraints lead to higher returns form acquisitions is partially supported. Another finding of these difference in results is that, since all three measures should in essence measure the same, the validity and replicability of exiting measures is questionable. As argued in a recent study by Farre-Mensa and Ljungqvist (2015), existing measures do not adequately measure the concept, but rather reflect differences in growth and financing policies of firms. Given that the only significant findings in this study are found for an index based on a firm’s size and age, this thesis also contributes to this debate. This thesis is structured as follows: It starts with providing an overview of related literature in the field of corporate acquisitions, financial constraints and investor protection. Subsequently, the data and methodology employed to examine the relationship between these three variables is described, followed by the empirical results. Additional robustness tests are performed in order to find robust results, followed by the conclusion and discussion.
2. RELATED LITERATURE AND HYPOTHESES
2. 1 CORPORATE ACQUISITIONS
Why do firms make acquisitions, and do they exist in the first place? In general, literature highlights three main value drivers of acquisitions.
Share ownership by managers is suggested to be positively related to bidder returns. (Campa and Hernando, 2004) On the other hand, agency motives (e.g. Berkovitch and Narayanan, 1993) posit that managers of bidders engage in acquisitions to extract value at the expense of its shareholders.
Empirical literature to a large extent agrees that acquisitions create value for target firms’ shareholders (e.g. Bruner, 2002; Georgen and Renneboog, 2004). Bruner (2002) finds that this is the result of the fact that shareholders of target firms receive a premium on their share price in a transaction. In addition, literature also documents positive returns to the bidder- and target firm combined, suggesting that overall, acquisitions create value. Moreover, it justifies the existence of acquisitions in the first place (see for example Mulherin, 2000; Campa and Hernando, 2004). However, Martynova and Renneboog (2008) attribute the overall value creation from an acquisition to the returns to shareholders of target firms, as target firms’ shareholders earn large positive returns and bidders’ shareholders do not lose on average. This conclusion is in line with the hubris hypothesis, as formalized by Roll (1986), which states that acquisitions neither create nor destroy value, but that it redistributes wealth from overbidding acquirers to target shareholders.
Indeed, evidence on returns to bidder firm’s shareholders is more ambiguous. Martynova and Renneboog (2008) find that bidder announcement returns are indistinguishable from zero in their review of M&As during merger waves, Moeller, Schlingemann and Stulz (2005) report that acquiring-firm shareholders lost 12 cents around announcements per dollar spent, with a total loss of $240 billion from 1998-2001, while others also report negative returns, including Walker (2000) and Goergen and Renneboog (2004).
Several reasons have been provided in existing literature to explain the differences and losses in returns to bidding firms’ shareholders when an acquisition is announced.
The size effect may be a result of the fact that the incentives of managers in small firms are more aligned with that of its shareholders compared to that of larger firms. Moreover, large firms offer higher acquisition premiums than smaller firms, in turn resulting in lower returns. Overall, literature finds inconclusive results on returns from acquisitions to shareholders of bidding firms, with the tendency to conclude that returns to bidder firms’ shareholder are insignificant from zero (Martynova and Renneboog, 2008) or negative (Moeler et al., 2004). The next section introduces a specific firm characteristic as a determinant that potentially affects bidder firms’ shareholder returns.
2.2 FINANCIAL CONSTRAINTS
Financing constraints occur when firms “face a wedge between the internal and external costs of funds” (Kaplan and Zingales, 1997 p.172) or when “frictions prevent the firm from funding all desired investment opportunities” (Lamont, Polk and Saá-Requejo, 2001, p. 529). But what makes the concept of financing constraints relevant in the first place?
In a world with perfect capital markets, firms can fund all value-increasing investment opportunities. In this situation, a firm’s cost of external capital equals that of internally generated funds, making investment decisions irrelevant (Modigliani and Miller, 1958). However, Myers and Majluf (1984) and Greenwald, Stiglitz and Weiss (1984) show that information asymmetry between corporate insiders and providers of external capital invalidate the perfect capital market theory, which results in capital market frictions. With the presence of capital market frictions, the cost of external capital relative to that of internal capital increases. In this situation, investment decisions depend on the availability of internal- and external capital, in the form of equity and debt, which imposes constraints on firms’ investment decisions (Fazarri and Athey, 1987). In other words, a firm’s level of investment is not only determined by the availability of positive NPV projects, but also by the availability of internal- and external- capital (Fazarri, Hubbard and Peterson, 1988).
Table 1. Overview of financial constraints measures
Authors (year)
Time-period Sample Main variable(s) Justification/ index
Fazarri et al. (1988) 1970-1984
422 U.S. manufacturing firms Dividend payout-ratio Dividends and investments are two alternative uses of funds. Financially constrained firms have lower payout-ratios in order to retain funds for future investments and show a higher sensitivity of investment to cash flow
Deveruex & Schiantarelli (1990)
1969-1986
720 U.K manufacturing firms Size, Age, Sector Large, mature firms are less likely to face information asymmetry because they are better known by the market. Sectors with high growth rates have better access to external financing
Hoshi, Kashyap and Sharfstein (1991)
1965-1986
245 Japanese manufacturing firms Group membership Unconstrained firms are closely related to banks, which serve as their primary source of external financing. Constrained firms without close ties face difficulties in raising finance.
Kashyap, Lamont and Stein (1994)
1980-1985
933 U.S. manufacturing firms Bond rating Small firms with low liquidity have limited access to the debt market, resulting from the lack of sufficient collateral to secure debt.
Gilchrist and Himmelberg (1995)
1979-1989
428 U.S. Manufacturing firms Commercial- and bond paper rating
A debt rating reflects the assessment by the market of a firm’s credit quality and indicates its ability to repay debt. Firms with debt and without a S&P rating are classified as financially constrained while firms with a rating are considered unconstrained.
Kaplan and Zingales (1997) 1970-1984
49 low-dividend paying manufacturing firms
Qualitative financial statement analysis
Financially constrained firms based on payout-ratio and investment-cash flow sensitivity do not show a monotonic relationship (criticizing Fazarri et al. (1988))
Faccio and Masulis (2005) 1997-2000
1349 EU firms Leverage Funds are primarily obtained by issuing new debt, hence highly levered firms are constrained in their ability to issue more, new debt.
