The Gains from Contracting with Equity. Myron B. Slovin Department of Finance Louisiana State University Baton Rouge, LA 70803

Transcription

1 The Gains from Contracting with Equity by Myron B. Slovin Department of Finance Louisiana State University Baton Rouge, LA Marie E. Sushka Department of Finance Arizona State University Tempe, AZ John A. Polonchek Department of Finance Oklahoma State University Stillwater, OK February 9, 2003 Address correspondence to: Professor Marie E. Sushka Department of Finance Box W. P. Carey College of Business Arizona State University Tempe, AZ Telephone (480) FAX (480)

2 The Gains from Contracting with Equity Abstract We analyze intercorporate asset sales that entail the use of equity as a medium of exchange. These transactions are valuing-enhancing for both sellers and buyers, with returns of 3% and 10%, respectively, in contrast to deleterious effects on acquirers reported for mergers when equity is the means of payment. There are strongly positive returns to buyers even when the shareholding created by the asset sale is less than 5% of buyer shares, enhancing the view that equity is an effective medium of contracting. An asset for equity sale is not a precursor to a seller acquisition of the buyer, and buyers are acquired by third parties at a rate comparable to control firms. Thus, the blockholdings created neither hinder nor enhance buyer openness to the market for corporate control. Positive announcement returns and improved subsequent operating performance of buyers suggest that contracting with equity induces sellers to be effective monitors and to share in the future improved performance of buyers that acquire the control of these assets.

3 The Gains from Contracting with Equity In this paper we analyze a set of control transactions that entail contracting with equity, more specifically, intercorporate sales of operating assets in exchange for buyer equity, events that we denote as asset for equity sales. We find that these transactions are value enhancing for both buyers and sellers, a finding that is in contrast to results established in the literature that the use of equity as the means of payment to acquire a public firm has a deleterious impact on the value of the acquirer relative to the use of cash. Our investigation analyzes the effects of using buyer equity as a medium of exchange within a context in which both the buyer and seller remain independent, and the seller becomes an equity blockholder in the buying firm. Since an asset for equity sale creates corporate blockholders of varying sizes, our study also addresses ownership structure issues that provide additional insight into the value of contracting with equity. Our results indicate that there are large, significantly positive wealth effects for both parties at the announcement of asset for equity sales, with average excess returns of approximately 3% for sellers and 10% for buyers. This evidence is in broad accord with Hansen s (1987) asymmetric information model in which buyer equity can be an efficient form of contracting that allows a seller to share in future gains to the buyer, while permitting the buyer to hedge its risk of overpaying for an asset. We find that the magnitude of the transaction and the size of the block of buyer equity that the seller receives have significant effects on buyer value. Other characteristics of these transactions, such as the presence of a standstill agreement or board representation for the seller, have no significant impact on the value of buyers or sellers. Our study evaluates the sources of the positive wealth effects from these transactions. One, asset for equity transactions create corporate holdings of blocks of shares, and the size of these holdings has a significant, nonlinear effect on buyer returns. Morever, for the more than 20% of the sample in which the size of the block is small so that no 5% blockholding is created, the average excess return is quite large, 9%, indicating that the use of equity as a medium of exchange is an important factor in enhancing buyer value. Two, we find that an asset for equity transaction is rarely the precursor of an acquisition of the buyer by the seller. Moreover, the frequency of acquisitions of buyers by third parties closely matches that of a set of

4 2 control firms, implying that asset for equity sales do not hinder or enhance the buyer s openness to the market for corporate control. Thus, the gains to buyers do not reflect a market expectation that these transactions are a precursor to a subsequent control event. These findings support the view that equity can be an efficient medium for intercorporate contracting. Since both parties to the transaction continue to remain independently traded public firms, we can address two additional issues: the post-event performance of the participant firms and the ultimate disposition of the equity block. We find positive changes in buyer operating performance after the asset is acquired and a blockholding is formed, suggesting that the increase in buyer value at the announcement of these transactions capitalizes the market s anticipation of future improvements in buyer performance. This evidence gives credence to the view that seller firms expect certain assets to be more valuable under the direction of another corporation than being retained, and thus sellers find it advantageous to participate in a contracting arrangement in which they obtain a share in the improved subsequent performance of the buyer. We also examine the ultimate disposition of the block of shares in the buyer. Approximately onequarter of the buyers are acquired, and in a similar proportion of cases the seller continues to hold the equity as of the end of the period of this study. For the remaining half of the sample, the seller subsequently dissolves the blockholding established by the asset for equity sale through one of four alternative types of transactions, secondary stock offerings, spin-offs, repurchases, and block trades. The average seller return at the announcement of its decision to dispose of the equity block in the buyer is significantly positive. Buyer returns are typically negative, but differ in sign by the type of disposition transaction. Overall, these results suggest that the financial market views the selling firms as corporate blockholders that effectively monitor buyer firms and counter agency difficulties that can result from the separation of ownership from control, rather than viewing sellers as likely acquirers of buyer firms. The large increases in overall shareholder wealth that we document are consistent with the view that an asset for equity sale shifts control of an operating asset to a new team of corporate managers that enhances the productivity of the asset, and allows efficient risk-sharing between buyers and sellers. The

