Reference Point Theory and Pursuit of Deals

Transcription

1 Reference Point Theory and Pursuit of Deals Abstract We investigate and find separate and joint effects of target and bidder reference points on the outcome and structure of merger deals. First, a firm whose stock price is more distant from its 52-week high reference point is less likely to attract bids, and is less likely to achieve deal completion even if it receives a bid. Second, a target firm has a higher likelihood of being acquired by its own managers (versus unaffiliated bidders) when its stock price is distant from its reference point. Third, bidders rely on their own reference point when considering bids, as the propensity to submit a bid is more likely when the firm s prevailing stock price is closer to its 52-week high. Fourth, when both the bidder and the target have reference points close to their respective current stock prices, both parties are more willing to complete the deal. Finally, hostile deals are more likely when the bidder s stock price is closer to its reference point. Keywords: Target Reference Point, Bidder Reference Point, Probability of Acquisition, MBOs, Hostile Mergers. JEL Classification Code: G34 1

2 I. Introduction and Motivation Individuals make decisions based on the way the problem is framed and on the status quo or reference level, which allows for a psychological component in economic decision making. Monroe (2003) suggests that consumers base their purchasing decisions on the price of the item relative to a reference point. Kamins et al. (2004) apply laboratory experiments to real life situations such as auctions and find that seller specified reference points result in higher bidder prices from buyers. The influence of reference point on investment decisions has also been investigated. Shefrin and Statman (1985) explain why investors hold on to their losing stocks and sell their winning stocks too fast by drawing on Kahneman and Tversky s prospect theory. Odean (1998) confirms the explanation by empirically testing the behavior of 10,000 investment accounts at a large brokerage house. A common reference point for investors is the stock s highest price over the previous year. Barberis and Xiong (2009) show that investors are more prone to sell their stocks when they are priced near the 52-week high, while Grinblatt and Kelaharju (2001) and Huddart, Lang, and Yetman (2009) find that the sell volume is much higher when the stock reaches the 52-week high. Similar results are found by Kaustia (2004) in an IPO context. In general, studies have documented that decision-making is influenced by the perception of a price or valuation relative to a reference point. Kahneman and Tversky (1979, 1984, and 1986) suggest that reference points have also been used to explain managerial finance decisions. Shefrin and Statman (1984) apply Kahneman and Tversky s prospect theory to explain why so many firms accommodate investors by paying cash dividends. Fanto (2001) explains how a company s board is in a better legal position when planning to bid for a target if it uses the prices from recent similar mergers as reference points. Gervais (2009) investigates the impact of psychology biases in capital budgeting decisions, and finds that managers overinvest their firms free cash flows, engage in more acquisitions, and invest in more new projects. In a recent study that is most relevant for our purposes, Baker, Pan, and Wurgler (2012) show that target shareholders use the 52-week high as a reference point when negotiating the deal price. Target 2

3 shareholders perceive that their stock is capable of reaching its 52-week high again in the near future, and therefore extract a higher premium from bidders when the reference point is relatively high compared to the target s prevailing stock price. The authors also find that the probability of deal completion increases when the bidder offers a price that is above the target s 52-week high. We build on the application of reference point theory in a number ways. Our contribution revolves around four distinct points. First, we test whether a firm s reference point influences its appeal to bidders, as measured by the probability that it receives a bid. The findings by Baker, Pan, and Wurgler (2012) about the reference point influencing offer price and about deal completion are focused on targets that already have received a bid. They do not focus on whether a reference point can distinguish between firms that are targeted versus firms that are not. We argue that because targets whose prevailing stock prices are distant from their respective reference point expect a natural rebound, they may be less willing to accept bids, and are therefore less likely to be targeted. Second, we apply reference point theory to the perspective of the bidder. To our knowledge, this is the first time the concept of bidder reference point is used. Just as target stakeholders may consider a reference point when assessing their prevailing valuation when pursued by bidders, bidder managers may consider their own reference point (the bidder s 52-week stock price high) when deciding whether to pursue deals. They should be more willing to pursue deals when their prevailing stock price is near its reference point, especially when using stock as its method of payment. Third, we assess how the target reference point influences the likelihood that a target will attract a bid from its existing management versus an outside bidder. The reference point from the perspective of firm s own managers who consider a buyout is unique, because the managers are typically shareholders of the target firm. As shareholders, they may have a more optimistic assessment of their firm s stock price rebounding when it is much lower than its 52-week high reference point compared to the assessment of the stock s prospects by outside bidders. Thus, targets may be more likely to attract bids from existing management when their stock price is depressed relative to their 52-week high. 3