Erickson & Whited (2000) 1992-1995
737 U.S. manufacturing firms Tobin's Q Financially constrained firms have higher Tobin’s Qs, because these firms are able to fund fewer positive NPV projects. Therefore, they are only able to invest in the most profitable projects leading to higher return on investments and ultimately in a higher Tobin’s Q. Almeida et al. (2004)
1971-2000
29 954 firm-years in the manufacturing industry
Payout-ratio, size, bond-
commercial paper rating, Financially constrained firms have higher cash holdings in order to finance future investment opportunities. Unconstrained firms have no use for cash as they do not face restricted access to external capital
Lamont, Polk & Saá-Requejo (2001)
1968-1997
Avg. sample of 1059 in the U.S manufacturing industry
Cash flow, Tobin's Q, leverage, dividends, cash holdings
KZ index = −1.002*cash flow + 0.283*Tobin’s Q + 3.139*leverage −39.368*dividends − 1.315*cash holdings
Whited & Wu (2006) 1975-2001
between 131 -1390 non-financial U.S. Firms
Cash flow, leverage, size, industry sales growth, firm sales growth
WW index = -0.091*cash flow + 0.021*leverage - 0.044*(ln)size + 0.102*industry sales growth - 0.035*firm sales growth
Hadlock & Pierce (2010) 1995-2004
356 non-financial U.S. Firms Size, age
SA index = (-0.737 * Size) + (0.043 * Size2 ) − (0.040* Age)
Mulier, Schoors, Merlevede (2016)
1996-2008
between 7443-404366 EU firms from 6 countries
Size, age, cash flow, leverage
Fazarri et al. (1998) are considered one of the first to empirically test financial constraints, and base financial constraints on a firm’s payout-ratio. Dividends and investments are two alternative uses of funds, and as such financially constrained firms have lower payout-ratios, in order to retain funds to finance future investments. Subsequently, they test a firm’s sensitivity of investment to cash flow, where constrained firms show a higher sensitivity. Constrained firms increase their level of investment with the availability of more internal funds, as they are forced to choose among investments and face a higher cost of external finance due to capital market frictions. Unconstrained firms are not primarily affected by the availability of internal funds, and can more easily raise debt, resulting in a lower cost of capital. Their investment choices are thus less sensitive to its cash flows. A number of subsequent empirical studies find results in line with this finding. (e.g. Whited, 1992; Bond and Meghir, 1994)
Other studies use firm size and age as proxy for financial constraints (e.g. Deveruex and Schiantarelli, 1990; Gertler and Gilchrist, 1994). Large firms are better known to the market and have assets that can serve as collateral to secure debt, and can thus easily obtain debt. Older firms are also better known to the market and face less information asymmetry, making them less prone to financial constraints. In contrast, small and younger firms less creditable, less known to the market and face more information asymmetry, resulting in financing constraints.
Almeida et al. (2004) suggest that financially constrained firms save more cash out of cash flow, which they label the cash flow sensitivity of cash. A major advantage of a liquid balance sheet is that it allows managers of firms to undertake value-enhancing project when they arise. Hence, when firms face financing constraints, cash holdings should be higher in order to ensure that the most profitable investment opportunities are continued to be financed, which reflects the precautionary motive of holding cash. (Opler, Pinkowitz, Stulz and Williamson, 1999). Unconstrained firms however, able to easily secure debt, have no use of holding cash and can fund positive NPV investments.
as a result of endogeneity. To illustrate this issue, they argue that an exogenous increase in cash may lower a firm’s financial constraints, but that on the other hand it is possible that it holds cash for precautionary reasons, indicating that the firm is constrained in the first place. In a response to these shortcomings, they propose that the only variables adequately reflecting financial constraints are firm size and age, where financially constrained firms are smaller and younger. Based on these componnents, they construct the SA-index. When comparing their index to the most widely used KZ-index (Kaplan and Zingales, 1997; Lamont et al., 2001) they find that the correlation is close to zero, casting doubt on the validity of this measure. The moderate correlation with another popular index, the WW-index (Whited and Wu, 2006) is dedicated to the fact that it also includes firm size as one of its components. Mulier et al. (2016) also relate their ASCL-index to the existing three indices and show that it is moderately correlated with the WW- and SA-index and even negatively correlated with the KZ-index. To conclude the discussion, Farre-Mensa and Ljungqvist (2015) evaluate the extent to which existing measures of financial constraints actually measure financial constraints. In a natural experiment, where they exogenously increase taxes to increase a firms’ demand for debt, they show that constrained firms have no difficulty in obtaining credit. They conclude that existing indices do not adequately measure financing constraints, but rather reflect differences in growth and financing policies of firms.
As the discussion above shows, firms’ financing constraints affect a variety of attributes of investments. Several arguments exist why one of these attributes includes bidder returns from acquisitions.
As a result of being financially constrained, bidder firms cannot fund all value-enhancing investments. In fact, Almeida et al. (2004) observe that financially constrained firms forego profitable opportunities in order to preserve its liquidity in order to be able to fund future growth options. Here, the intuition follows that the investments that constrained bidders do undertake are only justified when these are perceived to result in high returns, or result in higher returns than other potentially profitable opportunities. In other words, being constrained forces a bidder to limit its investments to only the most profitable. In this sense, financial constraints can have a ‘disciplinary effect’. In a similar vein, as a result of market frictions, the cost of capital is higher for a financially constrained bidder, who faces a ‘wedge’ between internal- and external capital (Kaplan and Zingales, 1997). This requires constrained bidders to carefully make a trade-off between the perceived added value of an acquisition and the costs associated with it.
Jensen (1986) illustrates this with his free cash flow hypothesis, which posits that as a result of excess cash and agency motives, managers engage in empire building and the realization of personal gain, which results in value-destroying acquisitions.
Several additional arguments for higher bidder returns follow by assuming that the proposed measures capturing financial constraints are valid, the most important being firm size. Several authors, including Deveruex & Schiantarelli (1990) and Hadlock and Pierce (2010) find that financially constrained firms are smaller than unconstrained firms. The size effect (Moeller et al.,2004) in turn suggests that acquisitions of small bidders are profitable and acquisitions by large bidders result in losses. Similarly, Faccio and Masulis (2005) argue that financially constrained firms have higher leverage ratios, while Jensen (1986) finds that debt makes managers more efficient as providers of external finance closely monitor managers’ actions.
Based on these arguments, the following is hypothesized:
H1: Returns from acquisitions are higher when a bidder is financially constrained
2.3 INVESTOR PROTECTION AND FINANCIAL CONSTRAINTS
This section introduces another, country-level, determinant that literature has documented to affect bidder returns from acquisitions: The quality of the legal- and regulatory protection of a firm’s shareholders, a country’s investor protection (La Porta, Lopez-de-Silanes, Shleifer and Vishny, 1997). It also develops arguments for an interaction between investor protection and financial constraints on bidder returns.