5 3 large gains in buyer shareholder wealth suggest that such transactions, which entail the use of buyer equity as a means of payment, are an effective contracting mechanism that conveys valuable private information about the buyer, and induce seller firms to monitor buying firm managers. Thus, we conclude that the use of equity as a means of payment in these intercorporate asset sales is not the deleterious signal of buyer value that it is in mergers and acquisitions of public companies, and that the use of buying firm equity as a currency for such sales is value enhancing over a broad range of the blockholdings formed in these transactions. The remainder of the paper is organized as follows. In Section I we discuss hypotheses and previous research in areas of corporate finance that are important background for the issues we examine. In Section II we describe sample development and methodology. In Section III we report event study evidence, cross-sectional regressions that explain share price reactions as a function of characteristics of the transactions and participating firms, analysis of the subsequent operating performance of buyers, and the disposition of the equity blocks formed by these transactions. Conclusions are found in Section IV. I. Asymmetric Information and Agency Considerations in Asset for Equity Sales There is considerable evidence (Travlos (1987), Franks, Harris and Mayer (1988), Asquith, Bruner and Mullins (1987), and Servaes (1991)) that the use of buyer equity as a means of payment in the acquisition of a public firm has a more deleterious effect on bidding firm value than the use of cash. Several theoretical models explain this empirical regularity based on asymmetric information and agency considerations. Research on another type of intercorporate transaction, that is, asset sales (Klein (1986), Hite, Owers, and Rogers (1987), and Lang, Poulsen, and Stulz (1995)), does not consider the issue of alternative means of payment, and instead focuses on transactions in which cash or cash-related instruments are the means of payment. These prior studies of asset sales report that shareholders of seller firms earn statistically significant gains, while the returns to buyer firms are typically small and not statistically significant.

6 4 There are several reasons that motivate our study of asset for equity sales. One, these sales provide a benchmark to assess existing findings in the mergers and acquisitions literature about the effects of the means of payment on firm value. Two, an asset for equity sale results from private negotiations between managers of the buying and selling firms, acting pursuant to the business judgment rule which avoids the extensive mandates for disclosure and shareholder approval that apply to mergers and acquisitions of public companies, giving rise to the potential for agency difficulties. Three, in asset for equity transactions the seller accepts an equity interest in the buyer as payment for the asset, indicating that both parties to the transaction agree to establish the seller as a corporate blockholder in the buyer. In this regard, it is important to understand that an asset for equity sale is not a type of earnout transaction since the seller of the asset does not retain any residual interest per se in the performance of the operating asset, which is definitively transferred to the buyer. Instead, the seller obtains an equity interest in the buying firm as a whole. A. Asymmetric Information and Asset for Equity Sales In this section, we draw on existing asymmetric information models about mergers and acquisitions of public companies that provide alternative perspectives as to whether the use of equity as the means of payment in an asset sale is value enhancing or deleterious for firm value. This literature is used to generate predictions for our empirical tests of asset for equity sales and allows us to frame empirical work that can discriminate among alternative hypotheses. Myers and Majluf (1984) model security issuance in an environment in which managers are better informed about the value of the firm than outside investors. As a result, managers seeking to maximize the wealth of existing shareholders have an incentive to issue equity when they believe that their firm s shares are overvalued. Thus, managers of acquiring firms that view their stock as overvalued have an incentive to use equity as the means of payment, while acquirers that regard their stock as undervalued prefer to use cash. Financial market participants recognize these incentives, and reduce the value of the acquiring firm in response to news of an acquisition using equity. Berkovitch and Narayanan (1989) also develop a model in