4 Fourth, we assess how the target and bidder reference points can influence the negotiation between parties in hostile versus friendly deals. We expect that hostile takeovers are more likely when the bidder s prevailing stock price is near its own reference point while the target s stock is far below its own reference point. Conversely, when the stock prices of the bidder and target are close to their respective reference points, both parties should be more willing to complete the deal, which should result in a friendly takeover. Based on our analysis, we find that a higher target 52-week high reference point relative to the target s prevailing stock price decreases the likelihood that a target will be pursued or acquired. Second, we find that a higher bidder 52-week high reference point relative to its prevailing stock price decreases the likelihood that it will bid for a target, but it has no effect on the probability of the merger completion. Third, we find that a higher target 52-week high reference point relative to its prevailing stock price increases the likelihood of a management buyout. Fourth, we find that a higher bidder 52-week high reference point relative to its prevailing stock price decreases the likelihood that the resulting merger will be hostile. Overall, our results offer strong evidence that target and bidder reference points influence the tendency to pursue deals and also influence the structure of the deals. II. Hypotheses Impact on Reference Point on the Likelihood of Receiving Bid Baker, Pan, and Wurgler (2012) apply the target reference point notion to test investors disposition effect relative to the 52-week high. Targets may be able to negotiate for better deals when their stock price is distant from their 52-week high. Extending this argument, we hypothesize that the distance between the target s prevailing share price and its 52-week high price influences the probability of the target receiving a bid. Targets whose prevailing stock prices are heavily discounted relative to their respective 52-week high reference points may resist takeovers because they believe they are presently undervalued. Conversely, targets with high prevailing stock prices (small difference between current share price and 52-week high reference point), may believe that their prevailing market valuation is 4

5 appropriate (not undervalued), and may be more willing to consider bids. Therefore, bidders that recognize the target s reference point philosophy may be less willing to pursue a target whose prevailing stock price is low relative to the 52-week high reference point; bidders may anticipate much resistance by targets with that profile. Furthermore, we argue that the distance between the target s share price and its 52-week high price influences the likelihood of deal completion. If targets with low stock prices relative to their 52- week reference points perceive themselves to be undervalued, they will resist bids unless the premiums are sufficiently large to offset their low market values. Conversely, targets with high prevailing stock prices relative to their 52-week high reference points should be more willing to accept a takeover bid, leading to the completion of the deal. Impact of Reference Point on the Likelihood of Bidding We adapt the reference point theory developed by Baker, Pan, and Wurgler (2012) to explain bidder behavior. Just as targets may refer to a reference point to assess their potential valuation, bidders may refer to their own reference point when assessing their own potential valuation. In particular, bidders may consider their own 52-week stock price high when deciding whether to pursue targets. If there is a large distance between the bidder s current stock price and its own 52-week high reference point, the bidder may be less willing to bid for targets, especially if it plans to use stock as payment for the acquisition. Under these conditions, it may view its stock price as temporarily depressed, and may be unwilling to pursue a stock financed acquisition until its stock price rebounds toward its higher reference point. Conversely, if there is a small distance between a bidder s current stock price and its 52-week high reference point, it may perceive its prevailing stock valuation to be more appropriate and acceptable. Impact of Reference Point on the Likelihood of an MBO Morck, Shleifer and Vishny (1988) argue that management buyouts (MBOs) are unique, because they are not motivated by potential synergies like many other mergers. They suggest that MBOs are triggered to align interests and incentives because of the increased management ownership and the 5

6 possibility of buying underpriced assets. Yet, Lee (1992) argues that managers do not propose buyouts of their own firms when they have more insider information. There is little research on the effect of psychological motives and determinants of MBO transactions. Fox and Marcus (1992) analyze leveraged management buyouts in the 1980s and address behavioral causes of MBOs. However, their focus is quite limited and only mentions behavioral characteristics such as insider information and reduced stakeholder commitment. Baruch and Woodward (1998) conclude that although there is significant stress associated with MBOs, managers are able to cope with it based on the interactive nature of the management team culture. We hypothesize that a firm s reference point can influence the probability that it is acquired by its own management as opposed to an independent bidder. A firm s managers, like the firm s other shareholders, may perceive that their stock is capable of reaching its 52-week high price again in the future. Thus, they may be more willing than other potential (unaffiliated) bidders to pursue a buyout when the firm s prevailing stock price is low relative to its 52-week high reference point. Conversely, managers of other potential bidder companies that are not attached to this firm may not share the same optimistic perspective that the firm s share price will naturally revert back toward its 52-week reference point. Therefore, the likelihood that the firm is subject to buyout by management instead of other outside bidders should be higher when the target s stock price is distant from its 52-week high reference point. Impact of Reference Point on the Likelihood of a Hostile Bid In general, transaction costs of a friendly merger are much smaller than the costs of a hostile acquisition. Nevertheless, firms commit to the hostile acquisition under specific conditions (see Schnitzer, 1996). When a bidder s price is close to its own reference point, it may perceive that its market valuation is accurate. It may attribute a high valuation to high performance, even if the performance is due to the performance of the overall economy. Therefore, when the distance between its prevailing stock price and its 52-week high reference point is small, the bidder should have more confidence in its ability to acquire other firms, resulting in more hostile acquisitions. 6