Supportive arguments related to agency motives are given by Dyck and Zingales (2004), who find lower private benefits of control in countries with strong investor protection. Higher dividend payouts in countries with strong investor protection, satisfying firm shareholders, are reported by La Porta et al. (2000).
Besides affecting bidder returns, investor protection is also documented to relate to financing constraints in various ways. Most important in this regard is that investor protection improves firms’ access to external finance and reduces the costs of external capital, mainly as a result of better developed stock markets. (e.g. La Porta et al., 1997; 1998) Leuz, Nanda and Wysocki (2003) argue that strong investor protection encourages more accurate financial reporting, resulting in stock prices that more accurately reflect a firm’s fundamental value. In turn, this implies that information asymmetry between providers of external finance and the firm is reduced, so that external investors are more willing to provide capital. La Porta et al. (1997) show that countries with stronger investor protection have larger capital markets and more IPOs, indicating a higher presence of external finance providers and making it easier to raise external capital. In other words, empirical research finds that stronger investor protection reduces financing constraints.
McLean, Zhang and Zhao (2012) study the relation between a firm’s sensitivity of investment to cash flow (Fazarri et al.,1988), and investor protection. They find that investments of firms in countries with strong investor protection are less dependent on cash flows. Because investor protection reduces the cost of external finance, low cash flow firms can easily raise external capital for investments instead of relying on internal funds. Burkart, Gromb, Mueller and Panunzi (2014) examine the role of investor protection on the efficiency of acquisitions. In their model, stronger legal investor protection increases a bidders’ outside funding capacity (i.e. reduces financing constraints) and limits the ease with which the bidder can expropriate firm resources. They find that, under effective bidding competition for a target, stronger investor protection makes it less likely that efficient financially constrained bidders are outbid by less efficient unconstrained rivals.
Arguments from literature as described here indicate that investor protection reduces financing constraints (La Porta et al.,1997;1998; McLean et al.,2012). Investor protection also reduces agency costs and makes it more likely that managers make value-enhancing investments. (Shleifer and Vishny, 1997). Combining these arguments with that of the previous section, where it is proposed that bidders’ financial constraints positively affect returns from acquisitions, leads to the following hypothesis:
H2: Returns from acquisitions by financially constrained bidders are higher in countries with strong
3. DATA AND METHODOLOGY
3.1 SAMPLE DESCRIPTION
Deal level data to examine the effect of bidders’ financial constraints on returns from acquisitions is extracted from Zephyr, a database provided by Bureau van Dijk. Retrieved data from Zephyr includes identifiers (ISIN codes), country- and industry classifications- of bidders and targets, deal value, the announcement date, method of payment and whether the target firm is private or public. Firm level, financial data is retrieved from Thomson Reuter’s financial database Datastream, including Worldscope, by matching firm identifiers obtained through Zephyr.
A sample of 875 acquisitions by publicly listed European bidders between 2002-2013 is identified that meet the following criteria: (1) The current deal status is completed, (2) acquirers possess less than 50% of target firms’ shares and more than 50% after the acquisition, (3) the minimum deal value is €1 million, (4) targets’ minimum total assets are €1 million, (5) targets are independent firms (excluding acquisitions of subsidiaries of other firms), (6) bidder and target are not in the financial- (SIC 6000-6999) and regulated utility- (SIC 4909-4939) industries, and (7) bidders have non-missing financial data for every variable in Datastream and Worldscope. Next, (8) countries and industries with less than 10 acquisitions are excluded, in order to be able to generalize findings. To conclude, (9) bidders do not have any other acquisitions in the estimation window (t=-255/ t=-46).
3.1.2 SAMPLE DISTRIBUTION
Table 2. Acquisitions by country and measures of financial constraints.
All Acquisitions SA-index ASCL-index Payout-ratio
Acquirer Country N % FC NFC FC NFC FC NFC Belgium 10 1.1% 4 6 6 4 6 4 Finland 59 6.7% 15 44 17 42 7 52 France 69 7.9% 13 56 13 56 28 41 Germany 45 5.1% 7 38 13 32 18 27 Ireland 20 2.3% 1 19 4 16 7 13 Italy 10 1.1% 3 7 5 5 7 3 Norway 17 1.9% 8 9 7 10 7 10 Poland 28 3.2% 17 11 12 16 15 13 Spain 35 4.0% 4 31 15 20 12 23 Sweden 61 7.0% 28 33 23 38 26 35 Switzerland 29 3.3% 2 27 7 22 5 24 The Netherlands 32 3.7% 5 27 6 26 9 23 United Kingdom 460 52.6% 156 304 157 303 116 344 Total 875 100% 263 612 285 590 263 612
Note: FC denotes financially constrained bidders and NFC denotes non-financially constrained bidders
Table 2 reports the distribution of deals by year and measure of financial constraints. Years with the lowest number of acquisitions are 2002 (46) and 2003 (43). Most deals, 119 (13.6% of the total sample) are conducted in 2007, followed by a decline in 2008 (76; 8.7%) and 2009 (59; 6.7%), reflecting the impact of the financial crisis on deal volume. 2
Table 3. Acquisitions by year and measures of financial constraints.
All Acquisitions SA-index ASCL-index Payout-ratio
Year N % FC NFC FC NFC FC NFC 2002 46 5.3% 15 31 19 27 10 36 2003 42 4.8% 11 31 12 30 11 31 2004 43 4.9% 10 33 15 28 13 30 2005 72 8.2% 21 51 22 50 24 48 2006 90 10.3% 34 56 27 63 33 57 2007 119 13.6% 35 84 44 75 42 77 2008 76 8.7% 21 55 21 55 13 63 2009 59 6.7% 21 38 19 40 23 36 2010 84 9.6% 22 62 29 55 21 63 2011 81 9.3% 25 56 26 55 27 54 2012 88 10.1% 25 63 29 59 26 62 2013 75 8.6% 23 52 22 53 20 55 Total 875 100% 263 612 285 590 263 612
Note: FC denotes financially constrained bidders and NFC denotes non-financially constrained bidders
3.2 VARIABLE CONSTRUCTION
In the next three subsections, the measurement of four categories of variables is discussed: Bidder returns as the dependent variable, measures of financial constraints as the key explanatory variables, investor protection as interaction variable, and deal- and bidder specific characteristics as the control variables.
