Deal or No Deal Hormones and Completion of Mergers and

Deal or No Deal: Hormones and Completion of Mergers and Acquisitions* Maurice Levi Sauder School of Business University of British Columbia 2053 Main Mall, Vancouver, BC V6T 1Z2 604.822.8260 maurice.levi@sauder.ubc.ca Kai Li Sauder School of Business University of British Columbia 2053 Main Mall, Vancouver, BC V6T 1Z2 604.822.8353 kai.li@sauder.ubc.ca Feng Zhang Sauder School of Business University of British Columbia 2053 Main Mall, Vancouver, BC V6T 1Z2 604.827.5294 feng.zhang@sauder.ubc.ca First version: November, 2008 This version: December, 2008 Abstract Testosterone, a hormone linked to aggression, has been shown to influence the outcome of the two-player ultimatum game. Specifically, high-testosterone players tend to reject low offers even though this is against their interest. As a result, neither player receives any reward. This paper investigates whether such a “testosterone effect” is present in mergers and acquisitions (M&As). The relevant “players” are the CEOs of bidder and target companies. Hormone levels are proxied by male CEO age, based on the evidence that testosterone diminishes after middle age. The outcomes of interest are bid withdrawal, unfriendly bids, and the use of a tender offer. We find that bid withdrawal is more likely if the bidder or target male CEO is young. Further, young target male CEOs increase the chance that negotiations become unfriendly and that acquisitions take the form of tender offers. These effects are present controlling for other plausible influences on offer completion and friendliness. Keywords: male CEO testosterone, male CEO age, mergers and acquisitions, bid withdrawal, unfriendly bids, tender offers. JEL classification: G34 * Levi and Li acknowledge the financial support from the Social Sciences and Humanities Research Council of Canada. We also acknowledge the financial support from the Certified Management Accounting Society of British Columbia. All remaining errors are our own. Deal or No Deal: Hormones and Completion of Mergers and Acquisitions Abstract Testosterone, a hormone linked to aggression, has been shown to influence the outcome of the two-player ultimatum game. Specifically, high-testosterone players tend to reject low offers even though this is against their interest. As a result, neither player receives any reward. This paper investigates whether such a “testosterone effect” is present in mergers and acquisitions (M&As). The relevant “players” are the CEOs of bidder and target companies. Hormone levels are proxied by male CEO age, based on the evidence that testosterone diminishes after middle age. The outcomes of interest are bid withdrawal, unfriendly bids, and the use of a tender offer. We find that bid withdrawal is more likely if the bidder or target male CEO is young. Further, young target male CEOs increase the chance that negotiations become unfriendly and that acquisitions take the form of tender offers. These effects are present controlling for other plausible influences on offer completion and friendliness. Keywords: male CEO testosterone, male CEO age, mergers and acquisitions, bid withdrawal, unfriendly bids, tender offers. JEL classification: G34 1 1. Introduction In a number of respects, the merger and acquisition (M&A) process has close parallels with the ultimatum game involving the division of a sum of money between two players. In this game one of the players, the “proposer,” offers the other player, the “responder,” part of a sum that is to be divided between them. The offer is final – hence the term “ultimatum.” If the amount offered is rejected, neither player receives anything. In the case of M&As, the parallels are that the bidder CEO can generally be viewed as the proposer and the target CEO as the responder. Further, the outcome of the M&A ultimately involves the division of the estimated gain from the acquisition between the two companies: the more the bidding firm pays in an acquisition the greater is the target firm’s share of the estimated gain. Economic rationality suggests that in the ultimatum game, the proposer will make a low offer, believing that since a low amount is better than nothing and that there is no benefit from holding out for more in terms of future rounds of play – this is a one-round game – the low offer will nevertheless be accepted by the responder.1 The remarkable result is that while the proposer does generally make a low offer, the low offer is frequently rejected by the responder.2 This has been attributed to a preference for distributional fairness, with players willing to sacrifice financially to achieve a fairer – more equal – division of the contested sum.3 The outcome of the game has also been attributed to players wishing to appear being fair to those administering it, at least when contested sums are rather small.4 However, perhaps even more intriguing than possible explanations of the outcome is how the frequent failure to reach the rational equilibrium in the ultimatum game – low offer proposed and low offer accepted – depends on the hormone levels of the players. For example, see Rubinstein (1982) and Stahl (1972) for the theoretical equilibrium. Numerous papers report this outcome, including Guth, Schmittberger, and Schwarze (1982) and Roth (1995). This finding has been shown to occur even when the stakes are relatively large (Hoffman, McCabe, and Smith (1996)) and in different cultural environments (Roth, Prasnikar, Okuno-Fujiwara, and Zamir (1991) and Henrich et al. (2001)). 3 See Bolton (1991), Fehr and Schmidt (1999), and Rabin (1993). 4 See Alexander (1987) and Nowak, Page, and Sigmund (2000). 2 1 2 In a widely cited paper, Burnham (2007) describes how the outcome of the ultimatum game is related to levels of the steroidal hormone, testosterone, which in humans and other animals has been associated with elevated levels of aggression, and with challenges to achieve dominance in a competitive situation. Burnham’s game is played by male students in a laboratory setting and involves the players dividing $40. The terms of the game are public knowledge to all players. To simplify the game the proposer is permitted to offer either $5 or $25 to the responder, with the ultimatum being a once-and-only offer and a once-and-only response. In order to evaluate the potential role played by testosterone, saliva is swabbed from the mouths of all players. This is done on several days before the experiment and again prior to the experiment.5 It is observed that among players with above average testosterone, 45 percent of responders offered $5 reject the offer. This compares to a 7 percent rejection rate among responders with below-average testosterone. Burnham argues that the outcome is consistent with high-testosterone responders feeling more challenged by a low offer than are low-testosterone responders: low offers are seen as aggressive. Five of the seven responders with the highest testosterone levels reject the $5 offers versus just one of the 19 responders with the lowest levels. As for the proposers, there is a tendency for high-testosterone players to offer the higher sum, that is, $25, but this is not statistically significant. While there are strong parallels between the nature of the ultimatum game and M&As there are some differences. Most importantly, unlike the situation in the ultimatum game, in the course of M&A negotiations there are opportunities to respond to offers and for offered amounts to be revised (Boone and Mulherin (2007)). Furthermore, in addition to the possibility of responses to and revisions of offers , bidders may opt for less amiable approaches, including making an unfriendly unsolicited offer, or going around the target board and making a tender offer directly to the target shareholders. This paper empirically investigates the role of testosterone in M&As to see whether the prospects for reaching a negotiated agreement 5 Experimental details can be found in Burnham (2007). 3 correspond to the role testosterone plays in the ultimatum game, in particular whether it influences the likelihood that negotiations fail, become unfriendly, or that acquisitions take the form of a tender offer directly to the target shareholders. Before presenting the empirical evidence we review the related literature on the association between age and testosterone, and the association between testosterone and aggression, which together constitute an association between age and aggression. Then, if aggression is a factor in the failure of M&A negotiations or in negotiations becoming unfriendly we have a possible association between age and M&As. Using data from over 300 acquisition bids for the period 1997-2007, we cannot go back and measure the testosterone of M&A negotiation participants. Therefore, we proxy testosterone by male CEO age. We find a strong and positive association between the bidder male CEO being young and the withdrawal rate of initiated M&As, and between the target male CEO being young and the withdrawal rate of initiated M&As. We also find that the target male CEO being young is positively associated with negotiations becoming unfriendly and with the use of tender offers. We argue that these results are what we would expect based on the experimental evidence from the ultimatum game that motivates our study. We further argue that other possible channels of influence of CEO age, channels which are alternatives to testosterone, work in the opposite direction to what we find, or that they are controlled for by other factors in our estimation. The plan of the paper is as follows. The next section reviews the related literature on the relations between age and testosterone and between testosterone and aggression. Despite an absence of prior work on the role and consequences of CEO hormones in M&As, studies of hormones in other contexts provide a valuable link to the work presented here. Section 3 describes our M&A sample and the model specifications used to analyze the age-M&A connection. Section 4 presents our main results on the incidences of bid withdrawal, unfriendly bids, and tender offers. Section 5 considers whether age may be capturing effects other than those of hormones. Section 6 summarizes and concludes. 