7 5 which low value acquirers signal their quality through stock offers. Consistent with these hypotheses, Travlos (1987), Franks, Harris, and Mayer (1988), Asquith, Bruner, and Mullins (1987), and Servaes (1991) report that acquirers of public firms that use equity as the means of payment sustain negative excess returns that are unfavorable relative to the returns that accrue to acquirers that use cash. From the perspective of this research, we should observe unfavorable share price effects on buyers that utilize equity as the means of payment in an asset sale. In Hansen s (1987) asymmetric information model of acquisitions, both the bidder and seller have private information about the value of their own firms. His model implies that bidder common stock can be an optimal means of payment due to elements of contingent pricing that are intrinsic to equity. Selling firm managers with favorable asymmetric information about their firm s assets and the expected future productivity of the assets when combined with the acquirer s operations, will prefer to receive buyer equity in payment to capture part of the increase in the value of the buyer firm that they anticipate will ensue after the transaction is completed and relevant cash flows are revealed to financial market participants. In contrast, seller firm managers with negative asymmetric information prefer to receive cash to avoid future losses when the true valuation is revealed. Thus, the use of equity as the means of payment is a selfselection mechanism. In this informational setting, the acquirer bears the entire cost of overpayment in a cash offer, while a seller that accepts buyer equity bears some of the post-acquisition revaluation risk associated with the transaction. As a result, the use of buyer equity is a positive signal about the target s value as part of the consolidated firm, including expected synergies. As the target s size increases relative to the buyer, this effect should become more pronounced. From the perspective of Hansen s model, an asset for equity sale should enhance both selling firm and buying firm value relative to a cash transaction. In assessing the sharing of risk, an asset for equity sale has several key differences from an earnout transaction, a type of asset sales contract that also entails risk sharing about ex post value between a buyer and a seller. In an earnout, the seller receives cash payments subsequent to the sale that depend on the performance of the relevant asset, measured along a dimension and a reporting period negotiated between

8 6 the parties. Thus, bargaining about the terms of an earnout is complicated by the absence of a common time horizon for the buyer and seller, a difficulty that creates an incentive for strategic, opportunistic behavior. In effect, an earnout is a complex state-contingent contract in which the seller seeks to have the operating performance of the unit maximized over the relevant period specified in the contract. The buyer, in contrast, is concerned with the value of the unit over a longer time horizon than the seller, and thus has an incentive to take actions that maximize the unit s long term value but that may reduce its reported operating performance during the period relevant for the earnout payments. This difference in time horizons reduces the value of an earnout transaction, given that the seller anticipates that the buyer has an incentive to behave opportunistically. In contrast, an asset for equity sale has a common valuation dimension for the buyer and seller, so that both parties gain from actions that increase the market value of the buying firm s common stock and are harmed by actions that reduce it. However, an asset for equity sale exposes the seller to risks entailed in the overall ex post performance of the buyer firm, and not just the assets involved in the transaction. These risks increase with the size of the buyer relative to the asset. Thus, the willingness of the seller to accept this risk in an asset for equity sale suggests more favorable announcement returns for the buyer compared to a cash-based asset sale. An asset for equity sale can also be viewed as an intercorporate transaction in which private information about the value of the buyer, and about the value of the asset to be sold, is conveyed through private negotiations. Given asymmetry of information between the participating firms and uninformed outside investors, the willingness of a corporate seller to hold an equity interest in the buyer, signals positive private information about the buyer to financial market participants. This informational setting facilitates the use of buyer equity as the means of payment and counters the Myers and Majluf (1984) adverse selection problem about the issuance of buyer equity. From this perspective, an asset for equity sale transaction should enhance shareholder wealth for buyers and sellers. Overall, our empirical tests generate evidence that allows us to discriminate between theoretical predictions as to whether the use of equity as a means of payment is value enhancing or conveys negative

9 7 private information. We conclude that the evidence from asset for equity sales is consistent with theoretical models that view the use of equity as an effective means of risk sharing between the participants in such a transaction rather than a signal of buyer overvaluation. B. Agency Considerations, Monitoring, and Control A theme in the corporate governance literature is the recognition that the separation of ownership from control that characterizes public firms gives rise to the potential for agency difficulties. In this section we draw on literature that is concerned with the role of agency problems in mergers and acquisitions. Embedded within this literature is a continuing debate between the view that blockholders enhance value for dispersed shareholders as a result of their monitoring activities, versus the view that blockholders reduce value by fostering managerial entrenchment or the consumption of perks, or by reducing managerial initiative and discretion. Since selling firms become corporate blockholders in buyers as a result of asset for equity sales, we turn to this literature as a source to establish testable hypotheses about the market valuation effects of these transactions. Researchers such as DeAngelo and Rice (1983), Jensen (1986), Shleifer and Vishny (1989), and Jensen and Murphy (1990), argue that managers are inefficiently disciplined by market forces. They emphasize the potential for managers to pursue policies that advance their own interests and foster managerial entrenchment rather than the maximization of shareholder wealth, and use this reasoning to explain the prevalence of negative excess returns for acquirers at acquisition announcements. Jensen (1986) argues that firms with free cash flow have considerable discretion over corporate resources that may not be used in shareholders interests. For example, managers may pursue acquisitions that do not enhance shareholder wealth in order to increase the size of their firm and enhance their compensation or prestige. Related agency hypotheses explain acquisitions on the basis of managerial hubris (Roll (1986)) or diversification of managerial human capital (Amihud and Lev (1981)).