3.2.1 BIDDER RETURNS
Bidder returns are evaluated from a shareholder’s perspective, as shareholders are the residual owner of the firm. Therefore, shareholder wealth maximization should be the primary objective managers, justifying the shareholder’s perspective as an efficient evaluation criterion to study gains from acquisitions ((Martynova and Renneboog, 2008). The standard event study method (Brown and Warner, 1985) is employed to study share price reactions to the announcement of an acquisition (i.e. returns from acquisitions). The event study approach assumes that the market is efficient (Fama, 1970) where share prices reflect information, including short-term wealth effects of the announcement of an acquisition3. Announcement effects, or abnormal returns, reflect the difference in share price return
with the effect of the acquisition announcement and the ‘normal’, benchmark return, without the effect of the acquisition announcement. The market model is used to compute normal returns, which takes the form of:
(1) Where
R
it is the return of share i on day t andR
mt is the benchmark return on day t. For the benchmark return, the bidders’ home country stock market index is used. Parameters α and β are the estimated regression parameters by an ordinary least squares (OLS) regression and ε is the error term. Parameters α and β are estimated using an estimation window from day -255 to day -46 [-255, -46] before acquisition announcement, as in John et al. (2010)4 Following MacKinlay (1997), the abnormalreturn, i.e. the excess return in bidders’ share price, as a result of the acquisition announcement is then computed as:
(2) Where
AR
it is the abnormal return of share i on day t,R
it the benchmark return from equation 1 and and
ˆ
i and ˆi are the estimated parameters from equation 1.
3 Short-term wealth effects are examined, because in the long-term the unique effect of an acquisition is difficult
to isolate. Also, a number of methodological issues arise when employing an event study for the long term. (for an overview, see Barber and Lyon, 1997)
4 Schwert (1996) empirically shows that a potential price run-up in share prices resulting from rumors and news
about an acquisition is mostly manifested after the 42nd day prior to the announcement. By excluding this period
from the estimation window this does not affect its estimation.
Subsequently, 5-day cumulated abnormal returns (CARs) are computed [-2, +2], where t=0 is the day of the event, the acquisition of the announcement. Cumulating the abnormal return for multiple days accounts for potentially incorrect announcement dates in Zephyr, and possible information leakage of the acquisition before or after day 0. Therefore, using a 5-day event window should capture most (all) of the announcement effect. CAR is denoted as:
CAR
=
2 2 t it tAR
(3)If the day of the acquisition announcement is a non-trading day, the first trading day after the announcement is considered as day 0, the day of the event.
3.2.3 MEASURES OF FINANCIAL CONSTRAINTS
Three measures are used to evaluate bidders’ financial constraints: The size-age-index (Hadlock and Pierce, 2010), the age-size cash flow-leverage index (Mulier et al., 2016) and payout-ratio (Fazarri et al., 1988).
The size-age index (SA-index) (Hadlock and Pierce, 2010) bases the level of financial constraints on a firm’s size and age. In line with Hadlock and Pierce (2010), it is constructed as:
SA-index = -0.737 * Size + 0.043 * Size² - 0.040 * Age (4)
Size is the average of the natural logarithm of inflation-adjusted book value of assets in euro, at the year of the acquisition, t=0, and the year before the acquisition, t=-1). Age is the length of time in years since the bidder had a non-missing share price in Worldscope in the year of the acquisition (t=0). Size is winsorized at the (natural logarithm-) of €4.5 billion and age is winsorized at 37 years. Similar to Hadlock and Pierce (2010), size is inflation adjusted to the 2004 consumer price index level of the bidders’ home country. Based on their ranks in the SA-index, firms are then ranked and sorted into deciles. A dummy variable is constructed, where firms in the bottom three deciles are assigned to the financially constrained group (FC) and equal one, and firms in the remaining deciles are assigned to the non-financially constrained group (NFC) and equal zero.4
The age-size-cash flow-leverage (ASCL)- index of Mulier et al. (2016) also incorporates a firm’s cash flow and leverage ratio. In this index, age is a firm’s year of incorporation +1 at the year of the acquisition (t= 0) and size is average of total assets in million euro in the year of the acquisition (t=0) and the year before (t=- 1). Cash flow ratio is the average cash flow to tangible fixed assets in the year of the acquisition (t=0) and the year before (t= -1). A firm’s leverage ratio is average long-term debt to
4 From here on, the denotation of ‘FC’ refers to financially constrained bidders and ‘NFC’ to non-financially
total assets at t= 0 and t= -1. Thereafter, values of every firm for each component are related to the industry median of the sample. A firm receives a score of 1 if it is younger in age, smaller in terms of total assets, and if it has a lower cash flow ratio than the industry median, and 0 otherwise. For the leverage ratio, it receives a 1 if it is higher than the median and 0 otherwise. Scores are then summed, resulting in an index from 0 (strongest unconstrained group) to 4 (strongest constrained group). Bidders with a score of 3 and 4 are assigned to the FC group and take on a value of 1 with the construction of a dummy variable. Firms with values of 0, 1 or 2 on the ASCL index are assigned to the NFC group and equal 0.
Payout-ratio follows from Fazarri et al. (1998). The original methodology employed by Fazarri et al. computes the payout-ratio as dividends to operating income and constructs three groups. Firms with a ratio of less than 0.1 are low dividend-paying firms, and are thus financially constrained, the second group includes firms with ratios between 0.1 and 0.2, and the third group are unconstrained firms with high payout-ratios above 0.2. However, in line with recent literature (e.g. Denis and Sibilkov, 2010; Almeida et al., 2004) and similar to the construction of the SA-index, here firms are sorted in deciles. Firms in the bottom three deciles are assigned to the FC group and the rest to the NFC group. A dummy variable is created, where FC firms equal one and NFC equal zero. Bidders’ payout-ratio is defined as:
𝑃𝑎𝑦𝑜𝑢𝑡 − 𝑟𝑎𝑡𝑖𝑜 = 𝐷𝑖𝑣𝑖𝑑𝑒𝑛𝑑𝑠+𝐶𝑜𝑚𝑚𝑜𝑛 𝑠𝑡𝑜𝑐𝑘 𝑟𝑒𝑝𝑢𝑟𝑐ℎ𝑎𝑠𝑒𝑠
𝑂𝑝𝑒𝑟𝑎𝑡𝑖𝑛𝑔 𝑖𝑛𝑐𝑜𝑚𝑒 (5)
Firms with positive dividends plus common stock repurchases and negative operating income are assigned to the NFC group. Similar to the construction of the SA-index and the ASCL-index, a firm’s payout-ratio is the average of its value in the year of the acquisition, at t=0, and the year before the acquisition, t=-1.
