A Financial Perspective on Mergers and Acquisitions

The Free Cash Flow Theory of Takeovers: A Financial Perspective on Mergers and Acquisitions and the Economy Michael C. Jensen Harvard Business School [email  protected] edu  © Michael C. Jensen, 1987 â€Å"The Merger Boom†, Proceedings of a Conference sponsored by Federal Reserve Bank of Boston, Oct. 1987, pp. 102-143 This document is available on the Social Science Research Network (SSRN) Electronic Library at: http://papers. ssrn. com/ABSTRACT=350422 The Free Cash Flow Theory of Takeovers: A Financial Perspective on Mergers and Acquisitions and the Economy Michael C.Jensen* Harvard Business School [email  protected] edu From, â€Å"The Merger Boom†, Proceedings of a Conference sponsored by Federal Reserve Bank of Boston, Oct. 1987, pp. 102-143 Economic analysis and evidence indicate the market for corporate control is benefiting shareholders, society, and the corporate form of organization. The value of transactions in this market ran at a record rate of about $ 180 billion per year in 1985 and 1986—47 percent above the 1981 record of $122 billion.The number of transactions with purchase prices exceeding one billion dollars was 27 of 3300 deals in 1986 and 36 of 3000 deals in 1985 (Grimm, 1985). There were only seven billion-dollar plus deals in total, prior to 1980. In addition to these takeovers, mergers, and leveraged buyouts, there were numerous corporate restructurings involving divestitures, spinoffs, and large stock repurchases for cash and debt. The gains to shareholders from these transactions have been huge.The gains to selling-firm shareholders from mergers and acquisition activity in the period 1977-86 total $346 billion (in 1986 dollars). 1 The gains to buying-firm shareholders are harder Estimated from data in Grimm (1986). Grimm provides total dollar values for all merger and acquisition deals for which there are publicly announced prices amounting to at least $500,000 or 10 percent of the firm and in which at least on e of the firms was a U. S. company. Grimm also counts in its numerical totals deals with no publicly announced prices that it believes satisfy these criteria.I have assumed that the deals with no announced prices were on average equal to 20 percent of the size of the announced transactions and carried the same average premium. *Professor of Business Administration, Harvard Business School, and Professor of Finance and Business Administration, University of Rochester. The author is grateful for the research assistance of Michael Stevenson and the helpful comments by Sidney Davidson, Harry DeAngelo, Jay Light, Robert Kaplan, Nancy Macmillan, Kevin Murphy, Susan Rose-Ackerman, Richard Ruback, Wolf Weinhold, Toni Wolcott, and especially Armen Alchian.This research is supported in part by the Division of Research, Harvard Business School, and the Managerial Economics Research Center, University of Rochester. The analysis here draws heavily on that in Jensen (forthcoming 1988). 1 M. C. Je nsen 2 1987 to estimate, and to my knowledge no one has done so yet, but I estimate that they would add at least another $50 billion to the total. These gains, to put them in perspective, equal 31 percent of the total cash dividends (valued in 1986 dollars) paid to investors by the entire corporate sector in the past decade. Corporate control transactions and the restructurings that often accompany them can be wrenching events in the lives of those linked to the involved organizations: the managers, employees, suppliers, customers and residents of surrounding communities. Restructurings usually involve major organizational change (such as shifts in corporate strategy) to meet new competition or market conditions, increased use of debt, and a flurry of recontracting with managers, employees, suppliers and customers.This activity sometimes results in expansion of resources devoted to certain areas and at other times in contractions involving plant closings, layoffs of top-level and mi ddle managers and of staff and production workers, and reduced compensation. Change due to corporate restructuring requires people and communities associated with the organization to adjust the ways they live, work and do business. It is not surprising, therefore, that this change creates controversy and that those who stand to lose are demanding that something be done to stop the process.At the same time, shareholders in restructured corporations are clear-cut winners; in recent years restructurings have generated average increases in total market value of approximately 50 percent. Those threatened by the changes argue that corporate restructuring is damaging the U. S. economy, that this activity damages the morale and productivity of organizations and pressures executives to manage for the short term. Further, they hold that the value that restructuring creates does not come from increased efficiency and productivity; rather, the gains come from lower tax payments, broken contract s withTotal dividend payments by the corporate sector, unadjusted for inflation, are given in Weston and Copeland (1986, p. 649). I extended these estimates to 1986. 2 M. C. Jensen 3 1987 managers, employees and others, and mistakes in valuation by inefficient capital markets. Since the benefits are illusory and the costs are real, they argue, takeover activity should be restricted. The controversy has been accompanied by strong pressure on regulators and legislatures to enact restrictions to curb activity in the market for corporate control.Dozens of congressional bills in the past several years have proposed new restrictions on takeovers, but as of August 1987, none had passed. The Business Roundtable, composed of the chief executive officers of the 200 largest corporations in the country, has pushed hard for restrictive legislation. Within the past several years the legislatures of New York, New Jersey, Maryland, Pennsylvania, Connecticut, Illinois, Kentucky, Michigan, Ohio, Indi ana, Minnesota and Massachusetts have passed antitakeover laws.The Federal Reserve Board implemented new restrictions in early 1986 on the use of debt in certain takeovers. In all the controversy over takeover activity, it is often forgotten that only 40 (an all-time record) of the 3,300 takeover transactions in 1986 were hostile tender offers. There were 110 voluntary or negotiated tender offers (unopposed by management) and the remaining 3,100-plus deals were also voluntary transactions agreed to by management. This simple classification, however, is misleading since many of the voluntary transactions would not have occurred absent the threat of hostile takeover.A major reason for the current outcry is that in recent years mere size alone has disappeared as an effective takeover deterrent, and the managers of many of our largest and least efficient corporations now find their jobs threatened by disciplinary forces in the capital markets. Through dozens of studies, economists have accumulated considerable evidence and knowledge on the effects of the takeover market. Most of the earlier work is well summarized elsewhere (Jensen and Ruback (1983); Jensen (1984); Jarrell, Brickley and M. C.Jensen 4 1987 Netter (1988)). Here, I focus on current aspects of the controversy. In brief, the previous work tells us the following: †¢ Takeovers benefit shareholders of target companies. Premiums in hostile offers historically exceed 30 percent on average, and in recent times have averaged about 50 percent. †¢ Acquiring-firm shareholders on average earn about 4 percent in hostile takeovers and roughly zero in mergers, although these returns seem to have declined from past levels. †¢ Takeovers do not waste credit or resources.Instead, they generate substantial gains: historically, 8 percent of the total value of both companies. †¢ Actions by managers that eliminate or prevent offers or mergers are most suspect as harmful to shareholders. †¢ Golden pa rachutes for top-level managers do not, on average, harm shareholders. †¢ The activities of takeover specialists (such as Icahn, Posner, Steinberg, and Pickens) benefit shareholders on average. †¢ Merger and acquisition activity has not increased industrial concentration.Over 1200 divestitures valued at $59. 