7 Meanwhile, a target should tend to resist bids when its own stock price is low relative to its 52- week high reference point (as explained earlier), which increases the likelihood of hostile bids. Conversely, a target should be more willing to entertain bids when its prevailing stock price is high relative to its own reference point, which would allow for an increased probability of a friendly deal. III. Sample We identify all mergers within SDC with a merger announcement date between January 1992 and December Only transactions in excess of $1,000,000 are included. We only include mergers, acquisitions, and acquisitions of majority interest. For the control variables that are not available in SDC, we obtain the data from Compustat Fundamentals Annual based on the date closest (before) the merger announcement date. The specific samples vary based on the hypothesis being tested, as described throughout this section. To test each of our hypotheses, we apply logistic regression analysis to determine whether each dependent variable that represents a particular profile (for example, whether the firm received a bid or not) is related to the distance between the firm s prevailing stock price and 52-week high (target and/or bidder) reference point. Sample Used to Assess Likelihood of Becoming a Target and Completing the Deal First, we investigate the impact of the target distance to reference point on the likelihood of a merger by comparing which firms received bids compared to firms that did not. To conduct our investigation, we match the SDC sample of firms that received a bid, regardless of the success of completion, with a sample of firms that did not receive bids. This is a one to one match. We match on year, industry (4 digit SIC code), and size (proxied by total assets). Firms that received a bid are assigned 1 and the matched firms that did not receive a bid are assigned 0. Since there were 701 firms that received 7

8 bids and had sufficient data for our analysis, our final sample (which also contains the matching firms that did not receive bids) consists of 1,402 firms 1 Next, we assess the likelihood of deal completion by comparing the firms that received a bid and closed the deal against those that received a bid but did not close the deal. For this purpose, we assign 1 to target firms that were successfully acquired, and assign 0 to target firms that did not close the deal. The final sample consists of 601 firms 2. Sample Used to Assess Likelihood of Management Buyout To determine whether the likelihood of a firm being subject to an MBO is influenced by its distance to its reference point, we create a dependent variable in which target firms that are subject to a MBO are assigned 1 and those that are pursued by independent bidders are assigned 0. The sample consists of 699 firms. This sample is derived from the original sample of 701 firms from above. Sample Used to Assess Likelihood of Becoming a Bidder and Completing the Deal To investigate whether the likelihood of a firm becoming a bidder is influenced by its distance to its reference point, we match the SDC sample of public firms that made a bid, regardless of the success of completion, with a sample of public firms that did not make bids. This is a one to one match. We match on year, industry (4 digit SIC code), and size (proxied by total assets). Firms that made a bid are assigned 1 and the matched firms that did not make a bid are assigned 0. We restrict the sample to stock-financed deals only. This sample consists of 408 firms (204 firms that pursued a target and a matched sample of 204 firms that did not). We also assess the likelihood of the bidder completing the merger by comparing the firms that made a bid and closed the deal against the firms that made a bid but did not close the deal. For this 1 This number refers to the number of firms used in the comprehensive model that includes target/bidder reference points and all control variables. It represents the most restricted model; less restricted models consist of larger sample sizes as shown in the tables. 2 The sample drops from 701 to 601 firms when we include both the target and bidder reference points and all control variables in the model. 8

9 purpose, we assign 1 to bidders that completed the deal, and 0 to bidders that did not complete the deal. Once again, we restrict the sample to stock-financed deals only. The final sample consists of 204 firms. Sample Used to Assess Likelihood of Hostile Takeover Finally, we assign 1 to the target firms that were involved in hostile takeovers and 0 to the target firms that were identified in SDC as friendly. The sample is composed of 601 firms. This sample consists of mergers that were announced, regardless of completion status. A summary of all samples used in the paper is presented in table I panel A and the breakdown of the sample by year, in table I panel B. IV. Variables Used Measurement of Reference Point Our primary independent variable in our models is the percent distance of the prevailing stock price of each firm in the sample prior to acquisition announcement from its 52-week high reference point. We define the target reference point as the high stock price the target company recorded within the 52 weeks before the merger announcement date. Following Baker, Pan, and Wurgler (2012), we use the 52- week target high price over the 365 calendar days ending 30 days prior to the announcement date obtained from CRSP 3. Studies such as Keown and Pinkerton (1981), McPhee et al. (2007), and Agrawal and Nasser (2012) show significant leakage in merger announcements and actual price of the target stock in advance of the announcement date. If insiders take advantage of their private information before the announcement, the 52-week high will most likely be near the announcement date. In order to avoid this bias, we use the 52-week high that ends 30 days before the announcement in the calculation of our primary independent variable. 3 we also consider an alternative measure as a robustness check, in which we extract the 52-week high from the SDC database that is calculated in reference to the announcement date. 9

10 As with the target reference point, we measure the bidder reference point as its 52-week high price over the 365 calendar days ending 30 days before the announcement date. Although the existing literature does not provide any proxies for the bidder reference point, the logic for choosing this specific measure is similar to that for choosing the target reference point. To avoid a biased reference point that could be affected by the insider and speculative trading, we use the 52-week high that ends 30 days before the actual merger announcement date. Control Variables Studies by Palepu (1986) and others identify several variables representing target, bidder, or other characteristics that may be important determinants of the likelihood of being acquired. These characteristics are incorporated as control variables in our analysis 4. Target s Managerial Effectiveness. The managerial effectiveness of the target is important because some acquisitions are motivated by the opportunity to improve a poorly performing firm [see Jensen (1988), Morck, Shleifer and Vishny (1988) or Rappaport (1990) for a discussion]. Palepu (1986) suggests that firms with a low ROE may attract bids. Following the literature, we use target s ROE in the year prior to the merger announcement date as a proxy for managerial effectiveness (TMANEFF), adjusted for the industry performance. Target s Size. The size of the target should be inversely related to the likelihood of acquisition [see Palepu (1986), Ambrose and Megginson (1992) and Powell (1997)]. Powell (1997) explains the inverse relationship by the increased costs needed to absorb a larger target into the structure of the bidder s organization. In addition, a large target may have more resources and, therefore, more power to resist the takeover. Small targets may be most sensitive to the target reference point, as the information asymmetry is higher for small firms. As there is less information available to derive a correct assessment about the company, small firms will more heavily rely on the reference point. By comparison, large 4 Target-specific variable names begin with the letter T while bidder-specific variable names begin with the letter B. 10