10 8 From the perspective of the potential for managerial entrenchment, the use of buyer equity as the means of payment to acquire an operating asset could be interpreted as an anti-takeover device. A privately negotiated transaction that places a substantial block of the buying firm s stock in friendly hands could increase the probability of incumbent managers maintaining control with its associated benefits, since the creation of a friendly blockholding reduces the percentage of votes held by dispersed shareholders that potentially could be acquired by an outside bidder. Moreover, some of these transactions incorporate a standstill agreement that constrains the seller from acquiring additional equity and limits the seller s influence on the conduct of the buyer s business. Such a standstill agreement could serve as an indication of a friendly relationship between the two sets of managers that fosters managerial entrenchment at the buyer. In some cases the selling firm obtains representation on the board of directors of the buyer, which could also signal market participants that the seller is a passive ally of buyer firm managers rather than an active outside blockholder. As a result, an asset for equity sale could be interpreted by the market as a defensive tactic that fosters a misallocation of resources and exacerbates managerial entrenchment by insulating managers from the market for corporate control. From this entrenchment viewpoint, there should be an unfavorable announcement effect on buyer value at an asset for equity sale relative to a cash transaction. Shleifer and Vishny (1997) discuss how blockholders can promote their own interests at the expense of other shareholders. Holmstrom and Tirole (1993) argue that blockholders lessen firm value by reducing stock liquidity, because they inhibit information production in the stock market. Burkart, Gromb, and Panunzi (1997) argue that large outside shareholders may harm shareholder value because a blockholder reduces managerial discretion. They contend that managerial discretion encourages firmspecific investment and effort by managers that contribute to firm value, but that managers are less likely to show such initiative when there is a potential for expropriation by blockholders, thereby reducing firm value. From the perspective of these models, the creation of a corporate blockholding that results from an asset for equity sale should reduce buyer value.

11 9 An alternative perspective is that blockholdings, such as those formed by asset for equity sales, increase the exposure of the acquirer s managerial activities to the market for corporate control. More specifically, the creation of a block may reduce the fraction of the votes controlled, directly or indirectly, by management, weaken its control of the firm, and increase the probability that incumbent managers that perform poorly will subsequently lose control of their firm through a takeover or a proxy fight. Shleifer and Vishny (1986) contend that a blockholder has an incentive to engage in extensive monitoring of the acquirer s managerial performance to enhance firm value, and thus the value of the block. Such a corporate blockholder has a comparative advantage at monitoring managers relative to dispersed shareholders who are unlikely to engage in the costly information acquisition that underpins this task. As a result, a corporate blockholder can act to counter managerial actions that reduce shareholder wealth, prevent acquisitions that do not contribute to shareholder value, and insure that corporate resources are managed efficiently. If such monitoring activities are valuable, then the formation of a corporate blockholding through an asset for equity sale should enhance the value of the buyer, as well as increase overall wealth. Black (1992) offers an extension to the monitoring hypothesis by arguing that institutional shareholders effectively monitor managers in a manner comparable to non-institutional blockholders. From this perspective, the extent of existing sources of external monitoring could influence the valuation effects of asset for equity transactions. As a result, the greater the monitoring provided by existing large shareholders of a buyer firm, as measured by the extent of institutional ownership and non-institutional outside blockholdings prior to the transaction, the less the value of the incremental monitoring of the buyer that is contributed by the formation of the new corporate blockholding that is created by an asset for equity sale. Overall, the previous literature suggests an ambiguous view as to whether blockholders are valueenhancing for dispersed shareholders due to their monitoring of managers, or are value-decreasing because of their negative impact on stock liquidity and managerial initiative. Our empirical results are consistent with the view that a corporate blockholder created by an asset for equity sale is an effective and valuable monitor. Moreover, the value contributed by the introduction of such a corporate blockholder is not

12 10 diminished by the presence of other external monitors or blockholders already in the firm s ownership structure. Thus, our study of asset for equity sales and the valuation effects of these transactions provides a test of the contrasting predictions of alternative hypotheses about the use of equity as a means of payment, and about the impact of the creation of a corporate blockholding on firm value. II. Data and Descriptive Statistics We search several sources to obtain the sample of intercorporate sales of operating assets in which the seller is paid with buyer equity. We examine the lists of divestitures and the Top 100" transactions for each year from 1982 through 2000 that are published in the periodical Mergers and Acquisitions, and conduct an online search of asset sale agreements that are reported on the Dow Jones News Retrieval Service and Lexis-Nexis. To be retained in the sample of eligible transactions, we require that, one, both the buyer and seller of the operating asset are corporations; two, information is disclosed about the specific terms of the transaction; three, as a result of the sale of the asset, the seller obtains an equity interest in the buyer but there is no change in control of the buyer; and four, the operating asset is wholly-owned by the seller prior to the sale, and the sale transfers full ownership of the asset to the buyer. The latter requirement eliminates transactions in which the seller retains partial ownership of the asset or in which there is an earnout provision. When either the buyer or the seller does not have sufficient data on the daily returns file of the Center for Research in Security Prices (CRSP) to conduct an event study analysis, typically because the entity trades on a foreign (predominantly Canadian) exchange, we retain the party to the transaction that has sufficient CRSP returns. The final sample of asset for equity transactions consists of 69 selling firms, 89 buying firms, and 62 transactions in which the buying and selling firms are both in the CRSP database. We obtain ownership data from proxy statements that are reported on Edgar or found in the SEC File provided by the Q-Data Corporation. Other corporate information is obtained from the Wall Street Journal, Standard and Poor s Stock Reports, Stock Guide, and Directory of Corporations, as well as annual reports and 8K