9 billion occurred in 1986, also a record level (Grimm, 1986). †¢ Takeover gains do not come from the creation of monopoly power. Although measurement problems make it difficult to estimate the returns to bidders as precisely as the returns to targets,3 it appears the bargaining power of target managers, coupled with competition among potential acquirers, grants a large share of the acquisition benefits to selling shareholders. In addition, federal and state regulation of 3See Jensen and Ruback (1983, pp. 18ff). M. C. Jensen 5 1987 tender offers appears to have strengthened the hand of target firms; premiums received by target-firm shareholders increased substanti ally after introduction of such regulation. 4 Some have argued that the gains to shareholders come from wealth reallocations from other parties and not from real increases in efficiency. Roll (1986) argues the gains to target firm shareholders come from acquiring firm shareholders, but the data are not consistent with this hypothesis.While the evidence on the returns to bidding firms is mixed, it does not indicate they systematically suffer losses; prior to 1980 shareholders of bidding firms earned on average about zero in mergers, which tend to be voluntary, and about 4 percent of their equity value in tender offers, which more often are hostile Jensen and Ruback (1983). These differences in returns are associated with the form of payment rather than the form of the offer: tender offers tend to be for cash and mergers tend to be for stock (Huang and Walkling, 1987).Some argue that bondholders in acquired firms systematically suffer losses as substantial amounts of debt are added to the capital structure. Asquith and Kim (1982) do not find this, nor do Dennis and McConnell (1986). The Dennis and McConnell study of 90 matched acquiring and acquired firms in mergers in the period 1962-80 shows that the values of bonds, preferred stock and other senior securities, as well as the common stock prices of both firms, increase around the merger announcement. Changes in the value of senior securities are not captured in measures of changes in the value of common stock prices summarized previously.Taking the changes in the value of senior securities into account, Dennis and McConnell find the average change in total dollar value is positive for both bidders and target firms. Shleiffer and Summers (1987) argue that some of the benefits earned by target and bidding firm shareholders come from the abrogation of explicit and implicit longterm contracts with employees. They point to highly visible recent examples in the airline See Jarrell and Bradley (1980), Nathan and Oâ⠂¬â„¢Keefe (1986), however, provide evidence that this effect occurred in 1974, several years after the major legislation. M. C. Jensen 6 1987 industry, where mergers have been frequent and wages have been cut in the wake of deregulation. But given deregulation and free entry by low-cost competitors, the cuts in airline industry wages were inevitable and would have been accomplished in bankruptcy proceedings if not in negotiations and takeover-related crises. Medoff and Brown (1988) study this issue using data from Michigan. They find that both employment and wages are higher, not lower, after acquisition than would otherwise be expected; however, their sample consists largely of combinations of small firms.The Market for Corporate Control The market for corporate control is best viewed as a major component of the managerial labor market. It is the arena in which alternative management teams compete for the rights to manage corporate resources (Jensen and Ruback, 1983). Understandin g this point is crucial to understanding much of the rhetoric about the effects of hostile takeovers. Takeovers generally occur because changing technology or market conditions require a major restructuring of corporate assets (although in some cases, takeovers occur because incumbent managers are incompetent).Such changes can require abandonment of major projects, relocation of facilities, changes in managerial assignments, and closure or sale of facilities or divisions. Managers often have trouble abandoning strategies they have spent years devising and implementing, even when those strategies no longer contribute to the organization’s survival, and it is easier for new top-level managers with no ties to current employees or communities to make changes. Moreover, normal organizational resistance to change commonly is lower early in the reign of new top-level managers.When the internal processes for change in large corporations are too slow, costly, and clumsy to bring about the required restructuring or change in managers efficiently, the capital markets do so through the M. C. Jensen 7 1987 market for corporate control. Thus, the capital markets have been responsible for substantial changes in corporate strategy. Causes of Current Takeover Activity A variety of political and economic conditions in the 1980s have created a climate where economic efficiency requires a major restructuring of corporate assets.These factors include: †¢ †¢ The relaxation of restrictions on mergers imposed by the antitrust laws. The withdrawal of resources from industries that are growing more slowly or that must shrink. †¢ Deregulation in the markets for financial services, oil and gas, transportation, and broadcasting, bringing about a major restructuring of those industries. †¢ Improvements in takeover technology, including more and increasingly sophisticated legal and financial advisers, and innovations in financing technology (for example, the strip financing commonly used in leveraged buyouts and the original issuance of high-yield non-investment-grade bonds).Each of these factors has contributed to the increase in total takeover and reorganization activity. Moreover, the first three factors (antitrust relaxation, exit, and deregulation) are generally consistent with data showing the intensity of takeover activity by industry. Table 1 indicates that acquisition activity in the period 1981-84 was highest in the oil and gas industry, followed by banking and finance, insurance, food processing, and mining and minerals. For comparison purposes, the table also presents data on industry value measured as a percentage of the total value of all firms.All but two of the industries, retail trade and transportation, represent a larger fraction of total takeover activity than their representation in the economy as a whole, indicating that the takeover market is concentrated in particular industries, not spread evenly throughout the corpo rate sector. M. C. Jensen 8 1987 Table 1 Intensity of Takeover Activity, by Industry, 1981-84 Percent Percent of Total of Total Takeover Corporate Industry Classification of Seller Market Valueb Activitya Oil and Gas 26. 13. 5 Banking and Finance 8. 8 6. 4 Insurance 5. 9 2. 9 Food Processing 4. 6 4. 4 Mining and Minerals Conglomerate Retail Trade Transportation Leisure and Entertainment Broadcasting Other a 4. 4 4. 4 3. 6 2. 4 2. 3 2. 3 39. 4 1. 5 3. 2 5. 2 2. 7 . 9 . 7 58. 