4 2. Testosterone versus Age and Aggression Ideally, an investigation of the effects of bidder and target male CEO testosterone levels on the nature of M&A negotiations would be based on observations of these levels along the lines of Burnham’s laboratory investigation. However, given that we have to rely on historical M&A data on the characteristics of negotiations, this is not possible. Indeed, it would probably be difficult to observe CEO testosterone levels even if we were able to do an experiment in real time as not all CEOs would be likely to consent to providing samples. Furthermore, it is not easy to identify the pivotal moment of the deal when the initial bid is made and thus the saliva sample should be taken. Finally, there are both circadian cycles (highest in the morning and lowest in the evening) and seasonal variations (lowest in the spring and highest in the fall and early winter months) in testosterone levels (Dabbs (1990a, 1990b)), that might lead to poor reliability in testosterone measurement. Fortunately, while less satisfactory than direct measures, we base our study on male CEO age as this is strongly associated with the relevant testosterone levels. 2.1 Testosterone versus Age Studies of the age-testosterone relationship are divided into cross-sectional and longitudinal analyses. Typical of the cross-sectional research is Ferrini and Barrett-Connor (1998) who study age-associated variations in total as well as the more relevant bioavailable testosterone among 810 Caucasian men in Rancho Bernardo, California. Plasma samples from middle- and upper middle-class men aged 24-90 years were obtained in 1984-1987, frozen in polypropylene tubes at -70°C, and analyzed in 1993 using radioimmunoassay. The age-hormone association is obtained after adjusting for subjects’ weight, body mass index, alcohol consumption, smoking levels, exercise, caffeine intake, disease, and sample storage time. Diseases controlled for are diabetes and coronary heart problems, both of which are determined by subject questionnaire responses, reviews of medical records, as well as by physical examinations including plasma glucose measurement and electrocardiogram readings. 5 The results of the study are graphically illustrated in Figure 1, where we note that research subjects of less than 50 years of age are lumped together. The graph shows that the bioavailable testosterone declines steadily with age: each point on the graph is the average for each of the 5-year age group except for the <50 years sample that contains subjects from 24-49 years. Ferrini and Barrett-Connor (ibid. p. 754) conclude their study by declaring: “bioavailable testosterone … decreased dramatically with age in these community-dwelling men, independently of body size, health behavior, and chronic disease.” Typical of the longitudinal studies of age and testosterone levels is that of Harman et al. (2001) who investigate stored samples taken from 890 men in the Baltimore Longitudinal Study on Aging.6 The researchers support the use of such a longitudinal evaluation on the grounds that cross-sectional investigations confound age and disease, obscuring the direct effects of age, although this does not apply to the Ferinni and Barrett-Connor research which goes to great lengths to control for disease as well as other factors. As in the aforementioned cross-sectional investigation, the sample used in the Harman et al. (2001) consists of largely middle-class Caucasian men. These are examined at approximately 2-year intervals. The results for the effects of age on total testosterone and free testosterone (a calculated value related to bioavailable testosterone that is highly correlated with salivary testosterone examined in Burnham’s (2007) study) are summarized in Figure 2. The graph shows essentially the same decline in free testosterone as in the cross-sectional study of Ferrini and Barrett-Connor (1998) using bioavailable testosterone.7 Figure 3 from Travison et al. (2007) illustrates that while total testosterone has declined over time, the negative age-testosterone association has persisted. 2.2 Testosterone versus Aggression The large sample in Harman et al. (2001) longitudinal study follows two smaller studies by Zmuda, Cauley, and Kriska (1997) that sample a population with a 13-year interval, and Morley et al. (1997). 7 The Harman et al. plot is over a slightly longer age range, but analysis of the detailed data shows a similar relative decline to that found in the cross-sectional research of Ferrini and Barrett-Connor (1998). 6 6 While not every study of the human testosterone-aggression connection has identified a simple causal relationship comparable to what has been found in non-human animal research, the disparities have generally been linked to moderating variables such as metabolism, age, sex, circadian rhythm, stress, past experience, and social status. In order to investigate the underlying systematic relationship that might be obscured by these intervening factors, Book, Starzyk, and Quinsey (2001) perform a meta-analysis that begins with 106 articles on the human testosterone-aggression relationship. After carefully considering possible double-counting from including studies using the same or overlapping databases and preening out studies with information gaps, the study focuses on 45 research publications. When outcomes are weighed by the number of participants, a significant positive relationship is revealed. The conclusion is reinforced at numerous levels including studies of prison inmates, spousal violence, and normal adolescent males.8 The balance of evidence from prior work strongly suggests a positive association between age and testosterone and between testosterone and aggression, leading to our empirical proxy of aggression by male CEO age. 3. Our Data and Research Framework In this section we describe how our M&A sample is formed and present our empirical specifications to examine the role of bidder and target male CEO testosterone levels, proxied by age, in bid withdrawal, unfriendly bids, and tender offers. 3.1 Our Sample To form our M&A sample, we begin with all US acquisition attempts with announcement dates between January 1, 1997 and December 31, 2007 as identified by the Thomson Financial’s SDC database. The above time frame is chosen because the data on corporate boards from the See Harris, Rushton, Hampson, and Jackson (1996), Mazur and Booth (1998), and Olweus, Mattson, Schalling, and Low (1980). 8 7 RiskMetrics Group are available during this period. We identify all deals where both the bidder and the target are public firms and where the form of deal is coded as a merger, an acquisition of majority interest, or an acquisition of assets. After applying the above filters, we obtain 4,956 deals. Our final sample of 303 acquisition attempts is an intersection of RiskMetrics for director information, ExecuComp for CEO tenure,9 Compustat for accounting information, CRSP for stock prices for the period 1997-2007, and the requirement that both bidder and target CEOs are male.10 3.2 Sample Overview Table 1 presents the summary statistics of our sample. About 12 percent of the acquisition bids are withdrawn. Using a much larger sample of 2,150 announced bids over the period 1984-2001, Chen, Harford, and Li (2007) show that the withdrawal frequency is 16 percent. According to SDC, a bid is hostile if the target board officially rejects the offer but the acquirer persists with the attempted takeover; and a bid is unsolicited when the acquirer makes an offer for the target without prior negotiations. Following Betton, Eckbo, and Thorburn (2008a), unfriendly bids are the sum of outright hostile bids and unsolicited bids. Approximately 9 percent of the acquisition bids in our sample are perceived as unfriendly. Betton et al. (2008a) show that the frequency of unfriendly bids is about 3 percent based on a sample of over 35,000 bids over the period 1980-2005. Andrade, Mitchell, and Stafford (2001) report the frequency of hostile bids at any point in time during the 1990s is 4 percent. Note that our sample only contains deals for which both the bidder and the target firms are public. About 16 percent of our M&A We focus on bidder CEO tenure as the data availability on target CEO tenure is extremely poor; only ten percent of the sample has information on target CEO tenure. 10 By limiting the sample to male CEOs for whom testosterone is known to vary closely with age we omit ten observations from the sample where the target or bidder CEO is a woman. The effect of gender on M&As is explored in Levi, Li, and Zhang (2008). 