11 targets have more information available about them, which should allow prospective bidder firms to more easily assess their true value. We use the log of total assets as a proxy for target size (TSIZE). For samples that involve targets and bidders, we also consider the size of the target relative to the size of the bidder. A small bidder may have a lower probability of acquiring a target of a given size, because its resources are smaller than those of a larger bidder. We measure the relative size (RELSIZE) as the ratio of the log of target s total assets to bidder s log of total assets. Target s Stock Price Performance. Akhigbe, Martin, and Whyte (2007) argue that the stock performance is indicative of existing or expected operating performance of the firm. Value-maximizing bidders may be more interested in strong-performing targets that have a higher chance of adding value. Berger and Ofek (1996) state that firms that diversify and, as a result, lose value have a higher chance of being acquired by other firms in a break-up acquisition. Although that study specifically focuses on diversification losses, the conclusion resonates with a number of studies conducted in the late 1980s and early 1990s. The results of those studies show that firms with weak market valuations are more likely to be acquired. We use a control variable to proxy for the target stock performance (TSTPERF), based on its return over the previous year. Target s Market-Book Ratio. High market-book ratios coincide with increased merger activity and this is especially true for stock-financed mergers (see Maksimovic and Phillips (2001) and Jovanovic and Rousseau (2001)). Rhodes-Khoph and Viswanatghan (2004) argue that this relationship can either be explained by a more efficient deployment of assets or by overvaluation. In addition, growth is also related closely to the risk of the firm [see Cao, Simin and Zhao (2008), Bernardo, Chowdhry and Goyal (2007), Carlson, Fisher and Giammarino (2004, 2006), and Lantz and Sahut (2005)]. Some firms may achieve their growth goals by purchasing targets with high growth prospects. We use the target s Market-Book Ratio (TMB) as a proxy for growth, calculated as the market price 30 days before the announcement to the latest book value of the target before the merger announcement. 11

12 Target s Financial Leverage. Many theories are used to explain the relationship between target financial leverage and acquisition activity. Among them are tax shields, the existence of financial slack, and wealth transfers. According to Loughran (1997), including debt in acquisition analysis is important, because it changes the value of the enterprise. Firms that have higher financial leverage are more likely to merge with cash-rich bidders who would take care of their financial problems. Yet, a counter argument is that bidders may try to avoid targets with high financial leverage, as they know they will be responsible for the high debt burden of the target after the completion of the deal. We measure target leverage (TLEV) as the total debt to total assets ratio. Target s Liquidity. Firms with high liquidity may be more desirable because they should require less financial support for covering their short-term obligations. Yet, they might also be more willing to entertain offers if they cannot easily to meet their current financial obligations. The target s liquidity (TLIQ) is measured as the ratio of cash and marketable securities to total assets. Industry Tendencies. Targets in some financial and tech industries are more susceptible to acquisitions. Given that the recent financial crisis is closely associated with the financial firms and the high technology firms have been shown to differ in many aspects (such as growth and types of assets) from other sectors, we control for the finance and technology sectors analysis. We designate targets with a dummy variable (TFIN) set equal to 1 if the target is in the financial industry and 0 otherwise. We designate targets in the technology industries (TTECH) with a dummy variable set equal to 1 if the target is in one of the following SIC codes: 3571, 3572, 3575, 3577, 3578 (computer hardware), 3661, 3663, 3669 (communications equipment), 3674 (electronics), 3812 (navigation equipment), 3823, 3825, 3826, 3827, 3829 (measuring and controlling devices), 4899 (communication services), or 7370, 7371, 7372, 7373, 7374, 7375, 7379 (software). Target P/E Heuristic. According to Shefrin (2007), the P/E ratio is used by financial analysts and financial executives to value companies more than any other technique. When considering mergers, bidders apply the P/E valuation shortcut to estimate the value of the target firm. Firms that have a high 12