5 Value of merger and acquisition transactions in the industry as a percentage of total takeover transactions for which valuation data are publicly reported. Source: W. T Grimm, Mergerstat Review (1984, p. 41). bIndustry value as a percentage of the value of all firms, as of 12/31/84 Total value is measured as the sum of the market value of common equity for 4,305 companies, including 1,501 companies on the New York Stock Exchange, 724 companies on the American Stock Exchange, plus 2,080 companies in the over-the -counter market. Source: The Media General Financial Weekly, (December 31, 1984, p 17) Many sectors of the U. S. economy have been experiencing slower growth and, in some cases, even retrenchment. This phenomenon has many causes, including substantially increased foreign competition.The slow growth has meant increased takeover activity because takeovers play an important role in facilitating exit from an industry or activity. Changes in energy markets, for example, have required radical restructuring and retrenchment in that industry, and takeovers have played an important role in accomplishing these changes; oil and gas rank first in takeover activity, with twice their proportionate share of total activity. Managers who are slow to adjust to the new energy environment and slow to recognize that many old practices and strategies are no longer viable find that takeovers M. C.Jensen 9 1987 are doing the job for them. In an industry saddled with overcapacity, exit is cheaper to accompl ish through merger and the orderly liquidation of marginal assets of the combined firms than by disorderly, expensive bankruptcy. The end of the competitive struggle in such an industry often comes in the bankruptcy courts, with the unnecessary destruction of valuable parts of organizations that could be used productively by others. Similarly, deregulation of the financial services market is consistent with the number 2 rank of banking and finance and the number 3 rank of insurance in table 1.Deregulation has also been important in the transportation and broadcasting industries. Mining and minerals has been subject to many of the same forces impinging on the energy industry including the changes in the value of the dollar. The development of innovative financing vehicles, such as high yield noninvestment-grade bonds (junk bonds), has removed size as a significant impediment to competition in the market for corporate control. Investment grade and high-yield debt issues combined were associated with 9. percent of all tender offer financing from January 1981 through September 1986 (Drexel Burnham Lambert, undated). Even though not yet widely used in takeovers, these new financing techniques have had important effects because they permit small firms to obtain resources for acquisition of much larger firms by issuing claims on the value of the venture (that is, the target firm’s assets) just as in any other corporate investment activity. Divestitures If assets are to move to their most highly valued use, acquirers must be able to sell off assets to those who can use them more productively.Therefore, divestitures are a critical element in the functioning of the corporate control market and it is important to avoid inhibiting them. Indeed, over 1200 divestitures occurred in 1986, a record level (Mergerstat Review (1986)). This is one reason merger and acquisition activity has not increased industrial concentration. M. C. Jensen 10 1987 Divested plants and asse ts do not disappear; they are reallocated. Sometimes they continue to be used in similar ways in the same industry, and in other cases they are used in very different ways and in different industries.But in both cases they are moving to uses that their new owners believe are more productive. Finally, the takeover and divestiture market provides a private market constraint against bigness for its own sake. The potential gains available to those who correctly perceive that a firm can be purchased for less than the value realizable from the sale of its components provide incentives for entrepreneurs to search out these opportunities and to capitalize on them by reorganizing such firms into smaller entities.The mere possibility of such takeovers also motivates managers to avoid putting together uneconomic conglomerates and to break up existing ones. This is now happening. Recently many firms’ defenses against takeovers appear to have led to actions similar to those proposed by th e potential acquirers. Examples are the reorganizations occurring in the oil and forest products industries, the sale of â€Å"crown jewels,† and divestitures brought on by the desire to liquidate large debts incurred to buy back stock or make other payments to stockholders.The basic economic sense of these transactions is often lost in a blur of emotional rhetoric and controversy. Managerial Myopia versus Market Myopia It has been argued that, far from pushing managers to undertake needed structural changes, growing institutional equity holdings and the fear of takeover cause managers to behave myopically and therefore to sacrifice long-term benefits to increase short-term profits.The arguments tend to confuse two separate issues: 1) whether managers are shortsighted and make decisions that undervalue future cash flows while overvaluing current cash flows (myopic managers); and 2) whether security markets are shortsighted and undervalue future cash flows while overvaluing ne ar-term cash flows (myopic markets). M. C. Jensen 11 1987 There is little formal evidence on the myopic managers issue, but I believe this phenomenon does occur.Sometimes it occurs when managers hold little stock in their companies and are compensated in ways that motivate them to take actions to increase accounting earnings rather than the value of the firm. It also occurs when managers make mistakes because they do not understand the forces that determine stock values. There is much evidence inconsistent with the myopic markets view and no evidence that indicates it is true: (1) The mere fact that price-earnings ratios differ widely among securities indicates the market is valuing something other than current earnings. For example, it values growth as well.Indeed, the essence of a growth stock is that it has large investment projects yielding few short term cash flows but high future earnings and cash flows. The continuing marketability of new issues for start-up companies with li ttle record of current earnings, the Genentechs of the world, is also inconsistent with the notion that the market does not value future earnings. (2) McConnell and Muscarella (1985) provide evidence that (except in the oil industry) stock prices respond positively to announcements of increased investment expenditures and negatively to reduced expenditures.Their evidence is also, inconsistent with the notion that the equity market is myopic, since it indicates that the market values spending current resources on projects that promise returns in the future. (3) The vast evidence on efficient markets, indicating that current stock prices appropriately incorporate all currently available public information, is also inconsistent with the myopic markets hypothesis. Although the evidence is not literally 100 percent in support of the efficient market hypothesis, no proposition in any of the social sciences is better documented. 