13 11 reports filed with the SEC. The announcement date of the event is the initial report of the transaction in the Wall Street Journal. Descriptive statistics for the sample of buyers and sellers are reported in Table I. The data indicate that asset for equity sales are major corporate events. The average (median) value of the transactions, measured in constant (year 1997) dollars is $429.2 ($81.4) million for buying firms and $550.0 ($180.5) million for selling firms. Buyer common stock used in payment is valued at the buyer s share price prior to the announcement of the transaction. Buyer and seller mean (median) market capitalization in constant dollars is $2,132.2 ($206.5) million and $27,090.1 ($1,974.3) million, respectively. Although buyers are large firms, seller firms, on average, are considerably larger. The average (median) equity interest that the seller receives is 18.1% (12.5%) of the buyer s shares outstanding after the transaction, indicating that the seller typically becomes an important blockholder. More than half of the transactions are all-equity deals. The mean (median) proportion of the payment that is equity is 76.2% (100.0%) for the buyer sample and 77.1% (100.0%) for the seller sample. A broad range of industries is represented in the sample, with 62 different 4-digit SIC codes among the 89 buyers and 63 different 4-digit SIC codes among the 69 sellers. The data indicate that both buyers and sellers have relatively concentrated ownership structures. For insider ownership by top executives and members of the board of directors, mean (median) holdings are 22.6% (17.5%) for buyers and 16.7% (7.2) for sellers. By comparison, Morck, Shleifer and Vishny (1998) report that median insider holdings for Fortune 500 companies are 3.4%. These ownership data run contrary to a prediction that can be drawn from Amihud, Lev, and Travlos s (1990) view that managers with substantial ownership stakes in their firms are averse to using equity to finance mergers and acquisitions because of the potential risk of losing control. With regard to institutional ownership, defined as shares held by investment companies, banks, insurance companies, college endowments, and 13F money managers, mean (median) holdings are 34.4% (30.5%) for buyers and 46.2% (51.1%) for sellers, figures comparable to the mean of 37.6% for institutional holdings at NYSE/ASE firms reported by McConnell and Servaes (1990). Almost 60% of the buyers and approximately one-third of the sellers have pre-existing

14 12 outside blockholders that hold 5% or more of the firm s outstanding shares. Overall, considerable ownership of stock by insiders, institutions, and external blockholders characterizes both buyer and seller firms. Such data might suggest that the incremental gain in buyer value from the additional monitoring provided by a new blockholder created in an asset for equity sale would be small, but our evidence reported in Section 4 indicates the contrary finding. Our empirical tests in the next section also allow us to separate the effects of the use of equity as a means of payment from the effects of forming a corporate blockholding. Standstill agreements are not common since only 13% of the agreements contain this provision. The rarity of standstill agreements suggests that buyers are not likely to undertake these transactions as a defensive tactic against corporate control activity. Representation of the seller on the buyer s board of directors occurs in approximately one-third of the cases. Overall, the descriptive statistics indicate that asset for equity asset sales are intercorporate transactions that are comparable in many respects to other corporate control events. Thus, our findings about the implications of these transactions contribute to the corporate finance literature by providing evidence about the value of the use of equity as a means of payment and the formation of a corporate blockholding that are intrinsic to these intercorporate transactions. III. Empirical Results A. Valuation Effects on Buyers and Sellers In Table II, using standard event study methodology, we report the excess returns to transactions in which a firm sells an operating asset and receives payment in the form of buyer equity. The two-day average excess return for sellers is 3.17%, statistically significant at the one percent level given a t-statistic of 5.90; 67% of the returns are positive. This average excess return is considerably greater than comparable figures for sellers in asset sales reported by Klein (1986) (1.1%), Hite, Owers, and Rogers (1987) (1.7%), Lang, Poulsen, and Stulz (1995) (1.4%), and John and Ofek (1995) (1.5%). The magnitude of the excess returns to sellers paid with buyer equity suggests that these asset sales are positive net present value