13 P/E ratio are most likely to attract more bids, increasing the probability of acquisition. We measure the target s PE ratio (TPE) as the most recent price/earnings ratio before the acquisition announcement. Target s Growth in Revenue or EPS. Targets may be especially appealing when they have high growth (TGROWTH) in revenue or in EPS. Shefrin (2007) suggests that revenue growth is a measure of excessive optimism. Bidder s Recent Stock Price. Bidders whose stock performed well in recent periods may be more willing to buy targets, regardless of their performance. There is a direct relationship between the bidder stock performance and the probability of acquisition. We use the stock performance of the bidder over the year before the merger announcement (BSTPERF) up to 30 days before the merger announcement as the proxy for bidder stock performance. Bidder s Liquidity. The bidder s liquidity is viewed as a source of financial resource availability. A bidder with a large reserve of cash may be more willing to pursue a target and close a deal, even in absence of other good reasons. Almeida et al. (2007) find that firms that are not financially stable are acquired by liquid firms within the same industry even when there are no synergistic reasons for the merger. We measure bidder s liquidity (BLIQ) as its ratio of cash and marketable securities to total assets. Type of Bidder. Akhigbe, Martin, and Whyte (2007) include the type of the bidder (corporate vs. non-corporate) as a control variable in their study about the likelihood of a partial acquisition leading to a full acquisition. They argue that corporate bidders are more likely to be strategic bidders and, as a result, they are more committed to the acquisition. Private equity (financial) buyers may attach less sentimental value and will make negotiation of the deal a priority. We measure the type of bidder (BTYPE) with a dummy variable that is set equal to 1 for a corporate bidder, and zero for a financial bidder such as a private equity firm. Bidder Overconfidence. The bidder s experience with acquisitions may cause overconfidence, and therefore increase the probability of acquisition. Managers that have been involved in a prior acquisition may be more confident that they can create a successful merger. A counter argument is that previous bidders are consumed with integrating their target, or learned from experience how mergers can 13

14 backfire, which could discourage future mergers. We proxy for the confidence level of the bidder (BEXPER) based on its prior participation in other mergers. We assign a 1 to firms that have previously (since 1982) acquired at least one other firm, and 0 otherwise. In addition to the financial and high-tech firms being different, we control for the overall performance of the industry. Some industries display countercyclical behavior. Other industries are routinely experiencing consolidations and are more prone to acquisitions. Therefore, we control for the performance of the industry (INDPERF) by using the monthly performance of the Fama-French 48 portfolios over the last year. The hostility of the deal is included in some of the analysis. Hostile deals (as indicated by a dummy variable HOSTILE) are more likely to be resisted by targets and as a result, reduce the probability of acquisition. Friendly bids, on the other side are more likely to be welcomed by targets, resulting in a higher probability of acquisition. Finally, in some other specifications, we control for the method of payment by including the percent of equity used in the final deal (EQUITY). V. Models and Results Impact of Reference Point on Likelihood of Becoming a Target To determine whether the target reference point influences the firm s likelihood of attracting bids, our analysis is intended to distinguish between public firms that become targets versus those that do not. We first identify a set of control firms (from Compustat) that are matched to the target firms by the same year, industry, and size, but were not targeted. Then, we apply the following logit model: Prob(Firm becomes Target) =α+β 1 TDISTANCE+ β 2 TMANEFF + β 3 TSIZE + β 4 TSTPERF + β 5 TMB+ β 6 TLEV+ β 7 TLIQ+ β 8 TFIN+ β 9 TTECH+ β 10 TPE+ β 12 TGROWTH+ β 13 INDPERF + ε (1) Results from applying our model are disclosed in table II, panel A. The model results in the first column isolate the effect on TDISTANCE, our primary variable of concern. The second column includes TDISTANCE along with all the control variables. The complete model consists of 1,402 observations and is overall significant with a p-value of and a pseudo r-square if

15 The coefficient of TDISTANCE is consistently negative and significant, which supports the hypothesis that the likelihood a firm is chosen as a target is inversely related to the target s distance to its reference point. To the extent that targets with stock prices more distant from their reference point negotiate for larger premiums (as implied by Baker, Pan, and Wurgler (2012)), they are less appealing as targets to prospective bidders. Regarding the control variables, target stock performance (TSTPERF) is inversely related to the firm s likelihood of receiving a bid, providing some indication that weaker firms are more susceptible to takeover offers. We also find some evidence that a firm s target leverage (TLEV) is positively related to the probability of becoming a target. Yet, we also find that TLIQ is positively related to the probability of a firm becoming a target, implying that firms with higher liquidity are more likely to attract bidders. The TECH variable is negative and significant at the.10 level, offering modest evidence that firms in the tech industry are less susceptible to bids when controlling for many other characteristics. Finally, the coefficient of the target P/E variable (TPE) is positive and significant, which suggests that firms with higher P/E ratios are more likely to attract bids. Impact of Target Reference Point on Deal Completion Next, we focus solely on the subsample of targeted firms from our previous analysis (by removing the control firms), and record the deal completion outcome (completion or withdrawal) for each target. To determine whether the deal completion outcome is related to the target s distance to its reference point along with other control variables, we apply the following logit model: Prob(Deal Completion)= α+β 1 TDISTANCE+ β 2 TMANEFF + β 3 TSIZE + β 4 RELSIZE+ β 5 TSTPERF + β 6 TMB+ β 7 TLEV+ β 8 TLIQ+ β 9 TFIN+ β 10 TTECH+ β 11 TPE+ β 12 TGROWTH+ β 13 BSTPERF + β 14 BLIQ + β 15 BTYPE + β 16 BEXPER + β 17 INDPERF + β 18 HOSTILE+ β 19 EQUITY +ε (2) Results are presented in table II, panel B. Results isolate the effects of TDISTANCE in the first column, include target control variables in the second column, and include all control variables in the third column. All regression models are overall significant with pseudo r-squares as high as 11.16%. 15