5For an introduction to the literature and empirical evidence on the theory of efficient markets, see Elton and Gruber (1984, Chapter 15, p. 375ff), and the 167 studies referenced in the bibliography. For some anomalous evidence on market efficiency, see Jensen (1978). For recent criticisms of the efficient market hypothesis see Shiller (1981a; 1981b), Marsh and Merton (1983; 1986) demonstrate that the Shiller 5 M. C. Jensen 12 1987 (4) Recent versions of the myopic markets hypothesis emphasize increases in the amount of institutional holdings and the pressure funds managers face to generate high quarterly returns.It is argued that these pressures on institutions are a major cause of pressures on corporations to generate high current quarterly earnings. The institutional pressures are said to lead to increased takeovers of firms, because institutions are not loyal shareholders, and to decreased research and development (R&D) expenditures. It is hypothesized that because R&D expenditures reduce current earnings, firms making them are more like ly to be taken over, and that reductions in R&D are leading to a fundamental weakening of the corporate sector of the economy.A study of 324 firms by the Office of the Chief Economist of the SEC (1985a) finds substantial evidence that is inconsistent with this version of the myopic markets argument. The evidence indicates the following: †¢ Increased institutional stock holdings are not associated with increased takeovers of firms. †¢ Increased institutional holdings are not associated with decreases in R&D expenditures. †¢ †¢ Firms with high R&D expenditures are not more vulnerable to takeovers. Stock prices respond positively to announcements of increases in R&D expenditures.Moreover, total spending on R&D is increasing concurrent with the wave of merger and acquisition activity. Total spending on R&D in 1984, a year of record acquisition activity, increased by 14 percent according to Business Week’s annual survey. This represented â€Å"the biggest gain since R&D spending began a steady climb in tests depend critically on whether, contrary to generally accepted financial theory and evidence, the future levels of dividends follow a stationary stochastic process. Merton (1985) provides a discussion of the current state of the efficient market hypothesis and concludes (p. 0), â€Å"In light of the empirical evidence on the nonstationarity issue, a pronouncement at this moment that the rational market theory should be discarded from the economic paradigm can, at best, be described as ‘premature’. † M. C. Jensen 13 1987 the late 1970’s. † All industries in the survey increased R&D spending with the exception of steel. In addition, R&D spending increased from 2 percent of sales, where it had been for five years, to 2. 9 percent. In 1985 and 1986, two more record years for acquisition activity, R&D also set new records.R&D spending increased by 10 percent (to 3. 1 percent of sales) in 1985, and in 1986, R &D spending again increased by 10 percent to $51 billion (3. 5 percent of sales), in a year when total sales decreased by 1 percent. 6 Bronwyn Hall (1987), in a detailed study of all U. S. manufacturing firms in the years 1976-85, finds in approximately 600 acquisitions that firms that are acquired do not have higher R&D expenditures (measured by the ratio of R&D to sales) than firms in the same industry that are not acquired.Also, she finds that â€Å"firms involved in mergers showed no difference in their pre- and post-merger R&D performance over those not so involved. † I know of no evidence that supports the argument that takeovers reduce R&D expenditures, even though this is a prominent argument among many of those who favor restrictions on takeovers. Free Cash Flow Theory More than a dozen separate forces drive takeover activity, including such factors as deregulation, synergies, economies of scale and scope, taxes, managerial incompetence, and increasing globalization of U. S. markets. 7 One major cause of takeover activity, the gency costs associated with conflicts between managers and 6 The â€Å"R&D Scoreboard† is an annual survey, covering companies that account for 95 percent of total private-sector R&D expenditures. The three years referenced here can be found in â€Å"R&D Scoreboard: Reagan & Foreign Rivalry Light a Fire Under Spending,† Business Week, (, July 8, 1985, p. 86 ff. ); â€Å"R&D Scoreboard: Now, R&D is Corporate America’s Answer to Japan Inc. ,† Business Week, (, June 23, 1986, p. 134 ff. ); and â€Å"R&D Scoreboard: Research Spending is Building Up to a Letdown,† Business Week, (, June 22, 1987, p. 39 ff. ). In 1984 the survey covered 820 companies; in 1985, it covered 844 companies; in 1986, it covered 859 companies. 7 Roll (1988) discusses a number of these forces. M. C. Jensen 14 1987 shareholders over the payout of free cash flow,8 has received relatively little attention. Yet it has pla yed an important role in acquisitions over the last decade. Managers are the agents of shareholders, and because both parties are selfinterested, there are serious conflicts between them over the choice of the best corporate strategy.Agency costs are the total costs that arise in such cooperative arrangements. They consist of the costs of monitoring managerial behavior (such as the costs of producing audited financial statements and devising and implementing compensation plans that reward managers for actions that increase investors’ wealth) and the inevitable costs that are incurred because the conflicts of interest can never be resolved perfectly. Sometimes these costs can be large, and when they are, takeovers can reduce them.Free Cash Flow and the Conflict Between Managers and Shareholders Free cash flow is cash flow in excess of that required to fund all of a firm’s projects that have positive net present values when discounted at the relevant cost of capital. Suc h free cash flow must be paid out to shareholders if the firm is to be efficient and to maximize value for shareholders. Payment of cash to shareholders reduces the resources under managers’ control, thereby reducing managers’ power and potentially subjecting them to the monitoring by the capital markets that occurs when a firm must obtain new capital.Financing projects internally avoids this monitoring and the possibility that funds will be unavailable or available only at high explicit prices. Managers have incentives to expand their firms beyond the size that maximizes shareholder wealth. 9 Growth increases managers’ power by increasing the resources This discussion is based on Jensen (1986a). Gordon Donaldson (1984), in a detailed study of 12 large Fortune 500 firms, concludes that managers of these firms were not driven by maximization of the value of the firm, but rather by the maximization of â€Å"corporate wealth. He defines corporate wealth as â€Å" the aggregate purchasing power available to management for strategic purposes during any given planning period†¦. this wealth consists of 9 8 M. C. Jensen 15 1987 under their control. In addition, changes in management compensation are positively related to growth. 10 The tendency of firms to reward middle managers through promotion rather than year-to-year bonuses also creates an organizational bias toward growth to supply the new positions that such promotion-based reward systems require (Baker, 1986);.The tendency for managers to overinvest resources is limited by competition in the product and factor markets that tends to drive prices toward minimum average cost in an activity. Managers must therefore motivate their organizations to be more efficient in order to improve the probability of survival. Product and factor market disciplinary forces are often weaker in new activities, however, and in activities that involve substantial economic rents or quasi-rents. 