9 8 deals take the form of a tender offer. Betton et al. (2008a) show that the frequency of tender offers is 12 percent based on a sample of over 25,000 successful targets over the period 19802005. Overall, our sample is not very different from those much larger samples covering earlier sample periods. Across our sample, the average (median) age of bidding firm male CEOs is 56 (55) years old, and the average (median) age of target firm male CEOs is 55 (56) years old. Motivated by the studies of population testosterone levels reviewed earlier, we use the age of 50 years as the cut-off to separate young male CEOs from the rest. The data show that 19 percent of bidder male CEOs and 22 percent of target male CEOs are no older than 50 years.11 In terms of the target firm characteristics, we show that the average target board size is 9.8. On average, 67 percent of the target boards are occupied by independent directors. In 71 percent of the target companies the CEO is also the Chairman of the Board (COB). Compared to summary statistics reported in Bange and Mazzeo (2004) using 436 bids over the period 19791990, we observe that corporate boards become more independent over time, while the same fraction of firms have CEOs being COBs in the target firms. Target firms experience sales growth of 8 percent per year. The average ratio of market value of total assets to book value of total assets (Tobin’s Q) is 1.7. The average operating cash flow to total assets is 12 percent. The average book leverage is 41 percent. The average target price runup from day -42 to day -1 relative to the bid announcement date is 8.2 percent. Using a sample of 7,522 initial control bids for public targets over the period 1980-2002, Betton et al. (2008c) report that the average price runup for target firms is 8.3 percent. In terms of the bid characteristics, we show that about one tenth of the deals have multiple bidders. Using a much larger sample over a much longer sample period, Betton et al. (2008c) report that on average 11 percent of the bids in their sample have competing bids. We rely on the RiskMetrics data to retrieve information on poison pills in the year before the bid and 11 It is worth noting that our main conclusions remain the same if we use the cut off age of 45 years old. 9 we show that 64 percent of the targets have the pill in place. We retrieve toehold information at the time of the bid from SDC and only 10 out of 303 bids in our sample have positive toeholds, leading to a sample average of 0.4 percent. Conditional on positive toeholds, the sample average toehold size is 11.7%. Betton et al. (2008b) show that 1,363 bids out of their 10,806 bids over the period 1973-2002 have positive toeholds. They also show that the use of toeholds in takeovers declines over time and is more common for private bidders than for public bidders. Bid premium is defined as the ratio of the final offer price to the target stock price four weeks prior to the bid minus one. The sample average bid premium is 36 percent. About 15 percent of the deals use only cash as the method of payment, and more than one third of the deals are pure stock swaps. About 30 percent of the deals are diversifying, i.e., the bidder and target belong to different industry classifications as defined by Fama and French (1997). The relative rate of cash deals, stock deals, and diversifying deals in our sample is comparable to that in other research of takeovers during the 1990s (see, for example, Andrade et al. (2001), and Betton et al. (2008a)). The average relative size, defined as the ratio of the transaction value to the market value of total assets of the bidder, is 32 percent. In terms of the bidder firm characteristics, we show that the average bidder CEO tenure is about 8 years. The average bidder board size is 11.7. On average, 69 percent of the bidder boards are occupied by independent directors. In 79 percent of the bidder companies the CEO is also the COB. Bidder firms experience sales growth of 15 percent per year. The average bidder Tobin’s Q is 2.2. The average operating cash flow to total assets is 14 percent. The average book leverage is 40 percent. Overall, it seems that the bidder firms in our sample enjoy faster sales growth, higher Tobin’s Q, and higher operating cash flow than the target firms. 3.3. Model Specifications To explore the role of CEO hormones in M&As, we first run the following crosssectional benchmark regression focusing on the target firm characteristics: 10 Bid Outcomei = α0 + β1 Target Board Sizei + β2 Target Proportion of Independent Directorsi + β3 Target CEO is COBi + β4 Target Sales Growthi + β 5Target Tobin' s Qi + β 6 Target Operating Cash Flowi + β7Target Book Leveragei + β 8Target Runupi + Other Controls + ei , (1) where “Bid Outcome” could be bid withdrawal, unfriendly bid, or the use of a tender offer. The control variables are motivated by Schwert (2000), Bange and Mazzeo (2004), and Chen et al. (2007). Following estimation of the above benchmark model in equation (1), we add our key variables of interest, namely, the CEO age variables proxying for hormone levels. These are captured by the indicator variables that the bidder male CEO is young and that the target male CEO is young. The expanded regression is specified as follows: Bid Outcomei = α0 + β1 Bidder Male CEO is Youngi + β 2Target Male CEO is Youngi + β 3Target Board Sizei + β4Target Proportion of Independent Directorsi + β5 Target CEO is COBi + β6 Target Sales Growthi + β7Target Tobin' s Qi + β 8Target Operating Cash Flowi + β 9Target Book Leveragei + β10Target Runupi + Other Controls+ ei . (2) We separate the CEO age/hormone variable into the bidder versus the target CEO age/hormone levels to allow for possible differences in their roles as is found with Burnham’s (2007) study of the ultimatum game. Finally, we incrementally add the bid characteristics, and the bidder firm characteristics to equation (2). Bid characteristics include indicator variables for competing bid, poison pill, all cash, all stock, diversifying deal, and two continuous variables, toehold and relative size. Bidder firm characteristics are the same as those of target firm characteristics. In all model specifications, we include year fixed effects and industry fixed effects (based on one-digit SIC codes), and we employ robust standard errors. Before proceeding with our multivariate analysis, we examine the correlation between our three dependent variables and all right-hand-side variables. Table 2 presents the correlation 11 matrix. There is economically significant positive correlation between bid withdrawal and unfriendly bids (at 0.49), and between unfriendly bids and tender offers (at 0.21). There is significant positive correlation between the bidder male CEO age indicator variable and bid withdrawal. Also, there is significant positive correlation between the target male CEO age indicator variable and bid withdrawal, and between the target male CEO age indicator variable and the use of tender offers. Overall, the extent of correlation among most pairs of variables raises little concern for multicollinearity in our regression analysis. 4. The Hormone Effect in M&As Table 3 presents our probit regression results where the dependent variable is bid withdrawal, which is set equal to one if the bid is withdrawn, and zero otherwise. In our baseline regression (column (1)), we include only the target firm characteristics to establish a benchmark. We find that bids made to targets with large boards and to fast growing targets are positively associated with bid withdrawal. We introduce the CEO age variables in column (2). We find that a young and by implication high testosterone bidder (target) male CEO, as measured by the indicator variable that the bidder (target) male CEO is no older than 50 years, is significantly and positively associated with bid withdrawal. In economic terms, the bidder (target) male CEO being young increases the likelihood of bid withdrawal by over 6 (5) percent. This effect is statistically significant at the one-percent (five-percent) level. Recall that the sample frequency of bid withdrawal is 12.2 percent (Table 1). This is a very striking finding as it strongly supports an association between testosterone, as proxied by male CEO age, and M&As.12 Furthermore, we learn that the presence of a young bidder male CEO is slightly more important for noncompletion of an initiated M&A than the presence of a young target male CEO: a young bidder male CEO explains 6.4 percent of the 12.2 percent of withdrawn offers versus 5.1 percent 12 Later we address the issue of whether age could be reflecting factors other than testosterone. 