3.3.3 INVESTOR PROTECTION
its components reflect favorable requirements for shareholders. In theory, the index can take a value up to six (strongest investor protection), but has a maximum of five in the revised index by Djankov et al. (2008). The mean in their index is 3.37 and the median 5, with higher values indicating stronger investor protection. Table 3 shows the classification of investor protection of included countries in the sample. The mean (median) value of the sample is 3.46 (3.5). Countries are assigned to the group of countries with strong (weak) investor protection if their median is higher (equal to- or lower-) than the median of the sample.
In addition, the anti-self-dealing index from Djankov et al. (2008) is used as alternative measure of investor protection, which was developed to address the specific, ad hoc nature of the anti-director rights index. The anti-self-dealing index is based on a hypothetical transaction reflecting a conflict of interest by an insider with a majority of share- holdings. Components of the index, classified either as ex ante- or ex post- control of self-dealing, include the approval of disinterested shareholders and the requirement of disclosure to a transaction. Values of the anti-self-dealing index range from zero to one. For the 72 countries included in the revised index of Djankov et al. (2008) the mean (median) is 0.44 (0.42). Table 3 shows that the mean (median) of the sample in this study is 0.44 (0.38). Again, countries with an equal- and lower- median are assigned to the group of weakly investor protected countries and countries with a higher median to the group of countries with strong investor protection.
Table 4. Investor protection per country
Anti-director rights index Anti-self-dealing index Acquirer country Value Classification Value Classification
Belgium 3 low 0.54 high
Finland 3.5 low 0.46 high
France 3.5 low 0.38 low
Germany 3.5 low 0.28 low
Ireland 5 high 0.79 high
Italy 2 low 0.42 low
Norway 3.5 low 0.42 low
Poland 2 low 0.29 low
Spain 5 high 0.37 low
Sweden 3.5 low 0.33 low
Switzerland 3 low 0.27 low
The Netherlands 2.5 low 0.20 low
United Kingdom 5 high 0.95 high
mean 3.46 0.44
3.2.4 CONTROL VARIABLES
Several variables are included to control for other factors that literature has documented to affect bidder returns from acquisitions.
Bidders’ Tobin’s Q is included as control variable on firm level. It is computed as the ratio of the bidders’ market value of assets over its book value of assets. Prior studies find inconclusive results on the effect of Tobin’s Q on CAR. Lang, Stulz and Walkling (1991) interpret a higher Tobin’s q as indication of growth opportunities, resulting in higher CARs. Moeller, Schlingemann and Stulz (2004) document a negative relationship between Tobin’s Q and CAR and argue that it is a proxy for overvaluation.
Domestic acquisitions are documented to result in higher returns than cross-border acquisitions. Erel, Liao and Weisbach (2012) argue that this is the result of geographical- and cultural distance, leading to lower bidder returns. Bris (2008) dedicates lower returns from cross-border acquisitions to asymmetric information problems when valuing target firms. To account for this effect, a dummy variable is created with a value of one for domestic acquisitions and a zero for cross-border acquisitions. Based on the arguments by Erel et al. (2012) and Bris (2008), domestic acquisitions are expected to have a positive effect on CAR.
The industry dummy variable is set to one if bidder and target share the first two-digit SIC code and a zero otherwise. Industry related acquisitions are perceived to create more value for bidders than industry diversifying acquisitions. Industry related acquisitions can benefit from the exploitation of strategic synergies (Doukas, Holmen and Travlos, 2002). Morck, Shleifer and Vishny (1990) find that diversifying acquisitions destroy shareholder value. Hence, the coefficient of the industry dummy is expected to be positive.
The acquisition of private targets is consistently documented to result in positive bidder returns, whereas acquisitions of public targets result in lower, or negative returns (Fuller, Netter and Stegemoller, 2002). Officer (2007) argues that bidders of private targets receive a liquidity discount, because private targets cannot be acquired and sold as easily as public firms. To capture this effect, a dummy variable is created with a value of one when the target is private and a zero otherwise, and is expected to be positive.
All shares equals one for acquisitions financed with shares and zero otherwise, and is expected to have a negative effect on bidder CAR.
Relative deal size is controlled for, as Asquith, Bruner and Mullins (1983) and Moeller et al. (2004) find that bidder announcement returns increase with the relative size of the deal. It is computed by dividing the deal value in euro by bidders’ average total assets at t=0 and t=-1. Its effect is expected to positively related to bidders’ CAR.
Target size is included as the last control variable, which is the natural logarithm of the deal value in euro, as in Fuller, Netter, and Stegemoller (2002). They argue that larger firms are more difficult to integrate, resulting in lower returns. Contrary, Rossi and Volpin (2004) show that larger deals are associated with lower premiums to target shareholders, leading to higher returns to bidders. Officer (2007) differentiates between private- and public- firms for the effect of target size on return. His liquidity discount hypothesis predicts that the liquidity discount is higher for large, private firms, leading to higher returns. Based on mixed findings on the effect of target size, there are no ex ante expectations on the effect of target size on bidder returns.
Existing literature suggests that several other firm characteristics affect bidder returns from acquisitions. Jensen (1986) documents a positive relationship between leverage and returns, because it makes managers more efficient. Moeller et al. (2004) find that firm size negatively affects bidder returns. However, because these variables are a component of one of the employed measures of financial constraints, they are not included as control variables in order to avoid endogeneity in the form of simultaneous causality.