1 Activities yielding substantial economic rents or quasi-rents are the types of activities that generate large amounts of free cash flow. In these situations, monitoring by the firm’s internal control system and the market for corporate control are more important. Conflicts of interest between shareholders and managers over payout policies are especially severe when the organization generates substantial free cash flow. The problem is how to motivate managers to disgorge the cash rather than invest it below the cost of capital or waste it through organizational inefficiencies.Myers and Majluf (1984) argue that financial flexibility (unused debt capacity and internally generated funds) is desirable when a firm’s managers have better information about the firm than outside investors. Their arguments assume that managers act in the best interest of shareholders. The arguments offered here imply the stocks and flows of cash and cash equivalents (primarily credit) that management can u se at its discretion to implement decisions involving the control of goods and services† (p. 3, emphasis in original). In practical terms it is cash, credit, and other corporate purchasing power by which management commands goods and services† (p. 22). 10 Where growth is measured by increases in sales. See Murphy (1985). This positive relationship between compensation and sales growth does not imply, although it is consistent with, causality. 11 Rents are returns in excess of the opportunity cost of the permanent resources in the activity. Quasirents are returns in excess of the opportunity cost of the short-lived resources in the activity. M. C.Jensen 16 1987 that such flexibility has costs; financial flexibility in the form of free cash flow (including both current free cash in the form of large cash balances, and future free cash flow reflected in unused borrowing power) provides managers with greater discretion over resources that is often not used in the shareholder s’ interests. Therefore, contrary to Myers and Majluf, the argument here implies that eventually the agency costs of free cash flow cause the value of the firm to decline with increases in financial flexibility.The theory developed here explains (1) how debt-for-stock exchanges reduce the organizational inefficiencies fostered by substantial free cash flow; (2) how debt can substitute for dividends; (3) why â€Å"diversification† programs are more likely to be associated with losses than are expansion programs in the same line of business; (4) why mergers within an industry and liquidation-motivated takeovers will generally create larger gains than cross-industry mergers; (5) why the factors stimulating takeovers in such diverse businesses as broadcasting, tobacco, cable systems and oil are essentially identical; and (6) why bidders and some targets tend to show abnormally good performance prior to takeover.The Role of Debt in Motivating Organizational Efficiency The a gency costs of debt have been widely discussed (Jensen and Meckling (1976); Smith and Warner (1979)), but, with the exception of the work of Grossman and Hart (1980), the benefits of debt in motivating managers and their organizations to be efficient have largely been ignored. Debt creation, without retention of the proceeds of the issue, enables managers effectively to bond their promise to pay out future cash flows. Thus, debt can be an effective substitute for dividends, something not generally recognized in the corporate finance literature. 12 By issuing debt in exchange for stock, Literally, principal and interest payments are substitutes for dividends. Dividends and debt are not perfect substitutes, however, because interest is tax-deductible at the corporate level and dividends are not. 12 M. C. Jensen 17 1987 anagers bond their promise to pay out future cash flows in a way that simple dividend increases do not. In doing so, they give shareholder-recipients of the debt the ri ght to take the firm into bankruptcy court if they do not keep their promise to make the interest and principal payments. 13 Thus, debt reduces the agency costs of free cash flow by reducing the cash flow available for spending at the discretion of managers. These control effects of debt are a potential determinant of capital structure. Managers with substantial free cash flow can increase dividends or repurchase stock and thereby pay out current cash that would otherwise be invested in low-return projects or wasted.This payout leaves managers with control over the use of future free cash flows, but they can also promise to pay out future cash flows by announcing a â€Å"permanent† increase in the dividend. 14 Because there is no contractual obligation to make the promised dividend payments, such promises are weak. Dividends can be reduced by managers in the future with little effective recourse available to shareholders. The fact that capital markets punish dividend cuts wit h large stock price reductions (Charest (1978); Aharony and Swary (1980)) can be interpreted as an equilibrium market response to the agency costs of free cash flow. Brickley, Coles and Soo Nam (1987) find that firms that regularly pay extra dividends appear to have positive free cash flow. In comparison with a control group they have significantlyRozeff (1982) and Easterbrook (1984b) argue that regular dividend payments can be effective in reducing agency costs with managers by assuring that managers are forced more frequently to subject themselves and their policies to the discipline of the capital markets when they acquire capital. 14 Interestingly, Graham and Dodd (1951, Chapters 32, 34 and 36) in their treatise, Security Analysis, place great importance on the dividend payout in their famous valuation formula: V=M(D+. 33E). (See p. 454. ) V is value, M is the earnings multiplier when the dividend payout rate is a â€Å"normal two-thirds of earnings,† D is the expected di vidend, and E is expected earnings.In their formula, dividends are valued at three times the rate of retained earnings, a proposition that has puzzled many students of modern finance (at least of my vintage). The agency cost of free cash flow that leads to over retention and waste of shareholder resources is consistent with the deep suspicion with which Graham and Dodd viewed the lack of payout. Their discussion (chapter 34) reflects a belief in the tenuous nature of the future benefits of such retention. Although they do not couch the issues in terms of the conflict between managers and shareholders, the free cash flow theory explicated here implies that their beliefs, sometimes characterized as a preference for â€Å"a bird in the hand is worth two in the bush,† were perhaps well founded. 13 M. C. Jensen 18 1987 igher cash plus short-term investments, and earnings plus depreciation, relative to their total assets. They also have significantly lower debt-to-equity ratios. Th e issuance of large amounts of debt to buy back stock sets up organizational incentives to motivate managers to pay out free cash flow. In addition, the exchange of debt for stock helps managers overcome the normal organizational resistance to retrenchment that the payout of free cash flow often requires. The threat of failure to make debt-service payments serves as a strong motivating force to make such organizations more efficient. Stock repurchase for debt or cash also has tax advantages.Interest payments are tax-deductible to the corporation, that part of the repurchase proceeds equal to the seller’s tax basis in the stock is not taxed at all, and prior to 1987 tax rates on capital gains were favorable. Increased leverage also has costs. As leverage increases, the usual agency costs of debt, including bankruptcy costs, rise. One source of these costs is the incentive to take on projects that reduce total firm value but benefit shareholders through a transfer of wealth fro m bondholders. These costs put a limit on the desirable level of debt. The optimal debt/equity ratio is the point at which firm value is maximized, the point where the marginal costs of debt just offset the marginal benefits. The debt created in a hostile takeover (or takeover defense) of a firm suffering severe agency costs of free cash flow need not be permanent.Indeed, sometimes â€Å"overleveraging† such a firm is desirable. In