12 explained by a young target male CEO. It is worthy of note that the sum of the effects of the young bidder and young target male CEO is 11.5 percent (6.4 percent + 5.1 percent). Therefore, we find that the combined effects of young male CEOs in both the bidder and target companies can explain virtually all the withdrawals. The roughly equal importance of the bidder and target CEO is consistent with the fact that in M&As, unlike the ultimatum game, there is a potential for dialogue and back-and-forth negotiation. It should not be a surprise that each side to negotiations could be associated with failure to reach an agreement. In addition to the above main findings concerning CEO age, all other explanatory variables have the same effects as shown in our baseline specification. In column (3), we add the bid characteristics to the above specification. The significant effects of a young bidder and a young target male CEO remain. Indeed, they both become significant at better than the one-percent level. The combined effects of young bidder and target male CEOs still just about fully explain the 12.2 percent of withdrawn offers, with the contributions of young bidder and target male CEO being 5.5 and 5.4, respectively. The two sides have almost exactly the same association with offers being withdrawn. We also show that bids with multiple bidders and diversifying bids are positively associated with bid withdrawal. Bange and Mazzeo (2004) also find that deals involving competing bids are more likely to fail. Finally, in column (4), we present results using our encompassing specification that includes the target firm, bid, and bidder firm characteristics. We show that both a young male bidder CEO and a young male target CEO are still significantly positively associated with bid withdrawal even when the other factors are included. Both are significant at better than the onepercent level, and between them explain most of the 12.2 percent of withdrawn offers. With the all encompassing specification we find that a young bidder male CEO is more responsible for failure to satisfactorily reach an agreement: 6.4 percent versus 4.1 percent of the 12.2 percent of uncompleted negotiations. Further, we find that bidders with more independent boards and higher Tobin’s Q are negatively associated with bid withdrawal, while bidders whose CEOs are 13 also COBs are positively associated with bid withdrawal.13 Bidder CEO tenure has no effect on bid withdrawal. We find that CEO age and CEO tenure are clearly different in their effect on bid withdrawal, where it is age that is related to testosterone, not tenure. This helps rule out the possibility that age is proxying for experience, which is related to tenure, rather than testosterone. Table 4 presents the probit regression results when the dependent variable is unfriendly, which is set equal to one if SDC codes the bid as either hostile or as an unsolicited bid, and zero otherwise. In our baseline regression (column (1)), we find that bids made to targets with large boards are more likely to be unfriendly, while bids made to targets with more independent boards are less likely to be unfriendly. When we add the CEO age variables we obtain the results in column (2). We find that the presence of a young target male CEO is positively associated with unfriendly bids. A young bidder male CEO does not play a role. If we relate the target CEO to the responder in the ultimatum game and the bidder CEO to the proposer, unfriendliness in M&As follows the same pattern as failure to reach an agreement in the game: the responder (target) attitude is what scuttles the deals. In terms of the economic significance, the target CEO being a young male increases the likelihood of an unfriendly bid by 3.0 percent. This effect is significant at the ten-percent level. Recall that the sample frequency of unfriendly bids is 8.9 percent (Table 1), so CEO age is explaining a moderate proportion of unfriendly offers. All other control variables have similar effects on unfriendly bids as those in the baseline regression. We further add the bid characteristics in column (3). The young target male CEO effect remains: young, and by implication high testosterone, target male CEOs are more likely to be associated with unfriendly M&A situations. We also find that bids with multiple bidders, targets with poison pills, higher bid premiums, and all cash payment are positively associated with unfriendly bids. The multiple bidder effect may reflect the unsolicited nature of some of the bidders’ offers; recall that unfriendly bids include unsolicited offers, and negotiations are likely 13 We include bid premium to control for possible horizon effects of age. This is discussed later when we consider whether age could be affecting M&As in ways other than through being closely related to testosterone. 14 to be limited at any time to a single potential buyer. Executives in target firms with poison pills are more likely to regard the bids as hostile in order to defend their own positions. The influence of bid premium indicates that to succeed in an unfriendly takeover attempt it is, ceteris paribus, necessary to pay more. The effect of cash payment corroborates Schwert (2000) who finds that cash is more likely to be offered as a means of payment in hostile actions. Finally, in column (4), we present results from our encompassing specification that includes the target firm, bid, and bidder firm characteristics. We show that both a young bidder and a young target male CEO are significantly and positively associated with the occurrence of unfriendly bids. Further, we find bidders with larger boards, higher Tobin’s Q, and higher leverage are negatively associated with unfriendly bids, while bids made by firms where CEOs are also COBs and fast growing firms are positively associated with unfriendly bids. Bidder CEO tenure has no effect on the occurrence of unfriendly bids. Table 5 presents the probit regression results when the dependent variable is tender offer, which is set equal to one if an acquisition takes the form of a tender offer, and zero otherwise. In our baseline regression (column (1)), we find that bids made to targets with larger boards and higher Tobin’s Q are less likely to take the form of a tender offer. We include the CEO age variables in column (2). We find that the presence of a young target male CEO is positively associated with the use of a tender offer. However, the effect is not statistically significant. All other control variables have similar effects on tender offers as those in the baseline regression. We further add the bid characteristics in column (3). We find that when these extra controls are included the presence of a young target male CEO is positively associated with the use of tender offers. Since tender offers are likely to follow unsuccessful negotiations – going to the target shareholders after other negotiations have failed – we would expect such a target CEO effect: cooperative negotiations are less likely with young, and by implication high testosterone, target male CEOs. It is the target and not the bidder male CEO who is more important for the bidder resorting to a tender offer. Using the context of the ultimatum game, it is the responder 15 and not the proposer who is more important in determining if a cooperative resolution fails. In terms of the economic significance, the target CEO being a young male increases the likelihood of a tender offer by 3.3 percent. This effect is significant at the ten-percent level. Recall that the sample frequency of tender offers is 15.8 percent (Table 1). Further, we show that bids with multiple bidders, targets with poison pills, bids with toeholds, bids with higher premiums, bids with all cash payment, and diversifying bids are positively associated with tender offers, while bids with all stock payment are negatively associated with tender offers. Finally, in column (4), we present results from our encompassing specification that includes the target firm, bid, and bidder firm characteristics. We show that the target male CEO age/testoserone effect on the likelihood of resorting to a tender offer remains. Further, we find that bidders with more independent boards and higher Tobin’s Q are negatively associated with tender offers, while bids made by firms with high operating cash flows are positively associated with tender offers. Bidder CEO tenure has no effect on the use of tender offers. In summary, we present evidence suggesting a potentially important role of testosterone in M&As. Ceteris paribus, the presence of a young bidder male CEO is positively associated with bid withdrawal, while the presence of a young target male CEO is positively associated with bid withdrawal, unfriendly bids, and the use of tender offers. In the language of the ultimatum game, the proposer and the responder (i.e. the bidder CEO and the target CEO, respectively) are both influential in bid withdrawal. Unfriendly bids and tender offers depend more on the responder, as in Burnham’s experiments. (Recall that Burnham (2007) finds the testosterone of the responder to influence whether a cooperative, rational equilibrium is reached.) However, it is still necessary to consider whether the evidence described above is consistent with the hormone explanation rather than some other pathways relating to male CEO age. 5. Alternative Explanations of the Male CEO Age Effect 16 As mentioned earlier, investigating the effects of testosterone on the successful completion of M&A negotiations and on whether the negotiations become unfriendly or tender offers are employed should be based on direct measurement of CEOs’ hormone levels during the negotiation process. Since we are unable to do this in the context of historical data on M&As, we have suggested an alternative, specifically to proxy testosterone by age. The validity of this approach clearly depends on the extent of the close association of hormone levels with age. As we have explained in Section 2, the literature is clear on this association. However, this still begs the question of whether CEO age might be serving as a catch-all for something other than the bidder or target male CEO testosterone level. Could it be that CEO age is capturing the length of the horizon over which the CEO is valuing the relevance of a completed and non-hostile M&A on his long-run career? Two possible channels may be at work through CEO horizon. First, a young CEO may care more about reputation than an older CEO, and as a result might want to avoid failure of negotiations and/or negotiations deteriorating into hostility. If this is the case we would expect fewer negotiation failures and unfriendly situations, and fewer tender offers. However, we have found the opposite for all three dependent variables.14 Second, a young bidder CEO may view completion of an M&A at a low price of the target company to have more benefit in the long run. This would, were such an effect to be present, lead to more non-completions and unfriendly situations with young bidder male CEOs. However, this should also show up with a smaller bid premium for acquisitions where the bidder male CEO is young. However, in additional regression results not reported here, we find no effect of the bidder or target male CEO age on bid premium. Furthermore, when we include bid premium in our analysis of the withdrawal likelihood this variable is insignificant (Table 3 columns (3) and (4)). 