15 13 transactions for selling firms, and that the increase in seller shareholder value is considerably greater than that for cash transactions based on results documented in the asset sales literature. Since the means of payment is predominantly in the form of buyer equity, it is unlikely that sellers would undertake these transactions as a source of financing for projects or other corporate purposes. Although the block of equity that a seller obtains can be liquidated through a subsequent financial market transaction, it is unlikely that a seller would accept buyer equity with the intention of quickly selling the shares through a block trade or a secondary offering, given the size of the equity blocks involved. Moreover, the cumulative average excess return for sellers for the one year period prior to these announcements is 0.36% (t-statistic = 0.05), indicating that these transactions are undertaken by sellers after periods of normal stock price performance. Instead, the gains for sellers are consistent with the view that by shifting control of an operating asset to a buyer that will enhance its value, the buyer can offer the seller a premium for the asset sufficiently large to induce the seller to agree to the sale, enhancing seller value to an extent that exceeds that obtained in asset sales for cash. The average excess return to buyers is a positive 9.77%, with a t-statistic of 16.22, statistically significant at the one percent confidence level; 80% of the returns are positive. This result implies that buyer returns at these acquisitions are considerably greater than the small excess returns to buyers reported in previous studies of asset sales for cash. For example, Hite, Owers, and Rogers (1987) in a study of intercorporate asset sales, report a two-day average excess return to buyers of 0.83%. Moreover, the large gains to buyers that we observe are contrary to the statistically significant negative returns of -1% to -3% to buyers reported in studies of mergers and acquisitions of public companies in which buyer equity is the means of payment (Travlos (1987), Franks, Harris, and Mayer (1988), Asquith, Bruner, and Mullins (1987), and Servaes (1991)). Positive returns to buyers indicate that the gains sellers obtain in asset for equity transactions do not represent a redistribution of wealth from buyers due to overbidding, and instead suggest that the gains result from effective risk sharing. Our results for asset sales are not consistent with agency explanations for overbidding that have been used to explain large gains for targets but negative

16 14 returns to buyers in mergers and acquisitions of public companies (Roll (1986) and Morck, Shleifer, and Vishny (1990)). Instead, the sale of an operating asset in exchange for buyer equity generates, on average, large increases in shareholder wealth for both the buyer and seller, suggesting that these transactions contribute to economic value because they reallocate resources toward higher valued uses and generate greater value for the buyer because of market expectations of heightened future profitability and effective risk sharing. Based on the set of 62 events in which the buyer and the seller are both listed on CRSP, we calculate the market s assessment of the average combined dollar gains in value from announcements of these transactions relative to the market values of the participants and also relative to the size of the transaction. For this sample of events, the two-day average excess returns are 2.87% (t-statistic = 4.91) for the sellers and 11.38% (t-statistic = 14.38) for the buyers (not reported in the table). The weighted average return for these transactions, that is, weighted by the respective parties market capitalization prior to the transaction announcement, is 3.18% (t-statistic = 2.72); in 74% of the transactions the combined return is positive. The implied mean gain in combined shareholder value as a percent of the transaction price for the assets that are sold is 223.4%. Thus, on average, an asset for equity sale generates a large increase in the combined shareholder value of buyers and sellers relative to the transaction price negotiated by the parties. The magnitude of this gain suggests that much of the change in shareholder wealth is due to the use of equity as a means of payment. Out of the increase in total shareholder wealth generated by these transactions, sellers obtain 50.8% of the total gains and buyers 49.2%. Since the price negotiated in each transaction incorporates a specified number of buyer shares to be allocated to the seller, the change in the buyer s share price induced by the announcement of the sale generates a proportionate change in the market value of the buyer common stock that is conveyed to the seller. Thus, a portion of the excess return to the seller reflects the change in the buyer s share price induced by the transaction. Based on announcement excess returns, the number of shares conveyed, and relevant values for each transaction, the implied mean

17 15 percentage change in seller value that is due to this induced change in the value of the seller s holding of buyer shares is 1.10%. Since the size of the seller s blockholding is likely to have an influence on the valuation effects generated by the formation of the new blockholding, we disaggregate the excess returns into ranges of seller ownership. The results, reported in Table II, indicate that there are statistically significant and uniformly positive excess returns for buyers throughout the ownership ranges. Moreover, transactions that result in seller equity holdings that are less than 5% of the buyer s outstanding shares generate large, significant returns for buyers of 8.81% (t-statistic = 7.18). This result suggests that even those asset for equity transactions that do not result in the formation of a 5% blockholding sharply increase buyer value, implying that it is the use of buyer equity and not just the formation of a large corporate blockholding that creates buyer value. For levels of seller shareholdings that exceed 5% of buyer outstanding shares, there is a greater impact on buyer excess returns. In the range of 5% to 10% of buyer outstanding shares, the excess return to buyers is 9.02% (t-statistic = 7.46), and the largest return occurs in the 10% to 20% ownership range, 12.56% (t-statistic = 8.71). These results suggest that buyer shareholders gain additional value from an increase in the size of the blockholding that results from an asset for equity transaction over a considerable range of ownership. Above the 20% ownership range, returns to buyers are still strongly positive but somewhat lower than for the 10% to 20% range. When we calculate difference in means tests between the average excess returns in the middle ownership group (10% to 20%) versus other ownership groups, none of the calculated t-values are statistically significant. Nevertheless, since the pattern of returns is suggestive of a nonlinear relationship between the size of the blockholding and excess returns to buyers, in the subsequent section we test for such a relationship using cross-sectional regressions. Excess returns to sellers are also uniformly positive across the various ranges of blockholdings, but are only statistically significant in two ranges, below 5%, with an excess return of 3.72% (t-statistic = 3.95) and in the range of 20% to 30% ownership, with an excess return of 7.21% (t-statistic = 5.43). However,