these situations, leveraging the firm so highly that it cannot continue to exist in its old form yields benefits by providing motivation for cuts in expansion programs and the sale of divisions that are more valuable outside the firm. The proceeds are used to reduce debt to a more normal or permanent level. This process results in a complete rethinking of the organization’s strategy and structure. When it is successful, a much leaner, more efficient, and competitive organization results. M. C. Jensen 19 1987 The control hypothesis does not i mply that debt issues will always have positive control effects.For example, these effects will not be as important for rapidly growing organizations with large and highly profitable investment projects but no free cash flow. Such organizations will have to go regularly to the financial markets to obtain capital. At these times the markets have an opportunity to evaluate the company, its management, and its proposed projects. Investment bankers and analysts play an important role in this monitoring, and the market’s assessment is made evident by the price investors pay for the financial claims. The control function of debt is more important in organizations that generate large cash flows but have low growth prospects, and it is even more important in organizations that must shrink.In these organizations the pressure to waste cash flows by investing them in uneconomic projects is most serious. Evidence from Financial Transactions Free cash flow theory helps explain previously puzzling results on the effects of various financial transactions. Smith (Smith, 1986, tables 1 to 3) summarizes more than 20 studies of stock price changes at announcements of transactions that change capital structure as well as various other dividend transactions. These results and those of others are presented in table 2. For firms with positive free cash flow, the theory predicts that stock prices will increase with unexpected increases in payouts to shareholders and decrease with unexpected decreases in payouts.It also predicts that unexpected increases in demand for funds from shareholders via new issues will cause stock prices to fall. The theory also predicts stock prices will increase with increasing tightness of the constraints binding the payout of future cash flow to shareholders and decrease with reductions in the tightness of these constraints. These predictions do not apply to those firms with more profitable projects than cash flow to fund them. M. C. Jensen 20 1987 The predictions of free cash flow theory are consistent with all but three of the 32 estimated abnormal stock price changes summarized in table 2, and one of the inconsistencies is explained by another phenomenon.Panel A of table 2 shows that stock prices rise by a statistically significant amount with announcements of the initiation of cash dividend payments, increases in dividends and specially designated dividends, and fall by a statistically significant amount with decreases in dividend payments. (All coefficients in table 2 are significantly different from zero unless noted with an asterisk. ) Panel B shows that security sales and retirements that raise cash or pay out cash and simultaneously provide offsetting changes in the constraints bonding the payout of future cash flow are all associated with returns that are insignificantly different from zero.The insignificant return on retirement of debt fits the theory because the payout of cash is offset by an equal reduction in th e present value of promised future cash payouts. If debt sales are not associated with changes in the expected investment program, the insignificant return on announcement of the sale of debt and preferred also fits the theory. The acquisition of new funds with debt or preferred stock is offset exactly by a commitment bonding the future payout of cash flows of equal present value. If the funds acquired through new debt or preferred issues are invested in projects with negative net present values, the abnormal stock price change will be negative. If they are invested in projects with positive net present values, the abnormal stock price change will be positive.Sales of convertible debt and preferred securities are associated with significantly negative stock price changes (panel C). These security sales raise cash and provide little effective bonding of future cash flow payments; when the stock into which the debt is convertible is worth more than the face value of the debt, manageme nt has incentives to call the convertible securities and force conversion to common. M. C. Jensen 21 1987 Panel D shows that, with one exception, security retirements that pay out cash to shareholders increase stock prices. The price decline associated with targeted large block repurchases (often called greenmail) is highly likely to be due to the reduced probability that a takeover premium will be realized.These transactions are often associated with standstill agreements in which the seller of the stock agrees to refrain from acquiring more stock and from making a takeover offer for some period into the future (Mikkelson and Ruback (1985; 1986); Dann and DeAngelo (1983); and Bradley and Wakeman (1983);). Panel E summarizes the effects of security sales and retirements that raise cash and do not bond future cash flow payments. Consistent with the theory negative abnormal returns are associated with all such changes, although the negative returns associated with the sale of common t hrough a conversion-forcing call are statistically insignificant.Panel F shows that all exchange offers or designated use security sales that increase the bonding of payout of future cash flows result in significantly positive increases in common stock prices. These include stock repurchases and exchange of debt or preferred for common, debt for preferred, and income bonds for preferred. The twoday gains range from 21. 9 percent (debt for common) to 1. 6 percent for income bonds and 3. 5 percent for preferred. 15 The theory predicts that transactions with no cash flow and no change in the bonding of payout of future cash flows will be associated with returns that are insignificantly different from zero. Panel G of table 2 shows that the evidence is mixed; 15 The two-day returns of exchange offers and self-tenders can be affected by the offer.However, if there are no real effects or tax effects, and if all shares are tendered to a premium offer, then the stock price will be unaffecte d by the offer and its price effects are equivalent to those of a cash dividend. Thus, when tax effects are zero and all shares are tendered, the two-day returns are appropriate measures of the real effects of the exchange. In other cases the correct returns to be used in these transactions are those covering the period from the day prior to the offer announcement to the day after the close of the offer (taking account of the cash payout). See, for example, Rosenfeld (1982), whose results for the entire period are also consistent with the theory. M. C. Jensen 22 1987 he returns associated with exchange offers of debt for debt are significantly positive and those for designated-use security sales are insignificantly different from zero. All exchanges and designated-use security sales that have no cash effects but reduce the bonding of payout of future cash flows result, on average, in significant decreases in stock prices. These transactions include the exchange of common for debt or preferred or preferred for debt, or the replacement of debt with convertible debt and are summarized in Panel H. The two-day losses range from 7. 7 percent (preferred for debt) to 1. 