16 The coefficient of TDISTANCE ranges between ** (p-value of 0.015) and * (0.066), and is significant at the.10 level or better in all three models. These results support our hypothesis that the target s distance to its reference point at the time of the bid is inversely related to the likelihood that the deal will be competed. They complement the prior findings of Baker, Pan and Wurgler (2012), who found that the probability of a successful deal increases when the bidder offers a price that is at or above the target reference point. In addition to the target distance, other target and bidder characteristics contribute to the likelihood of a merger being completed successfully. We find limited evidence of an inverse relationship between target firm s management effectiveness (TMANEFF) and the probability of completion. Targeted firms that have a low ROE are more likely to experience deal completion (rather than withdrawal). We also find strong evidence that firm size is inversely related to deal completion. In addition to the target size, the relative size of the target to the bidder is inversely related to the probability of deal completion. The larger is the target compared to the bidder, the smaller is the probability that the merger will be successful. Hostile deals are less likely to be completed, which is not surprising given that the target is more likely to oppose hostile bids. Finally, target firms that are either in the technology (TTECH) or finance (TFIN) industries have a higher likelihood of deal completion. Regarding the bidder characteristics, the coefficient of BEXPER is negative, which suggests that deal completion is less likely when the bidder firm has more merger experience. This may be attributed to a learning experience that makes the bidder more cautious or to continuing issues that the experienced bidder has in integrating the target from its previous deal(s). Finally, we find some evidence that the better the overall performance of the industry (INDPERF), the more likely is the merger to be completed successfully. 16

17 Impact of Target Reference Point on Likelihood of MBO To test whether the target s distance to its reference point influences the likelihood of a management buyout (MBO) versus an acquisition by an unaffiliated bidder, we apply the following logit model to the sample of target firms that was used in the previous analysis: Prob(MBO)= α+β 1 TDISTANCE+ β 2 TMANEFF+ β 3 TSIZE+ β 4 TSTPERF + β 5 TMB+ β 6 TLEV+ β 7 TLIQ+ β 8 TFIN+ β 9 TTECH+ β 10 TPE+ β 11 TGROWTH+ β 12 INDPERF +ε (3) Results are displayed in table III. Results isolate the effects of TDISTANCE in the first column, and include all control variables in the second column. Both regression models are overall significant with pseudo r-squares of and The coefficient for the TDISTANCE variable is positive and highly significant in both models. The coefficients are and 0.060, respectively. The results suggest that the further is the target s stock price at the time of announcement from its reference point, the higher is the likelihood that the source of the bid will be the target s management. Managers are more willing to submit a bid than outside firms under these conditions, because as existing stakeholders and potentially, holders of private information, they may have more faith that the firm s stock price will eventually revert to its reference point. Regarding the control variables, the coefficient of TMB is positive and significant, which suggests that targets with higher market-book ratios are associated with a higher likelihood of an MBO. Conversely, coefficients of the variables representing target P/E ratio (TPE) and target growth (TGROWTH) are negative and significant, which suggest that likelihood of a management buyout is higher for targets that have low fundamental multiples. We also find evidence that the probability of an MBO is positively associated with industry performance. Impact of Reference Point on Likelihood of Becoming a Bidder To assess the effects of a firm s distance to its reference point on its likelihood of becoming a bidder, we identify control firms (from Compustat) that are matched to bidder firms by year, industry, and 17

18 size, but did not become bidders. Then we apply the following logit model to isolate the effects of the bidder reference point variable (BDISTANCE), while controlling for other bidder-specific variables: Prob (Firm becomes Bidder) =α+β 1 BDISTANCE+ β 2 BSTPERF + β 3 BLIQ+ β 4 BTYPE + β 5 BEXPER+ β 6 INDPERF+ ε (4) Results of our analysis are presented in table IV, panel A. Results isolate the effects of BDISTANCE in the first column and include all control variables in the second column. The coefficient of BDISTANCE is and , respectively and is negative and significant in both models. Results support our hypothesis that the likelihood of a firm becoming a bidder increases when the bidder s distance to its respective reference point is shorter. Regarding control variables, BLIQ is negative and significant, which implies that firms with less liquidity were more likely to become bidders (possibly to access the liquidity of the targets). In addition, BEXPER is positive and significant, which implies that firms with previous acquisition experience are more likely to become bidders. Impact of Bidder Reference Point on Likelihood of Deal Completion Next, we focus solely on the subsample of firms that submitted bids from the previous analysis (by removing the control firms), and record the deal completion outcome (completed or withdrawn) for each bid. We test whether the deal completion outcome of the bid is related to the bidder s distance to its respective reference point, along with control variables, by applying the following logit model: Prob(Deal Completion)= α+β 1 BDISTANCE+ β 2 STOCK + β 3 STOCK*DIST + β 4 TMANEFF+ β 5 TSIZE+ β 6 RELSIZE+ β 7 TSTPERF + β 8 TMB+ β 9 TLEV+ β 10 TLIQ+ β 11 TFIN+ β 12 TTECH+ β 13 TPE+ β 14 TGROWTH+ β 15 BSTPERF + β 16 BLIQ + β 17 BTYPE + β 18 BEXPER + β 19 INDPERF + β 20 HOSTILE+ β 21 EQUITY +ε (5) Results of our analysis are disclosed in table IV, panel B. The coefficient of BDISTANCE is not significant. Thus, we conclude that once the bid has been announced, the bidder reference point does not have an impact on the probability that the deal will be completed. In addition, the variables representing the percentage of stock planned for the deal and the interaction between the percent of stock planned for the deal and the bidder reference point are also not significant. These results are not surprising. Once the 18