Table 5. Correlation matrix. Definitions of variables in table A1 in the appendix
Variable CAR SA- Index ASCL -index Payout
-ratio Tobin's Q Domestic Industry Private All
cash All shares
Relative
size Target size
The observed correlations here show that, to a large extent, bidders assigned to the financially constrained group in one of the measures can be classified as unconstrained in another. Consequently, the discussion in literature about the validity and merits of each measure seems justified. Moreover, it might reflect some preliminary indications of different results for the included measures of constraints. 3.3 METHODOLOGY
To examine the effect of bidders’ financial constraints and the interaction of financial constraints and investor protection on shareholder returns from acquisitions, the following regression is estimated:
Returnsi =
+
1FCi +
2 IP i +
3 IPi *FCi
N k i k kX 1 ,
+
i (6)where Returnsi are the bidders’ shareholder returns from acquisition i, in 5-day cumulated abnormal return (CAR);
is the intercept, FCi is the main explanatory dummy variable and is equal to 1 if the bidder is classified as financially constrained in acquisition i and 0 otherwise; IPi is a dummy variable equal to 1 if in acquisition i the bidder is located in a country with stronger investor protection and a 0 otherwise;X ,
k i denote the vector of control variables (bidders’ Tobin’s Q, domestic acquisitions, industry-focused acquisitions, private acquisitions, whether the acquisition is financed with all cash or all shares, the relative size of the bidder to the target and the target size). First, the regression is estimated without the interaction effect of financial constraints and investor protection, to examine the single effect of financial constraints on bidder returns (hypothesis 1). Thereafter, results are reported with the interaction effect.4. EMPERICAL RESULTS
4.1 SUMMARY STATISTICS
Table 6 presents summary statistics of the sample, and differences in mean values between financially constrained- and unconstrained bidders for each measure. At firm level, the table shows consistent significant differences on all measures between FC bidders and NFC bidders for total assets and age. FC bidders are significantly younger and smaller in terms of total assets. For instance, based on the ASCL-index, the average age of a FC bidder is 21 years and average total assets are €1106 million,
whereas NFC bidders’ average age is around 51 year with total assets of €6205 million. These results are in line with previous findings (e.g. Gertler and Gilchrist, 1994) and confirm that FC firms typically are younger and smaller. It also provides some justification for the use of the SA-index of Hadlock and Pierce (2010), which is loaded on firm size and age. Tobin’s Q and leverage show no stable and significant differences across measures of constraints, lending no support to arguments that FC bidders have unexploited investment opportunities (Erickson & Whited, 2000) or have a higher leverage ratio (Faccio and Masulis, 2005). FC bidders tend to hoard more cash in order to be able to future finance investment opportunities (Almeida et al.,2004), which is indicated by significantly higher cash holdings for the SA-index and payout-ratio.
Table 6. Summary statistics
All acquisitions SA-index ASCL-index Payout-ratio
FC NFC p-value FC NFC p-value FC NFC p-value
Company level Total Assets (€m) 4544.240 65.396 6468.972 0.000 1106.076 6205.048 0.000 787.894 6158.486 0.000 Age 41.219 31.010 45.606 0.000 21.014 50.980 0.000 23.255 48.940 0.000 Tobin's Q 1.971 2.519 1.735 0.249 1.694 2.104 0.183 2.446 1.766 0.317 Leverage 0.194 0.129 0.222 0.000 0.228 0.178 0.000 0.183 0.199 0.178 Cash holdings 0.137 0.198 0.111 0.000 0.141 0.136 0.625 0.195 0.113 0.000 No. of observations 875 263 612 285 590 263 612 Deal level Deal value (€m) 309.269 10.347 437.727 0.000 147.972 387.183 0.084 102.170 398.267 0.012 Industry focused (%) 71.771 71.483 71.895 0.901 72.982 71.186 0.578 75.285 70.261 0.122 Domestic (%) 56.914 73.764 49.673 0.000 66.316 52.373 0.000 58.555 56.209 0.520 All cash (%) 50.857 47.909 52.124 0.254 50.526 51.017 0.892 43.726 53.922 0.006 All shares (%) 13.943 23.574 9.804 0.000 22.807 9.661 0.000 30.418 6.863 0.000 Private (%) 82.629 95.057 77.288 0.000 87.719 80.169 0.003 81.749 83.007 0.657 Relative size 0.193 0.389 0.109 0.000 0.316 0.134 0.001 0.397 0.106 0.000 Target size 16.875 15.578 17.432 0.000 16.341 17.133 0.000 16.409 17.075 0.000 No. of observations 875 263 612 285 590 263 612 High investor protection
Anti-director index (%) 58.857 61.217 57.843 0.351 61.754 57.458 0.224 51.331 62.092 0.003
Anti-self-dealing index (%) 62.743 66.920 60.948 0.090 64.561 61.864 0.438 51.711 67.484 0.000
No. of observations 875 263 612 285 590 263 612
Because FC bidders are constrained as a result of low internal funds in the first place, and because they cannot easily raise debt, they are forced to resort to the most expensive from of financing an acquisition: by issuing equity, as table 5 indicates.
The variables related to investor protection do not show stable significant differences across measures of financial constraints, except for payout-ratio. Here, 62% and 67% of NFC bidders are located in countries with strong investor protection based on the anti-director- and anti-self-dealing index respectively. This is in line with arguments by La Porta et al. (1997;1998) and Leuz et al. (2003), who argue that stronger investor protection improves firms’ access to external capital.
Table 7 shows summary statistics for the dependent variable, bidders’ 5-day CAR. Contrary to findings of Martynova and Renneboog (2008) and Moeller et al. (2005), who respectively find insignificant and negative returns, bidder returns are positive and significantly different from zero. For the whole sample, the mean (median) of bidders’ CAR is positive: 0.020 (0.007) and highly significant at the 1% level. The mean bidders’ CAR is also significant for FC- and NFC- bidders in every measure, where the highest CAR is found for FC bidders when classified as such by the SA-index (0.020 and highly significant at the 1% level). The lowest mean bidders’ CAR is 0.010 (significant at 1%), when a bidder is NFC based on the SA-index. Except for the median in the payout-ratio measure, all medians are also highly significant (at the 1% level). Table 6 also shows the differences in mean (median) CAR between FC- and NFC- bidders for every measure of financial constraints. The mean (median) CAR of FC bidders is significantly higher than that of NFC bidders at the 10% when for the SA-index, with a difference of 0.010 (0.005). This provides some preliminary evidence of higher returns from acquisitions when a bidder is financially constrained. The other two measures, the ASCL-index and a bidders’ payout-ratio do not show significant differences in CARs between bidders. Overall, the summary statistics suggest that the degree of bidders’ constraints is a positive determinant affecting shareholder returns only when classified as constrained by the SA-index. However, these univariate results do not control for the previously described variables that can affect bidder returns. Therefore, the next section will examine this relationship in the proposed multivariate framework.
Table 7. Summary statistics CAR
All
acquisitions SA-index ASCL-index Payout-ratio FC NFC Diff. FC NFC Diff. FC NFC Diff. Mean 0.013*** 0.020*** 0.010*** 0.010* 0.016*** 0.012*** 0.004 0.013** 0.013*** 0.000 Median 0.007*** 0.011*** 0.006*** 0.005* 0.009*** 0.006*** 0.003 0.003 0.008*** -0.005 Observations 875 263 612 285 590 263 612
4.2 FINANCIAL CONSTRAINTS AND BIDDER RETURNS
Table 8 reports the estimates for the regression model of five-day cumulated abnormal return (CAR) to bidders for the sample of financially constrained- and unconstrained bidders. The effect of each measure is separately examined in models 1-3. Every model includes industry- and fixed-effects and standard errors are robust using White’s standard errors.