1 percent (common for debt). In summary, the results in table 2 are remarkably consistent with free cash flow theory hich predicts that, except for firms with profitable unfunded investment projects, stock prices will rise with unexpected increases in payouts to shareholders (or promises to do so) and will fall with reductions in payments or new requests for funds from shareholders (or reductions in promises to make future payments). Moreover, the size of the value changes seems to be positively related to the change in the tightness of the commitment bonding the payment of future cash flows. For example, the effects of debtfor-preferred exchanges are smaller than the effects of debt-for-common exchanges. Tax effects can explain some of the results summarized in table 2, but not all.For example, the ex change of preferred for common, or replacement of debt with convertible debt, has no tax effects and yet is associated with price increases. The last column of table 2 denotes whether the individual coefficients are explainable by pure corporate tax effects. The tax theory hypothesizes that all unexpected changes in capital structure that decrease corporate taxes increase stock prices and vice versa. 16 Therefore, increases in dividends and reductions of debt interest should cause stock prices to fall, and vice versa. 17 Fourteen of the 32 coefficients are inconsistent with the corporate tax See, however, Miller (1977) who argues that allowing for personal tax effects and the equilibrium response of firms implies that no tax effects will be observed. 7 Ignoring potential tax effects due to the 85 percent exclusion of dividends received by corporations on holdings of preferred stock. 16 M. C. Jensen 23 1987 Table 23 Summary of Two-Day Average Abnormal Stock Returns Associated with th e Announcement of Various Dividend and Capital Structure Transactionsa Average Sample Size Average Abnormal Return (Percent) Free Cash Flow Theory Agreement with Tax Predicted Agreement Theory Sign with Theory? Type of Transaction A. Dividend changes that change the cash paid to shareholders Dividend initiation1 Dividend increase2 Specially designated dividend Dividend decrease2 3 Security Issued Security Retired 160 281 164 48 3. 7% 1. 0 2. 1 -3. 6 + + + – es yes yes yes no no no no B. Security sales (that raise cash) and retirements (that pay out cash) that simultaneously provide offsetting changes in the constraints bonding future payment of cash flows Security sale (industrial) 4 Security sale (utility) 5 Security sale (industrial) 6 Security sale (utility) Call8 7 debt debt preferred preferred none none none none none debt none none none common common common common 248 140 28 251 133 74 54 9 147 182 15 68 – 0. 2* -0. 1* -0. 1* -0. 1* -0. 1* -2. 1 -1. 4 -1. 6 15. 2 3. 3 1. 1 -4. 8 0 0 0 0 0 – – – + + + + yes yes yes yes yes yes yes yes yes yes yes no b no no yes yes no no no no yes yes yes no b C.Security sales that raise cash and bond future cash flow payments only minimally Security sale (industrial) 4 conv. debt 7 Security sale (industrial) conv. preferred 7 Security sale (utility) conv. preferred D. Security retirements that pay out cash to shareholders Self tender offer 9 Open market purchase10 Targeted small holdings11 Targeted large block repurchase12 none none none none M. C. Jensen 24 1987 E. Security sales or calls that raise cash and do not bond future cash flow payments Security sale (industrial) 13 common none Security sale (utility)14 common none Conversion-forcing call15 common conv. preferred Conversion-forcing call15 common conv. debt F.Exchange offers, or designated use security sales that increase the bonding of payout of future cash debt common Designated use security sale16 Exchange offer 17 debt comm on 17 Exchange offer preferred common 17 Exchange offer debt preferred Exchange offer 18 income bonds preferred G. Transaction with no change in bonding payout of future cash flows Exchange offer 19 debt 20 Designated use security sale debt debt debt 215 405 57 113 flows 45 52 10 24 18 36 96 -3. 0 -0. 6 -0. 4* -2. 1 21. 9 14. 0 8. 3 3. 5 1. 6 0. 6 0. 2* -2. 4 -2. 6 -7. 7 -4. 2 -1. 1 – – – – + + + + + 0 0 – – – – – yes yes no yes yes yes yes yes yes no yes yes yes yes yes yes yes yes yes yes yes yes no yes yes no yes yes no yes yes yes H.Exchange offers, or designated use security sales that decrease the bonding of payout of future cash flows Security sale 20 conv. debt debt 15 Exchange offer 17 common preferred 23 17 Exchange offer preferred debt 9 20 Security sale common debt 12 Exchange offer 21 common debt 81 a Returns are weighted averages, by sample size, of the returns reported by the respective studies All returns are significantly different from zero unless noted otherwise by *. b Explained by the fact that these transactions are frequently associated with the termination of an actual or expected control bid. The price decline appears to reflect the loss of an expected control premium. Source: 1 Asquith and Mullins (1983). 2 Charest (1978); Aharony and Swary (1980). 3 From Brickley (1983). Dann and Mikkelson (1984); Eckbo (1986); Mikkelson and Partch (1986). 5 Eckbo (1986). 6 Linn and Pinegar (1985); Mikkelson and Partch (1986). 7 Linn and Pinegar (1985). 8 Vu (1986). 9 Dann (1981); Masulis (1980); Vermaelen (1981); Rosenfeld (1982). 10 Dann (1980); Vermaelen (1981). 11 Bradley and Wakeman (1983). 12 Calculated by Smith (1986), table 4, from Dann and DeAngelo (1983); Bradley and Wakeman (1983). 13 Asquith and Mullins (1986); Kolodny and Suhler (1985); Masulis and Korwar (Korwar and Masulis); Mikkelson and Partch (1986). 14 Asquith and Mullins (1986); Masulis and Korwar (1986); Pettway and R adcliffe (1985). 15 Mikkelson (1981). 16 Others with more than 50% debt Masulis (1980). 17 Masulis (1983).These returns include announcement days of both the original offer and, for about 40 percent of the sample, a second announcement of specific terms of the exchange 18 McConnell and Schlarbaum (1981). 19 Dietrich (1984). 20Eckbo (1986); Mikkelson and Partch (1986). 21Rogers and Owers (1985); Peavy and Scott (1985); Finnerty (1985). (Allen, 1987; Auerbach and Reishus, 1987; Biddle and Lindahl, 1982; Bradley, Desai, and Kim, 1983; Bradley and Rosensweig, 1986; Comment and Jarrell, 1986; 1986; Crovitz, 1985; Easterbrook, 1984a; Eckbo, 1985; 1985; Fama and Jensen, 1983a, b, 1985; Franks, Harris, and Mayer, 1987; Golbe and White, 1987; Herzel, Colling, and Carlson, 1986; Holderness and Sheehan, 1985; 1985; Jarrell, Poulsen, and Davidson, 1985; Jensen, 1985, 1986b; Jensen and Smith, 985; Kaplan and Roll, 1972; Koleman, 1985; Lambert and Larcker, 1985; Malatesta and Walkling, 1985; Mart in, 1985; Morrison, 1982; Mueller, 1980; Myers, 1977; Office of the Chief Economist, 1984, 1985b, 1986; Paulis, 1986; Ravenscraft and Scherer, 1985a, b; Ricks, 1982; Ricks and Biddle, 1987; Ruback, 1988; Ryngaert, 1988; Shoven and Simon, 1987; Sunder, 1975; You et al. ) Jensen 25 1987 hypothesis. Simple signaling effects, where the payout of cash signals the lack of present and future investments promising returns in excess of the cost of capital, are also inconsistent with the results-for example, the positive stock price changes associated with dividend increases and stock repurchases. If anything, the results in table 2 seem too good, for two reasons.The returns summarized in the table do not distinguish firms that have free cash flow from those that do not have free cash flow, yet the theory says the returns to firms with no free cash flow will behave differently from those which do. In addition, only unexpected changes in cash payout or the tightness of the commitments bonding the payout of future free cash flow should affect stock prices. The studies summarized in table 2 do not, in general, control for the presence or absence of free cash flow or for the effects of