19 bidder commits to the acquisition, it is less likely to reverse its decision, regardless of how far its stock price is from its reference level. A number of characteristics are associated with the probability of a firm completing a merger successfully. The coefficient of TSIZE is negative and significant, which suggests that deal completion is more likely when smaller firms are targeted. TLIQ and BLIQ are both negative and significant, which suggest that deal completion is more likely when the bidder and target have lower levels of liquidity. Deal completion is also more likely when the bidder s recent stock performance (BSTPERF) is more favorable, while deal completion is less likely when the bidder has acquired a firm in the past (BEXPER). The coefficient of HOSTILE is negative, which suggests that hostile bids result in a lower probability of deal completion. Joint Impact of Target and Bidder Reference Points We extend our analysis by focusing on the joint impact of the bidder and the target reference points on the probability of deal completion. For this purpose, we focus on a sample that consists of all merger announcements that were either completed or withdrawn, for which we can calculate both the bidder and the target reference points. Then and we apply the following logit model: Prob(Completed Deal)= α+β 1 DIFDIF+ β 2 TMANEFF+ β 3 TSIZE+ β 4 RELSIZE+ β 5 TSTPERF + β 6 TMB+ β 7 TLEV+ β 8 TLIQ+ β 9 TFIN+ β 10 TTECH+ β 11 TPE+ β 12 TGROWTH+ β 13 BSTPERF + β 14 BLIQ + β 15 BTYPE + β 16 BEXPER + β 17 INDPERF + β 18 HOSTILE+ β 19 EQUITY +ε (6) When the bidder and the target simultaneously have reference points close to their respective current stock prices, both parties should be more willing to complete the deal. Conversely, when one party s stock price is far from its respective reference point, it may find the deal less appealing because it sees itself as being at more of a disadvantage compared to its counterparty. We build a new variable called DIFDISTANCE (differences in differences) which is calculated as TDISTANCE minus BDISTANCE to examine the joint impact of the bidder and target reference points on the probability of acquisition. The results are presented in table V. 19

20 The pseudo R-squared of the complete model (with control variables) is and the overall model is highly significant with a p-value of The coefficient DIFDISTANCE is consistently negative and significant, which supports our hypothesis that as the differences in differences decrease, the probability of deal completion increases. In addition to the DIFDISTANCE, a number of other factors are significant, and contribute to the probability of the merger being successful. The coefficient of TSIZE is negative, indicating a higher likelihood of deal completion when the target size is relatively small, while TSTPERF is positive, indicating that targets with relatively high recent stock price performance are more likely to attract bids. Financial target firms (TFIN) are more likely to complete the deal. Targets exhibiting relatively high P/E heuristics are less likely to complete the deal. Hostile deals are less likely to be successful and the overall performance of the industry contributes positively to the probability of a successful acquisition. Impact of Target and Bidder Reference Points on Likelihood of Hostile Takeovers To determine whether target and bidder reference points are associated with hostile takeovers, we use the same sample as in the previous analysis (announced mergers that were either completed or withdrawn), and apply the following logit model: Hostile Bid= α+β 1 TDISTANCE+ β 2 BDISTANCE + β 3 TDISTANCE*BDISTANCE + β 4 TMANEFF+ β 5 TSIZE+ β 6 RELSIZE+ β 7 TSTPERF + β 8 TMB+ β 9 TLEV+ β 10 TLIQ+ β 11 TFIN+ β 12 TTECH+ β 13 TPE+ β 14 TGROWTH+ β 15 BSTPERF + β 16 BLIQ + β 17 BTYPE + β 18 BEXPER + β 19 INDPERF +ε (7) Results of our analysis are presented in table VI. The coefficient of TDISTANCE is consistently negative, but insignificant. Thus, there is no convincing evidence that TDISTANCE is related to the probability that the merger will be hostile. However, the coefficient of the bidder distance variable (BDISTANCE) is consistently negative and significant in all both models, which suggests that the likelihood of a hostile merger is higher when the bidder s distance to its 52-week high reference point is short. These results for BDISTANCE suggest that when a bidder s stock price close to its reference point, it is more confident and therefore more willing to engage in a hostile bid. 20

21 In addition, the probability of a hostile bid is higher when the target s size (TSIZE) is smaller, when the target s recent stock performance (TSTPERF) is stronger, and when it is in the technology or finance sector. There is also modest evidence that a hostile bid is more likely when the bidder has greater liquidity (BLIQ). VI. Conclusion We adapt the theory established by Baker, Pan, and Wurgler (2012) to investigate how reference points influence the likelihood of a firm being acquired, the likelihood that a targeted firm is subject to a management buyout, the likelihood of a firm implementing a bid, and the likelihood that a targeted firm is subject to a friendly versus hostile takeover. We find that the firm s distance to its 52-week high reference point is inversely associated with its likelihood of becoming a target. This result is distinguished from the foundation established by Baker, Pan, and Wurgler (2012) because it is focused on the impact of reference point on whether a firm attracts a bid, while Baker, Pan, and Wurgler (2012) focused on offer price and completion of the merger after the bid was offered. We also confirm that the probability of a successful deal completion is inversely related to the distance to the target reference point. Next, we investigate whether the reference points influence the likelihood that the target is subject to a management buyout. We find that the probability of a MBO versus a bid from an unaffiliated bidder is associated with the target reference point. The larger is the distance between the stock price and the reference point, the higher is the likelihood that the merger will occur in the form of an MBO. This result supports our hypothesis that managers of a firm with a weak stock price (relative to its respective 52-week high) may have more confidence than unaffiliated prospective bidders about its future prospects. We also develop a theory that like the target, the bidder relies on its own reference point to make merger decisions. We find that a firm s reference point is a significant determinant of the probability that it will pursue a merger. The smaller is the distance between the bidder stock price at the time of the announcement and the bidder high price in the 52 weeks before the merger announcement, the higher is the probability that the firm will become a bidder. However, unlike the target reference point, once the bid 21