14 An alternative argument might be that it is good for a CEO to appear tough for future deals. However, a horizon effect of this type would be more likely to show up as relating to the target CEO age instead of the bidder CEO age. We find significant associations for both CEOs. 17 It is also possible that CEO age is serving as a proxy for experience. Perhaps, an older CEO with more experience can better judge the correct price for an acquisition and thereby avoid failure of the offer. This possibility is questioned by the observation of no effect of the CEO tenure variable across Tables 3-5: we show that bidder CEO tenure has no material effect on any of our three outcome variables. What we find is that even when we allow for CEO tenure the bidder male CEO’s age or the target male CEO’s age matters, and sometimes, both CEOs’ age variables matter. Finally, we find that depending on the variable of interest – bid withdrawal, unfriendly bids, and tender offers – there are effects of bidder (proposer) and target (responder) age/testosterone levels. However, what shows up in explaining all three dependent variables is the target’s (responder’s) age. This is in line with Burnham’s conclusion concerning the importance of the responder’s testosterone. Unlike Burnham (2007) we also find that the bidder’s (proposer’s) age/testosterone is also importance in some contexts, particularly in bid withdrawal, and in unfriendly bids when all target, bid, and bidder characteristics are included. This is, perhaps, not surprising in M&As since, unlike the ultimatum game, there is room for back-andforth exchanges, and hence for negotiators on both sides of the discussion to play a role. 6. Conclusions In this paper we examine whether testosterone, which is associated with male aggression and which we have proxied by male CEO age, is associated with the course of M&A negotiations. Consistent with evidence from studies of the ultimatum game, we show that the bidder male CEO being young is positively associated with bid withdrawal. We also find that young target male CEOs are positively associated with bid withdrawal, the acquisition bids being unfriendly, and the use of a tender offer. The combined effects of young bidder and young target male CEOs explain just about all the proportion of withdrawn offers. The more pervasive effects of the target (responder) versus the bidder (proposer) male CEO age, where the former is 18 significantly associated with all three measures of bid withdrawal, unfriendliness, and resorting to tender offers, is consistent with the observations by Burnham (2007) in the ultimatum game: responders’ reactions are generally more relevant than proposers. The connections we find are present even after controlling for other plausible influences on merger outcomes. It would also appear that age is not influencing these merger outcomes through some pathways other than through testosterone. At least, from the evidence we have gathered, it would appear that our age proxy for testosterone and associated aggression, and does not proxy for CEO horizon or experience. M&As represent an arena in which the aggressive instinct of male CEOs may come to the fore and influence whether cooperative outcomes can be achieved. Narratives documenting troubled encounters, often leading to withdrawn offers and unfriendly takeover attempts including tender offers, are common. We believe the results reported here are in line with this experience. However, confidence in the potential importance of what we have documented would be enhanced if hormones of decision makers are found to be influential in other corporate settings beyond the arena of M&As where “contesting” CEOs meet head to head. We hope the present work leads to such further studies. 19 References: Alexander, Richard D., 1987, The Biology of Moral Systems, Aldin de Gruyter. Andrade, Gregor, Mark Mitchell, and Erik Stafford, 2001, New evidence and perspectives on mergers, Journal of Economic Perspectives 15, 103-120. Bange, Mary M., and Michael A. Mazzeo, 2004, Board composition, board effectiveness, and the observed form of takeover bids, Review of Financial Studies 17, 1185-1215. Betton, Sandra, B. Espen Eckbo, and Karin Thorburn, 2008a, Corporate takeovers, in: Eckbo, B.E. (Ed.), Handbook of Corporate Finance: Empirical Corporate Finance Vol. II, Elsevier/North-Holland, 291-429. Betton, Sandra, B. Espen Eckbo, and Karin S. Thorburn, 2008b, Merger negotiation and the toehold puzzle, Journal of Financial Economics, forthcoming. Betton, Sandra, B. Espen Eckbo, and Karin S. Thorburn, 2008c, Markup pricing revisited, Tuck School of Business Working Paper No. 2008-45. Bolton, Gary E., 1991, A comparative model of bargaining: Theory and evidence, American Economic Review 81, 1096-1136. Book, Angela S., Katherine B. Starzyk, and Vernon Quinsey, 2001, The relationship between testosterone and aggression: A meta-analysis, Aggression and Violent Behavior 6, 579599. Boone, Audra L., and J. Harold Mulherin, 2007, How are firms sold?, Journal of Finance 62, 847-875. Burnham, Terence C., 2007, High-testosterone men reject low ultimatum game offers, Proceedings of The Royal Society Series B 274, 2327-2330. Chen, Xia, Jarrad Harford, and Kai Li, 2007, Monitoring: Which institutions matter?, Journal of Financial Economics 86, 279-305. Dabbs, James M. Jr., 1990a, Age and seasonal variation in serum testosterone concentration among men, Chronobiol International 7, 245-249. Dabbs, James M. Jr. 1990b, Salivary testosterone measurements: Reliability across hours, days, and weeks, Physiology and Behavior 48, 83-86. Fama, Eugene, and Kenneth French, 1997, Industry costs of equity, Journal of Financial Economics 43, 153-193. Fehr, Ernst, and Klaus Schmidt, 1999, A theory of fairness, competition, and cooperation, 20 Quarterly Journal of Economics 114, 817-868. Ferrini, Rebecca L., and Elizabeth Barrett-Connor, 1998, Sex hormones and age: A crosssectional study of testosterone and estradiol and their bioavailable fractions in community-dwelling men, American Journal of Epidemiology 147, 750-754. Guth, Werner, Rolf Schmittberger, and Bernd Schwarze, 1982, An experimental analysis of ultimatum bargaining, Journal of Economic Behavior and Organization 3, 367-388. Harman, S. Mitchell, E. Jeffrey Metter, Jordan D. Tobin, Jay Pearson, and Marc R. Blackman, 2001, Longitudinal effects of aging on serum total and free testosterone levels in healthy men, Journal of Clinical Endocrinology and Metabolism 86, 724-731. Harris, Julie Aitken, J. Phillippe Rushton, Elizabeth Hampson, and Douglas N. Jackson, 1996, Salivary testosterone and self-report aggressive and pro-social personality characteristics in men and women, Aggressive Behavior 22, 321-331. Henrich, Joseph, Robert Boyd, Samuel Bowles, Colin Camerer, Ernst Fehr, Herbert Gintis, and Richard McElreath, 2001, In search of Homo economics: Behavioral experiments in 15 small-scale societies, American Economic Review 91, 73-78. Hoffman, Elizabeth, Kevin McCabe, and Vernon Smith, 1996, On expectations and monetary stakes in ultimatum games, International Journal of Game Theory 25, 289-301. Levi, Maurice D., Kai Li, and Feng Zhang, 2008, Mergers and acquisitions: The role of board gender diversity, University of British Columbia working paper. Mazur, Allan, and Alan Booth, 1998, Testosterone and dominance in men, Behavioral and Brain Sciences 21, 353-397. Morley, John E., F. E. Kaiser, H. M. Perry, P. Patrick, P. M. K. Morley, P. M. Stauber, B. Vellas, R. N. Baumgartner, and P. J. Garry, 1997, Longitudinal changes in testosterone, luteinizing hormone, and follicle-stimulating hormone in healthy older men, Metabolism, 46, 410-413. Nowak, Martin A., Karen M. Page, and Karl Sigmund, 2000, Fairness versus reason in the ultimatum bargaining game, Science 289, 1773-1775. Olweus, Dan, Ake Mattsson, Daisy Schalling, and Hans Low, 1980, Testosterone, aggression, physical, and personality dimensions in normal adolescent males, Psychosomatic Medicine 42, 253-269. Rabin, Matthew, 1993, Incorporating fairness into game theory and econometrics, American Economic Review 83, 1281-1303. Roth, Alvin E., 1995, Bargaining experiments, in: Kagel, J. and Roth, A. (Eds.), Handbook of 21 Experimental Economics, Princeton University Press, 253-348. Roth, Alvin E., Vesna Prasnikar, Masahiro Okuno-Fujiwara, and Shmuel Zamir, 1991, Bargaining and market behavior in Jerusalem, Ljubljana, Pittsburgh, and Tokyo: An experimental study, American Economic Review 81, 1068–1095. Rubinstein, Ariel, 1982, Perfect equilibrium in a bargaining model, Econometrica 50, 97-109. Schwert, G. William, 2000, Hostility in takeovers: In the eyes of the beholder?, Journal of Finance 55, 2599-2640. Stahl, Ingolf, 1972, Bargaining Theory, Economic Research Institute, Stockholm. Travison, Thomas, Andre B. Araujo, Amy B. O’Donnell, Varant Kupelian, and John B. McKinlay, 2007, A population-level decline in serum testosterone levels in American men, Journal of Clinical Endocrinology and Metabolism 92, 196-202. Zmuda, Joseph M., Jane A. Cauley, Andrea Kriska, Nancy W. Glynn, James P. Gutai, and Lewis H. Kuller, 1997, Longitudinal relation between endogenous testosterone and cardiovascular disease risk factors in middle-aged men, A 13-year follow-up of former Multiple Risk Factor Intervention Trial participants, American Journal of Epidemiology 146, 609-617. 