expectations. If free cash flow effects are large and if firms on average are in a positive free cash flow position, the predictions of the theory will hold for the simple sample averages. To see how the agency costs of free cash flow can be large enough to show up in the uncontrolled tests summarized in table 2, consider the graph of equilibrium firm M.C. Jensen 26 1987 value and free cash flow in figure 1. Figure 1 portrays a firm whose manager values both firm value (perhaps because stock options are part of the compensation package) and free cash flow. The manager, however, is willing to trade them off according to the given indifference curves. By definition, firm value reaches a maximum at zero free cash flow. The point (V*, F*) represents the equilibrium level of firm value and free ca sh flow for the manager. It occurs at a positive level of free cash flow and at a point where firm value is lower than the maximum possible. The difference Vmax – V* is the agency cost of free cash flow.Because of random factors and adjustment costs, firms will deviate temporarily from the optimal F*. The dashed line in figure 1 portrays a hypothetical rectangular distribution of free cash flow in a cross section of firms under the assumption that the typical firm is run by managers with preferences similar to those portrayed by the given indifference curves. Changes in free cash flow (or the tightness of constraints binding its payout) will be positively related to the value of the firm only for the minority of firms in the cross section with negative free cash flow. These are the firms lying to the left of the origin, 0. The relation is negative for all firms in the range with positive free cash flow.Given the hypothetical rectangular distribution of firms in figure 1, the majority of firms will display a negative relation between changes in free cash flow and changes in firm value. As a result the average price change associated with movements toward (V*, F*) will be negatively related to changes in free cash flow. If the effects are so pervasive that they show up strongly in the crude tests of table 2, the waste due to agency problems in the corporate sector is probably greater than most scholars have thought. This waste is one factor contributing to the high level of activity in the corporate control market over the past decade. More detailed tests of the propositions that control for growth prospects and expectations will be interesting. M. C. Jensen 27 1987Evidence from Going-Private and Leveraged Buyout Transactions Many of the benefits in going-private and leveraged buyout transactions seem to be due to the control function of debt. These transactions are creating a new organizational form that competes successfully with the open corporate form because of advantages in controlling the agency costs of free cash flow. In 1985, going-private and leveraged buyout transactions totaled $37. 4 billion and represented 32 percent of the value of all public acquisitions. 18 Most studies have shown that premiums paid for publicly held firms average over 50 percent,19 but in 1985 the premiums for publicly held firms were 31 percent (Grimm, 1985). Leveraged buyouts are frequently financed with high debt; 10:1 ratios of debt to equity are not uncommon, and they average 5. 5:1 (Schipper and Smith (1986); Kaplan (1987); and DeAngelo and DeAngelo (1986)). Moreover, the use of â€Å"strip financing† and the allocation of equity in the deals reveal a sensitivity to incentives, conflicts of interest, and bankruptcy costs. Strip financing, the practice in which investors hold risky nonequity securities in approximately equal proportions, limits the conflict of interest among such securityholders and therefore limits bankruptcy costs. T op managers and the sponsoring venture capitalists hold disproportionate amounts of equity. A somewhat oversimplified example illustrates the organizational effects of strip financing. Consider two firms identical in every respect except financing.Firm A is entirely financed with equity, and Firm B is highly leveraged with senior subordinated debt, convertible debt, and preferred as well as equity. Suppose Firm B securities are sold only in strips; that is, a buyer purchasing a certain percentage of any security must purchase the same percentage of all securities, and the securities are â€Å"stapled† together See W. T. Grimm, Mergerstat Review (1985, Figs. 29, 34 and 38). See DeAngelo, DeAngelo and Rice (1984), Lowenstein (1985), and Schipper and Smith (1986). Lowenstein also mentions incentive effects of debt but argues tax effects play a major role in explaining the value increase. 19 18 M. C. Jensen 28 1987 o they cannot be separated later. Security holders of both firms have identical unlevered claims on the cash flow distribution, but organizationally the two firms are very different. If Firm A managers withhold dividends to invest in value-reducing projects or if they are incompetent, the shareholders must use the clumsy proxy process to change management or policies. In Firm B, strip holders have recourse to remedial powers not available to the equity holders of Firm A. Each Firm B security specifies the rights its holder has in the event of default on its dividend or coupon payment; for example, the right to take the firm into bankruptcy or to have board representation.As each security above equity goes into default, the strip holder receives new rights to intercede in the organization. As a result, it is quicker and less expensive to replace managers in Firm B. Moreover, because every security holder in the highly leveraged Firm B has the same claim on the firm, there are no conflicts between senior and junior claimants over reorganization of the claims in the event of default; to the strip holder it is a matter of moving funds from one pocket to another. Thus, Firm B will not go into bankruptcy; a required reorganization can be accomplished voluntarily, quickly, and with less expense and disruption than through bankruptcy proceedings. The extreme form of strip financing in the example is not normal practice.Securities commonly subject to strip practices are often called â€Å"mezzanine† financing and include securities with priority superior to common stock yet subordinate to senior debt. This arrangement seems to be sensible, because several factors ignored in our simplified example imply that strictly proportional holdings of all securities is not desirable. For example, IRS restrictions deny tax deductibility of debt interest in such situations and bank holdings of equity are restricted by regulation. Riskless senior debt need not be in the strip because there are no conflicts with other claimants in the event of reorganization when there is no probability of default on its payments. M. C. Jensen 29 1987Furthermore, it is advantageous to have the top-level managers and venture capitalists who promote leveraged buyout and going-private transactions hold a larger share of the equity. Top-level managers on average receive over 30 percent of the equity, and venture capitalists and the funds they represent generally retain the major share of the remainder (Schipper and Smith (1986); Kaplan (1987)). The venture capitalists control the board of directors and monitor the managers. Both managers and venture capitalists have a strong interest in making the venture successful because their equity interests are subordinate to other claims. Success requires (among other things) implementation of changes to avoid investment in low-return projects in order to generate the cash for debt service and to increase the value of equity.Finally, when the equity is held by a small number of people, efficiencies in risk-bearing can be achieved by placing more of the risk in