22 has been announced, the bidder reference point does not have an impact on the probability that the merger will be completed. Then, assess the simultaneous impact of the target and bidder reference point on the probability of the a successful acquisition by using a variable that represents the difference in differences (between the target distance to its own reference point versus the bidder distance to its own reference point). We find that when both the bidder and the target have reference points close to their respective current stock prices, both parties are more willing to complete the deal. We attribute this result to how a distant stock price from the reference point by the target or bidder may inflict a psychological barrier on negotiations, thereby preventing an agreement between parties. Finally, we test whether the target and bidder distance to their respective reference points influence whether a bid is hostile versus friendly. We find no relationship between the target s distance to its reference point and the likelihood of a hostile deal. However, we find that when bidder s stock price is closer to its reference point, the bid is more likely to be hostile. Overall, results from our analyses offer strong evidence that target and bidder reference points serve as potent anchors that shape the outcome and the structure of mergers. References Agrawal, A. and T. Nasser, 2012, Insider Trading in Takeover Targets, Journal of Corporate Finance, 18 (3), Akhigbe, A., Martin, A. and A.M. Whyte, 2007, Partial Acquisitions, the Acquisition Probability Hypothesis and the Abnormal Returns to Partial Targets, Journal of Banking and Finance, 31 (10): Almeida, H., and M. Campello, 2007, Financial Constraints, Asset Tangibility and Corporate Investment, Review of Financial Studies, 20: Ambrose, B. and W. Megginson, 1992, The Role of the Asset Structure, Ownership Structure, and Takeover Defenses in Determining Acquisition Likelihood, The Journal of Financial and Quantitative Analysis, 27 (4): Baker, M., Xin Pan, and J. Wurgler, 2012.The Effect of Reference Point Prices on Mergers and \ Acquisitions, Journal of Financial Economics (forthcoming). Barberis, N.s and W. Xiong, 2009, What Drives the Disposition Effect? An Analysis of a Long-Standing 22

23 Preference-Based Explanation, The Journal of Finance, 64 (2): Baruch,Y. and S. Woodward, 1998, Stressful situations? The case of management buyout/buyins, Management Decision, 36 (10): Berger, P. and E. Ofek, 1996, Bustup Takeover of Value-Destroying Diversified Firms, The Journal of Finance, 51 (4): Bernando, A., Chowdhry, B. and A. Goyal, 2007, Growth options, Beta and the Cost of Capital, Financial Management, 36 (2): Cao, C., Simin, T. and J. Zhao, 2008, Can Growth options Explain the Trend in Idiosyncratic Risk?, The Review of Financial Studies, 21 (6): Carlson, M., Fisher, A. and R. Giammarino, 2006, Corporate Investment and Asset Price Dynamics: Implications for the Cross Section of Returns, The Journal of Finance, 59 (6): Harford, J., 2005, What Drives Merger Waves?,Journal of Financial Economics, 77: Fanto, J. A., 2001,Quasi-Rationality in Action: A Study of Psychological Factors in Merger Decision- Making, 62 Ohio St. L.J: Fox, I. and A. Marcus, 1992, The Causes and Consequences of Leveraged Management Buyouts, The Academy of Management Review, 17 (1): Gervais, S. and I. Goldstein. 2007, The Positive Effects of Biased Self-Perceptions In Firms, Review of Finance 11(3): Grinblatt, M. and M. Keloharju, 2001, What makes investors trade? Journal of Finance, 56 (2): Huddart, S., Lang, M. and M. Yetman, Volume and Price Patterns Around a Stock s 52-week Highs and Lows: Theory and Evidence, Management Science, 55 (1): Huddart, S., Lang, M. and M. Yetman, Volume and Price Patterns Around a Stock s 52-week Highs and Lows: Theory and Evidence, Management Science, 55 (1): Jovanovic, B., and P. Rousseau, The Q-Theory of Mergers, American Economic Review, 92 (2): Kahneman, D. and A. Tversky, 1979, Prospect Theory: Theory: An Analysis of Decision Under Risk. Econometrica, 47: Kahneman, D. and A, Tversky, 1984, Choices, Values and Frames, American Psychologist, 39: Kahneman, D. and A. Tversky, 1986, Rational Choice and the Framing of Decision, The Journal of Business, 59(4): Kaustia, M., 2004, Market-Wide Impact of the Disposition Effect: Evidence from IPO Trading Volume, Journal of Financial Markets, 7 (2):