18 16 none of the differences between seller returns based on ranges of ownership are statistically significant. Thus, the results suggest that the valuation effect on the seller is not related to the size of the blockholding it obtains. We next disaggregate the event study results to gain insight about the influence of several characteristics of the transactions on excess returns. The results are reported in Table III. Three factors relate to the terms of the transaction. One factor is the proportion of the transaction price that is in the form of buyer equity. In Eckbo, Giammarino, and Heinkel s (1990) asymmetric information model of mergers and acquisitions, the proportion of equity affects bidder returns because it conveys private information about the value of the buyer and the target, including the synergies expected between the two entities. However, their empirical results, drawn from data on Canadian mergers, do not support their hypothesis. When we disaggregate between all equity transactions and those that involve a mix of equity and cash, both types of transactions have statistically significant positive results, but the all equity transactions generate modestly greater average excess returns to buyers and sellers. In all-equity deals, the returns are 10.27% (t-statistic = 10.96) for buyers and 4.27% (t-statistic = 5.00) for sellers. For transactions that involve some cash, the returns are 9.20% (t-statistic = 12.04) for buyers and 1.81% (t-statistic = 3.26) for sellers. Nevertheless, the difference in means calculated t-values for these sets of returns are not statistically significant. A second factor is whether the transaction entails a standstill agreement, a provision that precludes the seller from acquiring additional equity in the buyer. In a study of responses to takeover bid activity, Dann and DeAngelo (1988) find that announcements of standstill agreements (even if not accompanied by negotiated stock repurchases) generate negative returns to target firms. If the seller s holding of buyer equity that results from an asset for equity sale is viewed as a toehold that creates the prospect of a subsequent bid for control of the buyer by the seller or by a third party, the presence of a standstill agreement in an asset for equity sale should reduce returns to buyers. Buyer returns might also be lessened if a standstill agreement is viewed as an indicator of managerial entrenchment that serves to isolate incumbent managers from the threat of removal and strengthens their opportunity to derive private benefits

19 17 at the expense of dispersed shareholders. We find that transactions with standstill agreements have average excess returns of 8.19% (t-statistic = 8.83) for buyers and -0.01% (t-statistic = -0.14) for sellers, while for transactions without a standstill agreement, the returns are somewhat greater, 10.01% (t-statistic = 13.96) for buyers and 3.64% (t-statistic = 6.38) for sellers. Nevertheless, difference in means calculated t-values are low and not statistically significant, so we cannot conclude that standstill agreements lessen the value created by these transactions. A third factor is that some asset for equity transactions provide for representation of the seller on the board of directors of the buyer, a provision that could enhance the prospect that the seller will influence the conduct of the buyer s business. For events in which sellers obtain board representation, the average excess returns are 8.51% (t-statistic = 10.94) for buyers and 1.21% (t-statistic = 1.92) for sellers. For events in which sellers do not receive a board seat, the excess returns are 10.54% (t-statistic = 12.04) for buyers and 4.27% (t-statistic = 5.94) for sellers. Although board representation for sellers is associated with lower returns for both buyers and sellers, the differences generate low calculated t-statistics and are not statistically significant. Thus, we cannot conclude that seller representation on the board affects the gains from asset for equity sales. Overall, the event study results suggest that asset for equity transactions generate strongly positive excess returns for both buyers and sellers. The evidence also indicates that various characteristics of these transactions that have been examined in the acquisitions literature do not have statistically significant impacts on the valuation effects on parties that engage in asset for equity transactions, suggesting that it is the use of equity as the means of payment and the associated formation of a corporate blockholding that are the primary causal factors for the large gains in value that we observe. B. Cross-sectional Regression Analysis In the cross-sectional regressions, Tables IV and V, the dependent variables are the individual excess returns to buyers and sellers, respectively, and the independent variables reflect characteristics of the