22 Figure 1 Age versus Total and Bioavailable Testosterone in Cross-Sectional Investigation of 810 Caucasian Men Source: Rebecca L. Ferrini and Elizabeth Barrett Connor, “Sex Hormones and Age: A Cross-Sectional Study of Testosterone and Estradol and Their Bioavailable Fractions in Community-Dwelling Men, American Journal of Epidemiology, Vol. 147, No. 8, 1998, pp. 750 – 754. 23 Figure 2 Age versus Total and Free Testosterone in Longitudinal Investigation of 890 Caucasian Men Source: S. Mitchell Harman, E. Jeffrey Metter, Jordan D. Tobin, Jay Pearson and Marc E. Blackman, “Longitudinal Effects of Aging on Serum Total and Free Testosterone Levels in Healthy Men”, The Journal of Clinical Endocrinology and Metabolism, Vol. 86, No. 2, 2001, pp. 724 - 731. 24 Figure 3 Secular Decline in Total Testosterone Source: Thomas G. Travison, Andre B. Araujo, Amy B. O’Donnell, Varant Kupelain and John. B. McKinlay, “A Population-Level Decline in Serum Testosterone Levels in American Men”, The Journal of Clinical Endocrinology and Metabolism, Vol. 92, No. 1, 2007, pp. 196 – 202. 25 Table 1 Sample Summary Statistics Our sample consists of 303 merger and acquisition attempts announced in the period 1997-2007. The data are retrieved from the SDC database and have available data from RiskMetrics/ExecuComp/CRSP/Compustat. Withdrawal is set to one if the bid is withdrawn, and zero otherwise. Unfriendly is set to one if SDC regards the bidder’s attitude as hostile or the bid is unsolicited, and zero otherwise. Tender Offer is set to one if the bidder uses a tender offer, and zero otherwise. Bidder Male CEO Age is the age of the bidder firm CEO. Bidder Male CEO is Young is set to one if the bidder male CEO is not more than 50 years old, and zero otherwise. Target Male CEO Age is the age of the target firm CEO. Target Male CEO is Young is set to one if the target male CEO is not more than 50 years old, and zero otherwise. Board Size is the number of directors serving on the board. Proportion of Independent Directors is measured as the number of independent directors divided by the board size. CEO is COB is set to one if the CEO is also the Chairman of the Board (COB), and zero otherwise. Sales Growth is the growth rate in sales. Tobin’s Q is the market value of total assets divided by the book value of total assets. Operating Cash Flow is sales minus the cost of goods sold, sales and general administration expenses, and working capital change, divided by the book value of total assets. Book Leverage is the book value of total debt divided by the book value of total assets. Target Runup is the cumulative abnormal return to the target firm’s stock for trading days [-41, -1] before the bid announcement. Competing Bid is set to one if there are multiple bidders for the target firm, and zero otherwise. Poison Pill is set to one if a poison pill is in place for the target firm, and zero otherwise. Toehold is the percent of target shares owned by the initial bidder before the bid. Bid Premium is the ratio of the final offer price to the target stock price four weeks prior to the original announcement date minus one. All Cash, All Stock, and Diversifying are indicator variables that take the value of one if only cash is used to pay for the acquisition, or if only equity is used, or if the bidder and the target are in the same Fama-French industry (Fama and French (1997)), respectively, and zero otherwise. Relative Size is the transaction value divided by the market value of total assets of the bidder. Bidder CEO Tenure is the number of years the CEO has been on the job. All dollar amounts are in 2007 millions of dollars, and all percentages are in real numbers. All firm characteristics are measured at the fiscal year end prior to the bid announcement. 26 Variable Withdrawal Unfriendly Tender Offer Bidder Male CEO Age Bidder Male CEO is Young Target Male CEO Age Target Male CEO is Young Target Board Size Target Proportion of Independent Directors Target CEO is COB Target Sales Growth Target Tobin’s Q Target Operating Cash Flow Target Book Leverage Target Runup Competing Bid Poison Pill Toehold Bid Premium All Cash All Stock Diversifying Relative Size Bidder CEO Tenure Bidder Board Size Bidder Proportion of Independent Directors Bidder CEO is COB Bidder Sales Growth Bidder Tobin’s Q Bidder Operating Cash Flow Bidder Book Leverage N 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 303 mean 0.122 0.089 0.158 55.822 0.191 55.129 0.221 9.762 0.671 0.706 0.083 1.736 0.118 0.406 0.082 0.119 0.640 0.004 0.356 0.149 0.360 0.310 0.322 8.172 11.654 0.692 0.786 0.153 2.230 0.141 0.402 std 0.328 0.285 0.366 6.124 0.394 6.495 0.416 3.012 0.176 0.484 0.255 1.085 0.097 0.206 0.245 0.324 0.481 0.026 0.272 0.356 0.481 0.463 0.365 7.488 3.906 0.170 0.427 0.249 2.348 0.084 0.200 5th percentile 0.000 0.000 0.000 46.000 0.000 44.000 0.000 6.000 0.357 0.000 -0.198 0.907 0.015 0.058 -0.196 0.000 0.000 0.000 0.015 0.000 0.000 0.000 0.013 1.000 7.000 0.333 0.000 -0.142 1.001 0.026 0.071 median 0.000 0.000 0.000 55.000 0.000 56.000 0.000 9.000 0.700 1.000 0.085 1.350 0.121 0.437 0.045 0.000 1.000 0.000 0.320 0.000 0.000 0.000 0.190 6.000 11.000 0.727 1.000 0.105 1.537 0.136 0.422 95th percentile 1.000 1.000 1.000 64.000 1.000 65.000 1.000 15.000 0.900 1.000 0.441 3.972 0.260 0.692 0.473 1.000 1.000 0.000 0.866 1.000 1.000 1.000 1.112 25.000 18.000 0.906 1.000 0.607 5.720 0.275 0.712 27 Table 2 The Correlation Matrix Our sample consists of 303 merger and acquisition attempts announced in the period 1997-2007. The data are retrieved from the SDC database and have available data from RiskMetrics/ExecuComp/CRSP/Compustat. Withdrawal is set to one if the bid is withdrawn, and zero otherwise. Unfriendly is set to one if SDC regards the bidder’s attitude as hostile or the bid is unsolicited, and zero otherwise. Tender Offer is set to one if the bidder uses a tender offer, and zero otherwise. Bidder Male CEO Age is the age of the bidder firm CEO. Bidder Male CEO is Young is set to one if the bidder male CEO is not more than 50 years old, and zero otherwise. Target Male CEO Age is the age of the target firm CEO. Target Male CEO is Young is set to one if the target male CEO is not more than 50 years old, and zero otherwise. Board Size is the number of directors serving on the board. Proportion of Independent Directors is measured as the number of independent directors divided by the board size. CEO is COB is set to one if the CEO is also the Chairman of the Board (COB), and zero otherwise. Sales Growth is the growth rate in sales. Tobin’s Q is the market value of total assets divided by the book value of total assets. Operating Cash Flow is sales minus the cost of goods sold, sales and general administration expenses, and working capital change, divided by the book value of total assets. Book Leverage is the book value of total debt divided by the book value of total assets. Target Runup is the cumulative abnormal return to the target firm’s stock for trading days [-41, -1] before the bid announcement. Competing Bid is set to one if there are multiple bidders for the target firm, and zero otherwise. Poison Pill is set to one if a poison pill is in place for the target firm, and zero otherwise. Toehold is the percent of target shares owned by the initial bidder before the bid. Bid Premium is the ratio of the final offer price to the target stock price four weeks prior to the original announcement date minus one. All Cash, All Stock, and Diversifying are indicator variables that take the value of one if only cash is used to pay for the acquisition, or if only equity is used, or if the bidder and the target are in the same FamaFrench industry (Fama and French (1997)), respectively, and zero otherwise. Relative Size is the transaction value divided by the market value of total assets of the bidder. Bidder CEO Tenure is the number of years the CEO has been on the job. All dollar amounts are in 2007 millions of dollars, and all percentages are in real numbers. All firm characteristics are measured at the fiscal year end prior to the bid announcement. The corresponding p-value is reported in the brackets below each correlation coefficient. 