The coefficient of the SA-index in model 1 is positive (0.012) and significant at the 5 % level. This implies that when a bidder is financially constrained, when classified as constrained by the SA-index, returns from acquisitions are 1.2% higher compared to unconstrained bidders. The coefficients of the other measures of constraints, the ASCL-index and payout-ratio in model 2 and 3 are positive, 0.003 and 0.004 respectively, but insignificant. Hence, when classified as constrained by these measures, bidder returns are not significantly higher than unconstrained bidders. These results indicate partial support for hypothesis one, that returns from acquisitions are higher when a bidder is financially constrained. However, this is only true when classified as constrained by the SA-index.
Most control variables on deal- and firm level are insignificant. No evidence is found for differences in bidder returns from acquisitions which involve domestic acquisitions (Erel et al.,2012); private acquisitions (Morck et al.,1990); method of payment (Mayers and Majluf, 1984) and relative size of the acquisition (Moeller et al.,2004). In model 1, the SA-index, target size is significant at the 5% level and has a coefficient of 0.003, implying that the acquisition of larger targets result in higher returns, contrary to findings of Fuller et al. (2002) who argue that larger firms are more difficult to integrate. Given that the summary statistics in table 5 show that over 82% of acquisitions involve private targets, this might be the result of a high liquidity discount for large private targets, as argued by Officer (2007). In line with findings of Lang et al. (1991), bidders’ Tobin’s Q is significant in every index of financial constraints (at the 1%, 5% and 5% level respectively) with a coefficient of 0.0004. Tobin’s Q can be interpreted as an indication of growth opportunities, resulting in higher acquisition returns. 4.3 FINANCIAL CONSTRAINTS, BIDDER RETURNS AND INVESTOR PROTECTION
Table 8. Financial constraints, regression of bidder returns
(1) (2) (3)
Measures of financial constraints:
SA-index 0.012** (0.006) ASCL-index 0.003 (0.006) Payout-ratio 0.004 (0.007)
Deal & Firm characteristics
Domestic 0.005 0.006 0.007 (0.005) (0.005) (0.005) Industry-focused -0.006 -0.006 -0.006 (0.005) (0.005) (0.005) Private -0.003 -0.003 -0.003 (0.006) (0.006) (0.006) All cash -0.007 -0.006 -0.006 (0.005) (0.005) (0.005) All shares 0.002 0.003 0.003 (0.010) (0.010) (0.011) Relative size -0.007 -0.005 -0.005 (0.007) (0.007) (0.007) Target size 0.003** 0.002 0.002 (0.002) (0.001) (0.001) Tobin's Q 0.0004* 0.0004** 0.0004** (0.000) (0.000) (0.000) Intercept -0.030 -0.007 -0.011 (0.038) (0.035) (0.035)
Country Fixed Effects Yes Yes Yes
Year Fixed Effects Yes Yes Yes
Industry Fixed Effects Yes Yes Yes
Adjusted R² 0.046 0.041 0.042
Observations 875 875 875
Given are the regression results for the effect of bidders' financial constraints on bidders' 5-day CAR. SA-index, ASCL-index and payout-ratio are dummy variables that equal 1 when the bidder is constrained and 0 otherwise; Domestic is a dummy variable set equal to 1 when the acquisition is domestic and 0 otherwise; Private is a dummy variable set equal to 1 when the target is private and 0 otherwise; All cash and All shares are dummy variables set equal to 1 if the acquisition is financed only with cash or shares and 0 otherwise. Robust standard errors are reported below the estimated coefficient. *,**,*** report significance at the 10%, 5% and 1% level respectively. Definitions of variables are given in table A1.
Table 9. Financial constraints and investor protection, regression of bidder returns
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) Measures of financial constraints:
SA-index 0.014** 0.019* 0.015** 0.013 (0.006) (0.010) (0.006) (0.010) ASCL-index 0.004 0.004 0.004 0.002 (0.006) (0.008) (0.006) (0.009) Payout-ratio 0.000 0.000 0.002 0.001 (0.000) (0.006) (0.006) (0.008) Deal & Firm characteristics
Domestic 0.003 0.004 0.004 0.003 0.004 0.004 0.002 0.003 0.003 0.002 0.003 0.003 (0.004) (0.004) (0.004) (0.004) (0.005) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004) (0.004) Industry-focused -0.004 -0.003 -0.003 -0.004 -0.003 -0.003 -0.004 -0.004 -0.004 -0.004 -0.004 -0.004 (0.005) (0.005) -(0.003) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Private -0.002 -0.002 -0.001 -0.002 -0.002 -0.002 -0.004 -0.004 -0.003 -0.004 -0.004 -0.003 (0.005) (0.006) -(0.001) (0.005) (0.007) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) (0.006) All cash -0.006 -0.006 -0.005 -0.006 -0.006 -0.005 -0.008 -0.007 -0.007 -0.007 -0.007 -0.006 (0.005) (0.005) -(0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) All shares 0.002 0.004 0.005 0.002 0.004 0.005 0.003 0.005 0.005 0.003 0.005 0.005 (0.010) (0.010) (0.005) (0.010) (0.008) (0.011) (0.010) (0.010) (0.011) (0.010) (0.010) (0.011) Relative size -0.006 -0.004 -0.004 -0.006 -0.004 -0.004 -0.006 -0.004 -0.004 -0.006 -0.004 -0.004 (0.007) (0.007) -(0.004) (0.007) (0.004) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) (0.007) Target size 0.002 0.000 0.000 0.002 0.000 0.000 0.002 0.001 0.001 0.002 0.001 0.001 (0.002) (0.001) (0.000) (0.002) (0.002) (0.001) (0.002) (0.001) (0.001) (0.002) (0.001) (0.001) Tobin's Q 0.0004** 0.0004** 0.0004** 0.0004* 0.0004 0.0004 0.0004** 0.0004** 0.0005 0.0004** 0.0005** 0.0005 (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) (0.000) Investor protection Anti-director 0.002 0.012 0.002 0.005 0.002 0.000 (0.005) (0.029) (0.005) (0.005) (0.006) (0.005) Anti-director x financial constraint -0.007 0.000 0.004
(0.013) (0.010) (0.008)
Anti-self dealing 0.009* 0.009* 0.009 0.009* 0.008 0.009 (0.005) (0.005) (0.005) (0.005) (0.005) (0.005) Anti-self-dealing x financial constraint 0.003 0.004 0.001
(0.013) (0.012) (0.013) Intercept -0.014 0.012 0.015 -0.016 0.012 0.015 -0.020 0.006 0.008 -0.015 0.008 0.012
(0.033) (0.029) (0.015) (0.034) (0.031) (0.030) (0.033) (0.030) (0.030) (0.032) (0.031) (0.029) Country Fixed Effects No No No No No No No No No No No No Year Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Industry Fixed Effects Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Adjusted R² 0.025 0.019 0.019 0.026 0.019 0.019 0.028 0.023 0.022 0.028 0.023 0.021 Observations 875 875 875 875 875 875 875 875 875 875 875 875
Models 4-6 include the interaction effect of bidders’ financing constraints and investor protection measured by the anti-director right index. Again, the only significant coefficient is that of financially constrained bidders in the SA-index (model 4), which is 0.019 at the 10% level. Acquisitions from FC bidders result in a 1.9% higher CAR than NFC bidders.