20 18 transactions and the participating firms. Regression analysis is used to determine the salient relationships among the variables, taking into account several continuous variables, such as the size of the transaction, that are not readily susceptible to event study analysis. For the buyer regressions, the coefficients suggest several major results. First, we find that there is a positive and statistically significant relationship between the buyer s share price reaction and the size of the transaction, measured in terms of the transaction price and the buyer s market value. Assuming that selling firms are better informed than outsiders about the value of the assets to be sold, sellers will prefer to accept buyer equity when they believe that the assets to be sold are undervalued so that the seller can participate in the post-consolidation gains that will accrue as the productivity of the assets as part of the buyer s operations is revealed to the market. Thus, these results are consistent with Hansen s (1987) risk sharing hypothesis. Second, there is a positive, non-linear relationship between the buyer s excess returns and the size of the block of common stock that the seller receives relative to the buyer s total shares outstanding. We estimate several non-linear specifications for this ownership variable, and focus on the results for a cubic specification. If the corporate blockholding and the monitoring initiated by the asset for equity transaction align managerial and shareholder interests at the buyer firm, the ceteris paribus effect of an increase in the size of the seller-held block on buyer returns should be positive for a broad range of ownership. The estimated coefficients are statistically significant and imply that for low levels of the seller s blockholding, an increase in the size of the blockholding has a positive effect on the share price reaction of the buyer. This effect reaches a local maximum when approximately a 10 percent ownership interest in the buyer is conveyed to the seller. The incremental effect of an increase in the blockholding on the buyer s share price then declines, reaching a minium at approximately 35 percent. Beyond this level of ownership, the marginal effect of an increase in the size of the seller s blockholding on buyer value is again positive. This regression evidence provides a more detailed analysis of the effects of the size of the blockholding, but the pattern of these ownership effects is broadly consistent with the event study evidence.

21 19 Third, we estimate regressions utilizing various specifications that incorporate variables for the buyer s ownership structure. Lewellen, Loderer, and Rosenfeld (1985) suggest that the buyer s share price reaction to an acquisition of a public firm is a positive function of buyer insider ownership, since concerns over agency problems in acquisitions are mitigated when there is high insider ownership. Consistent with this view, Amihud, Lev, and Travlos (1990) report that the typically negative returns to buyers associated with the use of common stock as a medium of exchange by buyers apply primarily to acquirers with low managerial ownership. Although buyers in asset for equity sales have relatively high insider ownership, we find no evidence that insider ownership affects the buyer gains generated by these transactions. Likewise, there is no evidence that holdings of either institutions or external blockholders affect returns to buyers at asset for equity sales that create a new blockholder. Four, the regression results consistently indicate that none of the variables that reflect various provisions sometimes included in asset for equity sales agreements affect returns to buyers. Qualitative variables for the presence of a standstill agreement, and board representation, typically have negative signs, but the t-statistics are consistently low. When we specify a variable for the proportion of the transaction price that is in the form of buyer equity, the coefficients are consistently small and not significant, a finding that is not consistent with Eckbo, Giammarino, and Heinkel s (1988) theoretical model. Corporate law treats an asset sale as a fiduciary responsibility of the board of directors that need not be submitted for a shareholder vote, unless a firm is selling substantially all of its assets. Thus, unlike a merger, shareholders who disagree with these asset sale transactions have no recourse other than to sell their shares. Nevertheless, for some transactions (10 sellers and 32 buyers) in our sample, the asset for equity sale is made contingent on a shareholder vote. Such a vote could delay closing and add to transactions costs because of applicable legal mandates and compliance rules, but could still be value-enhancing if it strengthens the market s expectation that the transaction is in the interests of shareholders. When qualitative variables that reflect the provision for shareholder voting are specified, the coefficients are small and not

22 20 significant. These results suggest that shareholder voting is a pro forma exercise rather than a signal of buyer value for an asset for equity sale. Additionally, we test whether strategic relationships and product market dealings between the parties affect the returns to asset for equity transactions, given that Allen and Phillips (2000) report that alliances and joint ventures enhance the change in target firm value for private placements of equity. However, when we include a qualitative variable that indicates the presence of such relationships, the coefficient is small and not significant. Thus, there is no evidence that such strategic relationships alter the valuation effects of asset for equity sales. The coefficient of a qualitative variable for the six transactions that are announced but not completed, so that no resulting block is formed, obtains a negative sign, suggesting that the market may partially anticipate that these transactions will not come to fruition. However, the t-statistic for this variable is not statistically significant. Overall, the regression results suggest that returns to buyers are closely related to the magnitude of the transaction and the size of the blockholding that the seller takes in the buyer, but are not affected by other provisions included in the agreement or by firm characteristics. With respect to the regressions for seller excess returns, there is evidence of a positive effect of the size of the transaction relative to the seller s market capitalization. Thus, the size of the transaction has a positive effect on the returns to sellers as well buyers. However, for seller firms none of the coefficients are statistically significant for any of the remaining variables that are specified to reflect the terms of the transaction or the characteristics of the seller, including ownership structure. The one exception is for a qualitative variable for a provision requiring a seller shareholder vote, which obtains a positive coefficient that is statistically significant, suggesting that an asset for equity sale that requires an approval vote by seller shareholders has a positive effect on seller value.