28 1 2 3 4 5 6 7 8 9 Withdrawal Unfriendly Tender Offer Bidder Male CEO is Young Target Male CEO is Young Target Board Size Target Proportion of Independent Directors Target CEO is COB Target Sales Growth 1 1.000 0.485 [0.000] 0.031 [0.586] 0.100 [0.081] 0.117 [0.042] 0.110 [0.055] -0.036 [0.530] -0.044 [0.441] 0.132 [0.022] 0.019 [0.748] 0.036 [0.538] 0.093 [0.105] -0.019 [0.736] 0.299 [0.000] 0.007 [0.910] 0.006 [0.922] -0.033 [0.570] 0.043 [0.459] -0.007 [0.910] 0.120 [0.036] 0.130 [0.023] -0.045 [0.435] -0.008 [0.887] -0.155 [0.007] 0.022 [0.701] 0.023 [0.694] -0.013 [0.821] 0.051 [0.375] 0.083 [0.149] 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 1.000 0.213 [0.000] -0.005 [0.932] 0.085 [0.142] 0.048 [0.407] -0.057 [0.323] -0.074 [0.202] -0.004 [0.951] -0.041 [0.481] 0.033 [0.566] 0.090 [0.116] -0.011 [0.853] 0.315 [0.000] 0.114 [0.048] 0.001 [0.987] 0.069 [0.234] 0.130 [0.024] -0.138 [0.016] 0.066 [0.254] 0.107 [0.062] -0.041 [0.474] -0.073 [0.204] -0.115 [0.046] 0.076 [0.188] 0.029 [0.620] -0.008 [0.885] 0.090 [0.118] 0.047 [0.413] 1.000 -0.050 [0.383] 0.096 [0.097] -0.152 [0.008] -0.115 [0.046] -0.017 [0.770] 0.009 [0.872] -0.038 [0.509] 0.077 [0.181] 0.098 [0.087] 0.009 [0.872] 0.204 [0.000] 0.080 [0.163] 0.223 [0.000] 0.282 [0.000] 0.302 [0.000] -0.288 [0.000] 0.119 [0.038] -0.098 [0.088] -0.006 [0.913] -0.057 [0.327] -0.034 [0.554] -0.015 [0.796] -0.067 [0.249] 0.025 [0.660] 0.273 [0.000] 0.081 [0.160] 1.000 0.084 [0.143] -0.221 [0.000] -0.087 [0.131] 0.001 [0.991] 0.075 [0.191] 0.152 [0.008] -0.027 [0.642] 0.056 [0.332] -0.009 [0.877] -0.049 [0.395] -0.037 [0.518] -0.009 [0.882] -0.039 [0.500] 0.009 [0.875] 0.125 [0.030] 0.000 [0.998] 0.070 [0.222] -0.162 [0.005] -0.204 [0.000] -0.010 [0.868] -0.188 [0.001] -0.067 [0.247] 0.120 [0.037] -0.044 [0.451] 0.077 [0.182] 1.000 -0.180 [0.002] -0.128 [0.026] -0.153 [0.008] 0.074 [0.202] 0.050 [0.387] 0.009 [0.882] 0.014 [0.805] -0.068 [0.241] -0.024 [0.682] -0.114 [0.047] 0.040 [0.493] -0.005 [0.926] 0.024 [0.684] 0.081 [0.159] -0.065 [0.259] -0.004 [0.944] 0.019 [0.747] 0.009 [0.882] 0.013 [0.825] -0.049 [0.395] 0.076 [0.187] 0.064 [0.264] 0.093 [0.107] -0.002 [0.979] 1.000 0.133 [0.021] 0.047 [0.412] -0.017 [0.768] -0.138 [0.016] -0.035 [0.541] -0.204 [0.000] -0.080 [0.167] 0.117 [0.041] 0.048 [0.403] -0.052 [0.365] -0.167 [0.004] -0.233 [0.000] 0.094 [0.104] -0.059 [0.310] 0.088 [0.125] 0.041 [0.478] 0.430 [0.000] -0.009 [0.883] 0.068 [0.235] -0.003 [0.964] -0.138 [0.016] -0.294 [0.000] -0.177 [0.002] 1.000 0.105 [0.068] -0.048 [0.405] -0.095 [0.098] -0.096 [0.096] -0.024 [0.678] -0.004 [0.950] 0.034 [0.551] 0.122 [0.033] -0.087 [0.132] -0.031 [0.594] -0.051 [0.379] -0.018 [0.758] -0.002 [0.979] 0.086 [0.138] 0.065 [0.259] 0.036 [0.535] 0.073 [0.204] 0.085 [0.142] 0.029 [0.611] -0.092 [0.109] -0.102 [0.078] -0.037 [0.526] 1.000 -0.022 [0.697] 0.088 [0.125] 0.079 [0.171] 0.024 [0.682] 0.017 [0.768] -0.051 [0.375] 0.128 [0.026] 0.042 [0.468] 0.030 [0.605] -0.111 [0.054] 0.142 [0.013] 0.039 [0.504] 0.086 [0.134] -0.021 [0.719] 0.109 [0.059] -0.014 [0.803] 0.031 [0.596] -0.056 [0.335] -0.033 [0.563] -0.031 [0.586] 0.023 [0.695] 1.000 0.199 [0.001] 0.211 [0.000] -0.025 [0.662] -0.134 [0.019] -0.073 [0.203] -0.045 [0.432] -0.004 [0.951] 0.001 [0.988] -0.058 [0.313] 0.159 [0.006] -0.024 [0.674] -0.035 [0.541] 0.024 [0.684] -0.015 [0.797] -0.053 [0.357] -0.019 [0.746] 0.164 [0.004] 0.143 [0.013] 0.061 [0.291] 0.051 [0.373] 1.000 0.287 [0.000] -0.011 [0.842] -0.037 [0.518] 0.084 [0.147] 0.103 [0.074] 0.070 [0.227] 0.035 [0.547] -0.023 [0.685] 0.185 [0.001] 0.134 [0.020] 0.100 [0.083] -0.002 [0.974] -0.118 [0.041] 0.036 [0.528] -0.071 [0.220] 0.043 [0.456] 0.340 [0.000] 0.214 [0.000] 0.116 [0.044] 1.000 0.177 [0.002] -0.082 [0.154] 0.076 [0.189] 0.023 [0.692] 0.168 [0.003] -0.134 [0.020] -0.017 [0.763] -0.134 [0.019] 0.141 [0.014] 0.230 [0.000] 0.002 [0.971] -0.066 [0.249] 0.017 [0.773] -0.043 [0.459] -0.013 [0.823] 0.012 [0.835] 0.373 [0.000] 0.231 [0.000] 1.000 -0.040 [0.494] 0.015 [0.800] 0.013 [0.817] -0.048 [0.404] 0.068 [0.237] -0.078 [0.174] -0.296 [0.000] 0.034 [0.552] 0.225 [0.000] -0.117 [0.042] -0.242 [0.000] 0.095 [0.099] 0.084 [0.147] -0.040 [0.487] 0.039 [0.505] 0.253 [0.000] 0.537 [0.000] 1.000 0.023 [0.687] -0.053 [0.361] 0.000 [0.995] 0.208 [0.000] 0.068 [0.240] 0.000 [0.996] 0.044 [0.441] 0.029 [0.615] 0.052 [0.367] -0.037 [0.527] -0.073 [0.208] 0.034 [0.553] -0.078 [0.176] -0.025 [0.662] 0.024 [0.675] 0.006 [0.912] 1.000 0.084 [0.145] 0.007 [0.900] -0.019 [0.747] -0.039 [0.503] -0.190 [0.001] 0.018 [0.751] 0.200 [0.001] 0.041 [0.480] 0.025 [0.668] -0.154 [0.007] 0.041 [0.475] 0.009 [0.874] -0.027 [0.637] 0.028 [0.630] 0.007 [0.905] 1.000 -0.119 [0.039] 0.056 [0.333] -0.054 [0.346] -0.011 [0.845] 0.042 [0.468] 0.047 [0.419] -0.049 [0.395] 0.029 [0.620] -0.029 [0.619] -0.038 [0.505] -0.031 [0.587] 0.095 [0.100] 0.137 [0.017] 0.034 [0.560] 1.000 0.024 [0.673] 0.186 [0.001] -0.112 [0.051] 0.027 [0.641] -0.064 [0.266] -0.041 [0.473] -0.022 [0.706] 0.041 [0.474] -0.226 [0.000] -0.098 [0.090] -0.010 [0.859] 0.087 [0.132] 0.006 [0.917] 1.000 0.110 [0.056] -0.122 [0.033] 0.010 [0.862] -0.110 [0.056] 0.053 [0.360] -0.084 [0.143] -0.135 [0.019] 0.012 [0.834] -0.035 [0.543] 0.157 [0.006] 0.124 [0.031] -0.042 [0.469] 1.000 -0.313 [0.000] 0.121 [0.035] -0.227 [0.000] 0.028 [0.632] -0.084 [0.143] 0.052 [0.369] -0.073 [0.206] -0.124 [0.031] 0.042 [0.471] 0.206 [0.000] 0.106 [0.066] 1.000 -0.072 [0.214] -0.079 [0.173] 0.038 [0.510] 0.178 [0.002] -0.090 [0.120] -0.075 [0.196] 0.188 [0.001] 0.168 [0.003] -0.173 [0.003] -0.245 [0.000] 1.000 -0.029 [0.614] -0.077 [0.179] -0.012 [0.839] 0.023 [0.693] -0.014 [0.809] 0.096 [0.094] 0.094 [0.103] 0.060 [0.300] 0.088 [0.125] 1.000 -0.100 [0.082] -0.208 [0.000] -0.108 [0.060] -0.073 [0.204] 0.126 [0.029] -0.024 [0.679] 0.030 [0.601] 0.230 [0.000] 1.000 0.088 [0.128] -0.194 [0.001] 0.104 [0.071] 0.066 [0.254] 0.012 [0.839] -0.050 [0.382] -0.081 [0.160] 1.000 -0.062 [0.281] -0.009 [0.876] 0.107 [0.063] -0.166 [0.004] -0.245 [0.000] -0.298 [0.000] 1.000 0.164 [0.004] -0.122 [0.034] -0.086 [0.135] 0.030 [0.598] 0.139 [0.015] 1.000 0.020 [0.731] -0.042 [0.470] -0.030 [0.601] 0.009 [0.872] 1.000 0.174 [0.002] -0.038 [0.512] -0.065 [0.262] 1.000 0.347 [0.000] -0.017 [0.775] 1.000 0.294 [0.000] 10 Target Tobin's Q 11 Target Operating Cash Flow 12 Target Book Leverage 13 Target Runup 14 Competing Bid 15 Poison Pill 16 Toehold 17 Bid Premium 18 All Cash 19 All Stock 20 Diversifying 21 Relative Size 22 Bidder CEO Tenure 23 Bidder Board Size 24 Bidder Proportion of Independent Directors 25 Bidder CEO is COB 26 Bidder Sales Growth 27 Bidder Tobin's Q 28 Bidder Operating Cash Flow 29 Bidder Book Leverage 29 Table 3 Explaining Bid Withdrawal Our sample consists of 303 merger and acquisition attempts announced in the period 1997-2007. The data are retrieved from the SDC database and have available data from RiskMetrics/ExecuComp/CRSP/Compustat. Withdrawal is set to one if the bid is withdrawn, and zero otherwise. Bidder Male CEO Age is the age of the bidder firm CEO. Bidder Male CEO is Young is set to one if the bidder male CEO is not more than 50 years old, and zero otherwise. Target Male CEO Age is the age of the target firm CEO. Target Male CEO is Young is set to one if the target male CEO is not more than 50 years old, and zero otherwise. Board Size is the number of directors serving on the board. Proportion of Independent Directors is measured as the number of independent directors divided by the board size. CEO is COB is set to one if the CEO is also the Chairman of the Board (COB), and zero otherwise. Sales Growth is the growth rate in sales. Tobin’s Q is the market value of total assets divided by the book value of total assets. Operating Cash Flow is sales minus the cost of goods sold, sales and general administration expenses, and working capital change, divided by the book value of total assets. Book Leverage is the book value of total debt divided by the book value of total assets. Target Runup is the cumulative abnormal return to the target firm’s stock for trading days [-41, -1] before the bid announcement. Competing Bid is set to one if there are multiple bidders for the target firm, and zero otherwise. Poison Pill is set to one if a poison pill is in place for the target firm, and zero otherwise. Toehold is the percent of target shares owned by the initial bidder before the bid. Bid Premium is the ratio of the final offer price to the target stock price four weeks prior to the original announcement date minus one. All Cash, All Stock, and Diversifying are indicator variables that take the value of one if only cash is used to pay for the acquisition, or if only equity is used, or if the bidder and the target are in the same Fama-French industry (Fama and French (1997)), respectively, and zero otherwise. Relative Size is the transaction value divided by the market value of total assets of the bidder. Bidder CEO Tenure is the number of years the CEO has been on the job. All dollar amounts are in 2007 millions of dollars, and all percentages are in real numbers. All firm characteristics are measured at the fiscal year end prior to the bid announcement. We estimate probit models and present the marginal effect of each explanatory variable on the likelihood of a bid withdrawal. The corresponding p-value is reported in the brackets below each coefficient. Superscripts ***, **, and * correspond to statistical significance at the one, five, and ten percent levels, respectively. Dependent Variable (1) Bidder Male CEO is Young Target Male CEO is Young Target Characteristics Target Board Size Target Proportion of Independent Directors Target CEO is COB Target Sales Growth Target Tobin’s Q Target Operating Cash Flow Target Book Leverage Target Runup Withdrawal (2) (3) 0.0636*** 0.0553*** [0.008] [0.003] 0.0513** 0.0535*** [0.022] [0.002] 0.0119*** [0.000] -0.0199 [0.587] -0.0143 [0.362] 0.1036*** [0.007] -0.0042 [0.570] -0.0532 [0.443] 0.0400 [0.356] 0.0292 [0.206] 0.0066*** [0.000] 0.0051 [0.838] -0.0034 [0.694] 0.0753*** [0.000] -0.0052 [0.239] -0.0847** [0.023] 0.0424 [0.112] 0.0129 [0.358] (4) 0.0637*** [0.001] 0.0407*** [0.004] 0.0057*** [0.000] 0.0085 [0.651] -0.0035 [0.610] 0.0678*** [0.000] -0.0029 [0.381] -0.0953*** [0.003] 0.0325 [0.142] 0.0062 [0.568] 0.0102*** [0.002] -0.0547 [0.200] -0.0170 [0.347] 0.1288*** [0.005] -0.0038 [0.643] -0.0761 [0.286] 0.0384 [0.466] 0.0191 [0.474] 30 Bid Characteristics Competing Bid Poison Pill Toehold Bid Premium All Cash All Stock Diversifying Relative Size Bidder Characteristics Bidder CEO Tenure Bidder Board Size Bidder Proportion of Independent Directors Bidder CEO is COB Bidder Sales Growth Bidder Tobin’s Q Bidder Operating Cash Flow Bidder Book Leverage 0.2603*** [0.000] -0.0082 [0.409] 0.0724 [0.543] 0.0015 [0.930] 0.0325 [0.135] 0.0110 [0.367] 0.0386*** [0.006] 0.0087 [0.511] 