Both the anti-director rights variable and the interaction variable are insignificant in all models, lending no support for hypothesis 2, that returns from financially constrained bidders are higher in countries with strong investor protection.
When shareholder protection is the anti-self-dealing index, in models 7-9, stronger investor protection has a significantly positive effect on bidder returns for the SA-index (model 7) and the ASCL index (model 8). Both coefficients have a value of 0.009 and are significant at the 10% level. Hence, bidder returns in countries with strong investor protection are 0.9% higher than bidder returns in countries with weaker investor protection. This is in line with Shleifer and Vishny (1997) who find that stronger investor protection results in higher returns, by reducing agency costs and by a more efficient allocation of resources. The coefficient of the SA-index remains positive, 0.015, and is significant at the 5% level. Returns from acquisitions are 1.5% higher when the bidder is financially constrained. The interaction effect of the anti-self-dealing index and financial constraints is tested in models 10-12. In all models, bidders’ financial constraints do not have a significant effect on returns from acquisitions. Stronger investor protection leads to higher bidder returns in model 10, the SA-index, with a coefficient of 0.009 which is significant at the 10% level.
The interaction variable does not show any significant effect in all models, indicating no evidence for higher returns when a bidder is financially constrained in a country with strong investor protection. Bidders’ Tobin’s Q has a significant and positive effect when the SA-index is the measure of financial constraints in all models (models 1,4,7 and 10). Its coefficient is 0.004 and is significant at the 5% level. Its effect is also significantly positive in models 2,3, 8 and 11. These results are to a large extent in line with findings of Lang et al. (1991), who argue that Tobin’s Q can have a positive effect on bidder returns, as it is indication of growth opportunities.
protection for this index, and for that of payout-ratio, but for this index, the CAR itself does not significantly differ between constrained-, and unconstrained bidders.
Overall, the results in table 7 and 8 show that returns from acquisitions of financially constrained bidders are higher only when considered as constrained by the SA-index of Hadlock and Pierce (2010). In other words, because the SA-index is negatively loaded on a firm’s size and age, smaller and younger firms show higher returns from acquisitions. This partially supports hypothesis 1. In addition, no evidence is found for higher returns of financially constrained bidders in countries with strong investor protection Therefore, the results do not support hypothesis 2.
4.4 ROBUSTNESS
In an attempt to find robust results for the effect of bidders’ financial constraints and the combined effect of bidders’ financial constraints and strong investor protection, the main results are based on three different measures of financial constraints and two indices of investor protection. As the results show, significant findings are not robust across these measures. Therefore, in a further attempt to find stable, significant results, the next section introduces an alternative classification of financially constrained bidders. Moreover, the significance of cumulated abnormal returns, CARs is tested with a non-parametric model.
4.4.1 NON-PARAMETRIC TEST
Table 2A in the appendix indicates that 5-day bidders’ CAR are non-normally distributed. The summary statistics in table 7 show that bidders’ CAR is significant for the sample, and significantly higher for FC bidders when grouped by the SA-index. These results are based on a t-test, which assumes a normal distribution. In order to cope with non-normality, here a Wilcoxon signed-rank test is employed, which does not assume a normal distribution (Brooks, 2008), to see if these results hold for the non-normality assumption. Table 3A in the appendix reports the results. Results in table 3A are similar to the results in table 7 and show significance for bidders’ CAR for the sample. The significant difference in bidders’ CAR also remains persistent for the SA-index. Hence, bidder returns from acquisitions are significantly different and positive, contrary to findings by Martynova and Renneboog (2008), who argue that bidder returns are insignificantly different from zero.
4.4.2 ALTERNATIVE CLASSIFICATION FINANCIAL CONSTRAINTS
The classification of constrained bidders remains the same for the SA-index and payout-ratio, but to the unconstrained group are now assigned only bidders in the top three deciles, resulting in a sample of 613 bidders. For the ASCL-index, to the financially constrained group are assigned bidders with a score of 4 and to the unconstrained group bidders with a score of 0. This reduces the sample dramatically to 84 bidders.
The results in table 3A indicate several differences relative to the model in the empirical results. The coefficients of financial constraints in SA-index increase relative to their values in table 8 and 9 and remain significant throughout all models. The highest coefficient, 0.032 (significant at the 5% level) is found in model 4, which is an improvement to its value in the main results (0.019, significant at the 1% level). Being financially constrained results in 3.2% higher returns. This finding can directly be dedicated to the alternative specification of constrained- and unconstrained bidders, as it increases the difference between the two groups. The other two classifications remain insignificant. The effect of bidders’ Tobin’s Q on bidders’ CAR remains significantly positive, as well as that of strong investor protection when measured by the anti-self-dealing index. Consistent with the view that domestic acquisitions result in positive returns from acquisitions (Erel et al.,2015), model 1 (SA-index) reports a significantly positive coefficient of 0.01 (1%) at the 10% level.
Besides the stronger, positive relationship between being a financially constrained bidder when classified as such by the SA-index, this alternative classification does not result in different findings, as previously described in the main results.
5. CONCLUSIONS
Within the European context of the sample under observation, a bidder’s home country’s degree of investor protection is proposed to interact in the relation between financial constraints and returns. Strong investor protection reduces agency costs and makes value-enhancing investments more likely (Shleifer and Vishy, 1997). Despite the positive effect of strong investor protection on bidder returns from acquisitions in some models, results do not suggest that returns from acquisitions are higher when financially constrained bidders are from countries with strong investor protection.