0.2194*** [0.000] -0.0064 [0.429] 0.1364 [0.164] 0.0009 [0.945] 0.0202 [0.218] 0.0157 [0.142] 0.0356*** [0.001] 0.0082 [0.398] -0.0002 [0.747] -0.0002 [0.816] -0.0435** [0.030] 0.0167* [0.072] -0.0071 [0.606] -0.0022** [0.025] 0.0766 [0.114] 0.0100 [0.661] Year Fixed Effects Industry Fixed Effects Observations Pseudo R-squared Yes Yes 303 0.1821 Yes Yes 303 0.2257 Yes Yes 303 0.3800 Yes Yes 303 0.4213 31 Table 4 Explaining Unfriendly Bids Our sample consists of 303 merger and acquisition attempts announced in the period 1997-2007. The data are retrieved from the SDC database and have available data from RiskMetrics/ExecuComp/CRSP/Compustat. Unfriendly is set to one if SDC regards the bidder’s attitude as hostile or the bid is unsolicited, and zero otherwise. Bidder Male CEO Age is the age of the bidder firm CEO. Bidder Male CEO is Young is set to one if the bidder male CEO is not more than 50 years old, and zero otherwise. Target Male CEO Age is the age of the target firm CEO. Target Male CEO is Young is set to one if the target male CEO is not more than 50 years old, and zero otherwise. Board Size is the number of directors serving on the board. Proportion of Independent Directors is measured as the number of independent directors divided by the board size. CEO is COB is set to one if the CEO is also the Chairman of the Board (COB), and zero otherwise. Sales Growth is the growth rate in sales. Tobin’s Q is the market value of total assets divided by the book value of total assets. Operating Cash Flow is sales minus the cost of goods sold, sales and general administration expenses, and working capital change, divided by the book value of total assets. Book Leverage is the book value of total debt divided by the book value of total assets. Target Runup is the cumulative abnormal return to the target firm’s stock for trading days [-41, -1] before the bid announcement. Competing Bid is set to one if there are multiple bidders for the target firm, and zero otherwise. Poison Pill is set to one if a poison pill is in place for the target firm, and zero otherwise. Toehold is the percent of target shares owned by the initial bidder before the bid. Bid Premium is the ratio of the final offer price to the target stock price four weeks prior to the original announcement date minus one. All Cash, All Stock, and Diversifying are indicator variables that take the value of one if only cash is used to pay for the acquisition, or if only equity is used, or if the bidder and the target are in the same Fama-French industry (Fama and French (1997)), respectively, and zero otherwise. Relative Size is the transaction value divided by the market value of total assets of the bidder. Bidder CEO Tenure is the number of years the CEO has been on the job. All dollar amounts are in 2007 millions of dollars, and all percentages are in real numbers. All firm characteristics are measured at the fiscal year end prior to the bid announcement. We estimate probit models and present the marginal effect of each explanatory variable on the likelihood of a bid withdrawal. The corresponding p-value is reported in the brackets below each coefficient. Superscripts ***, **, and * correspond to statistical significance at the one, five, and ten percent levels, respectively. Dependent Variable (1) Bidder Male CEO is Young Target Male CEO is Young Target Characteristics Target Board Size Target Proportion of Independent Directors Target CEO is COB Target Sales Growth Target Tobin’s Q Target Operating Cash Flow Target Book Leverage Target Runup Unfriendly (2) (3) 0.0005 0.0077 [0.971] [0.314] 0.0302* 0.0232** [0.073] [0.023] 0.0080*** [0.000] -0.0500 [0.150] -0.0095 [0.459] 0.0116 [0.500] -0.0065 [0.464] 0.0015 [0.980] 0.0136 [0.639] 0.0096 [0.611] 0.0038*** [0.003] -0.0251 [0.155] 0.0023 [0.703] 0.0127* [0.091] -0.0049 [0.237] 0.0018 [0.957] 0.0106 [0.506] 0.0011 [0.902] (4) 0.0116*** [0.007] 0.0085*** [0.006] 0.0013*** [0.000] -0.0089** [0.029] 0.0014 [0.370] 0.0016 [0.425] -0.0010 [0.262] 0.0063 [0.409] 0.0043 [0.322] 0.0002 [0.938] 0.0072*** [0.001] -0.0604* [0.096] -0.0115 [0.377] 0.0163 [0.377] -0.0061 [0.486] -0.0075 [0.897] 0.0120 [0.678] 0.0055 [0.781] 32 Bid Characteristics Competing Bid Poison Pill Toehold Bid Premium All Cash All Stock Diversifying Relative Size Bidder Characteristics Bidder CEO Tenure Bidder Board Size Bidder Proportion of Independent Directors Bidder CEO is COB Bidder Sales Growth Bidder Tobin’s Q Bidder Operating Cash Flow Bidder Book Leverage 0.1464*** [0.000] 0.0113* [0.061] -0.0335 [0.801] 0.0212** [0.045] 0.0270* [0.060] -0.0052 [0.454] 0.0064 [0.424] 0.0123 [0.228] 0.0690*** [0.000] 0.0035** [0.020] 0.0280 [0.295] 0.0058** [0.028] 0.0185*** [0.008] -0.0002 [0.906] 0.0030 [0.155] 0.0032 [0.108] 0.0000 [0.747] -0.0004* [0.058] -0.0077 [0.112] 0.0084*** [0.000] 0.0041* [0.093] -0.0003* [0.092] 0.0089 [0.326] -0.0074* [0.077] Year Fixed Effects Industry Fixed Effects Observations Pseudo R-squared Yes Yes 303 0.1793 Yes Yes 303 0.1922 Yes Yes 303 0.3789 Yes Yes 303 0.4691 33 Table 5 Explaining Tender Offers Our sample consists of 303 merger and acquisition attempts announced in the period 1997-2007. The data are retrieved from the SDC database and have available data from RiskMetrics/ExecuComp/CRSP/Compustat. Tender Offer is set to one if the bidder uses a tender offer, and zero otherwise. Bidder Male CEO Age is the age of the bidder firm CEO. Bidder Male CEO is Young is set to one if the bidder male CEO is not more than 50 years old, and zero otherwise. Target Male CEO Age is the age of the target firm CEO. Target Male CEO is Young is set to one if the target male CEO is not more than 50 years old, and zero otherwise. Board Size is the number of directors serving on the board. Proportion of Independent Directors is measured as the number of independent directors divided by the board size. CEO is COB is set to one if the CEO is also the Chairman of the Board (COB), and zero otherwise. Sales Growth is the growth rate in sales. Tobin’s Q is the market value of total assets divided by the book value of total assets. Operating Cash Flow is sales minus the cost of goods sold, sales and general administration expenses, and working capital change, divided by the book value of total assets. Book Leverage is the book value of total debt divided by the book value of total assets. Target Runup is the cumulative abnormal return to the target firm’s stock for trading days [-41, -1] before the bid announcement. Competing Bid is set to one if there are multiple bidders for the target firm, and zero otherwise. Poison Pill is set to one if a poison pill is in place for the target firm, and zero otherwise. Toehold is the percent of target shares owned by the initial bidder before the bid. Bid Premium is the ratio of the final offer price to the target stock price four weeks prior to the original announcement date minus one. All Cash, All Stock, and Diversifying are indicator variables that take the value of one if only cash is used to pay for the acquisition, or if only equity is used, or if the bidder and the target are in the same Fama-French industry (Fama and French (1997)), respectively, and zero otherwise. Relative Size is the transaction value divided by the market value of total assets of the bidder. Bidder CEO Tenure is the number of years the CEO has been on the job. All dollar amounts are in 2007 millions of dollars, and all percentages are in real numbers. All firm characteristics are measured at the fiscal year end prior to the bid announcement. We estimate probit models and present the marginal effect of each explanatory variable on the likelihood of a bid withdrawal. The corresponding p-value is reported in the brackets below each coefficient. Superscripts ***, **, and * correspond to statistical significance at the one, five, and ten percent levels, respectively. Dependent Variable (1) Bidder Male CEO is Young Target Male CEO is Young Target Characteristics Target Board Size Target Proportion of Independent Directors Target CEO is COB Target Sales Growth Target Tobin’s Q Target Operating Cash Flow Target Book Leverage Target Runup Tender Offer (2) (3) -0.0462 -0.0065 [0.194] [0.597] 0.0617 0.0330* [0.122] [0.065] -0.0114* [0.066] -0.0635 [0.436] -0.0305 [0.358] 0.0281 [0.590] -0.0375** [0.013] 0.1487 [0.497] -0.0920 [0.307] -0.0333 [0.581] -0.0014 [0.532] -0.0307 [0.311] 0.0005 [0.965] 0.0449** [0.023] -0.0126** [0.021] -0.0107 [0.866] 0.0119 [0.729] -0.0301 [0.210] (4) -0.0017 [0.752] 0.0152* [0.062] -0.0001 [0.934] -0.0055 [0.655] 0.0020 [0.656] 0.0178* [0.066] -0.0037 [0.109] -0.0333 [0.114] 0.0114 [0.425] -0.0232** [0.017] -0.0123** [0.048] -0.0772 [0.341] -0.0398 [0.231] 0.0329 [0.557] -0.0392*** [0.008] 0.1804 [0.402] -0.0861 [0.338] -0.0352 [0.560] 34 Bid Characteristics Competing Bid Poison Pill Toehold Bid Premium All Cash All Stock Diversifying Relative Size Bidder Characteristics Bidder CEO Tenure Bidder Board Size Bidder Proportion of Independent Directors Bidder CEO is COB Bidder Sales Growth Bidder Tobin’s Q Bidder Operating Cash Flow Bidder Book Leverage 0.2608*** [0.000] 0.0283*** [0.007] 0.4124*** [0.004] 0.0791*** [0.000] 0.1142*** [0.001] -0.0375** [0.024] 0.0447*** [0.008] -0.0246 [0.169] 0.1950*** [0.000] 0.0118*** [0.008] 0.1793*** [0.002] 0.0356*** [0.000] 0.0641*** [0.002] -0.0158** [0.019] 0.0274*** [0.001] -0.0113 [0.158] 0.0000 [0.864] -0.0005 [0.621] -0.0229* [0.099] -0.0029 [0.580] 0.0149 [0.154] -0.0036* [0.087] 0.1469*** [0.000] 0.0082 [0.588] Year Fixed Effects Industry Fixed Effects Observations Pseudo R-squared Yes Yes 303 0.1847 Yes Yes 303 0.1981 Yes Yes 303 0.5042 Yes Yes 303 0.5617 35