Agency costs of corporate risk management.
While risk management surely offers benefits, it may have costs as well. However, there has been very little discussion of the costs of risk management, apart from the transaction costs and sometimes-important implementation details of executing risk management programs.(3) This paper establishes a potentially important cost of risk management in exacerbating agency conflicts, leading firms to poor investment decisions. Quite briefly, risk management can be used to facilitate the protection of managers' "pet" projects that enhance their welfare, but reduce shareholder value. More specifically, this potential cost arises from cash-flow hedging. The underlying principles of cashflow hedging have a long heritage, advanced first by Donaldson (1961), discussed by Lessard (1991), and most recently substantially refined and extended by Froot, Scharfstein, and Stein (1993, 1994). The cashflow hedging concept recommends that managers engineer their operating cash flows so as to be able to continue value-enhancing investment programs without having to resort to expensive capital markets. Managers grasp the appeal of the strategy easily; to paraphrase one CFO's excited interpretation when I was teaching the concept in an executive program, "If I can avoid going to a bank or Wall Street when I'm in trouble, that's great!"
This over-enthusiastic endorsement of cash-flow hedging may suggest a problem. If projects that managers seek to protect are negative-NPV investments to shareholders, but managers support them nevertheless because of some private benefits they will enjoy, the absence of capital market scrutiny can lead to improper resource allocation and the destruction of shareholder value. While cash-flow hedging takes away the need to obtain "expensive" financing, it does so by eliminating the discipline that the capital markets would impose on firms attempting to fund poor projects. Risk management strategies that allow managers to bypass these external monitors can heighten existing agency conflicts between managers and shareholders.
The remainder of this paper develops the proposition that risk management may exacerbate agency conflicts between managers and shareholders. Section I reviews the arguments for cash-flow hedging, as developed by Froot, Scharfstein, and Stein (1993, 1994). Section II revisits a hypothetical example that they present and discusses the costs of the risk management programs using a counterexample framed in terms of actual corporate-investment strategies. I present the results in the form of a simple numerical example (an algebraic model that develops the arguments more formally is available from the author). Section III discusses anecdotal evidence that suggests that the concerns raised here may be relevant at some firms, and it concludes by discussing the implications for managers and researchers of risk management.
I. The Benefits of Risk Management and Cash-Flow Hedging
Many real benefits of risk management have been advanced in theory, and in some cases, proven empirically. As a starting point, recognize that if there were no market imperfections of some sort, hedging would neither increase nor decrease firm value.(4) Rationales for risk management usually flow from a consideration of the real imperfections that firms face, which include taxes, financial distress, and costly external financing. For example, Smith and Stulz (1985) note how nonlinear (or more precisely convex or progressive) tax schedules give rise to a rationale for hedging, and Graham and Smith (1997) document the magnitude of the tax savings from hedging. Furthermore, just as nonlinear tax policies induce hedging, so do deadweight costs of financial distress create discontinuities that make hedging optimal (see Smith and Stulz, 1985).
More recent justifications of risk management have focused on imperfections in raising external funds that lead firms to make suboptimal investment decisions. The notion of cash-flow hedging suggests that firms should engineer their internal cash flows to meet their optimal investment needs so as to avoid having to bear the deadweight costs of external finance. Froot, Scharfstein, and Stein (hereafter FSS, 1994) summarize the fundamental notion of cash-flow hedging:
In general, the supply of internally generated funds does not equal the investment demand for funds. Sometimes there is excess supply; sometimes there is a shortage. Because external financing is costly, this imbalance shifts investment away from the optimal level. Risk management can reduce this imbalance and the resulting investment distortion. It enables companies to better align their demand for funds with their internal supply of funds. That is, risk management lets companies transfer funds from situations in which they have an excess supply to situations in which they have a shortage. In essence, it allows companies to borrow from themselves.
To illustrate their point, FSS (1994) consider hypothetical firms, including one they call Omega Oil. Figure 1 reproduces their graphical analysis of Omega Oil's conundrum. The firm's supply of internal funds (the lighter, solid line) is its cash flow from operations, which is positively related to the price of oil. The firm's optimal demand for funds (the dashed line) is its desired investment, which is also a positive function of the oil price. However, while both the internal supply and demand are positively related to the oil price, they are not precisely equal.
In the absence of risk management, Omega Oil would have too much cash when oil prices are high, and too little when oil prices are low. In the latter instance, it would make a suboptimal investment decision rather than going to the capital markets to obtain new funds to fund exploration and development. The firm's aversion to obtaining external financing is driven by its costs: information asymmetries between outsiders and insiders make equity and debt expensive. More precisely, FSS posit that deadweight costs to external financing could arise from multiple sources:
First, they could originate in costs of bankruptcy and financial distress, which include direct costs (e.g., legal fees) as well as indirect costs (e.g., decreased product-market competitiveness and underinvestment). Second, such costs could arise from informational asymmetries between managers and outside investors. Or to the extent that managers are not full residual claimants, there may be agency costs associated with motivating and monitoring managers who resort to certain types of external financing. Finally, managers may obtain private benefits from limiting their dependence on external investors. Thus, even though there are no observable costs to external finance, management may act as though external financing has real economic costs. Regardless of which interpretation one chooses, the deadweight costs should be an increasing function of the amount of external finance (FSS, 1993).
Later in their paper, they pose most of these costs as arising from costly state verification as developed by Townsend (1979) and Gale and Hellwig (1985), in which cash flows are costlessly observable to insiders, but observable to outside investors at a cost. As a result of the deadweight costs of financing, firms will either shun external financing altogether or underinvest relative to the optimal level in the absence of these costs. Through its risk management programs (whose implementation is assumed to have no deadweight costs), the firm can fashion a pattern of available cash flows that matches its demands for funds to invest, equating the marginal cost and marginal benefit of an additional dollar for investment.
One can also note that, under this hypothetical example, risk management can unwittingly serve to eliminate a potential agency problem. By hedging, the firm may give away the "extra" cash flow that it cannot use when oil prices are high in order to spend more when oil prices are low. Under this scheme, managers are no longer able to "overspend" when they are flush with extra cash. While cast in a different setting, this conclusion was reached by Stulz (1990), who found that reducing volatility (and simultaneously increasing debt) can increase the value of the firm when managers have a tendency to overinvest. In his setting, a collateral benefit of risk management is this reduction in agency problems, by taking resources away from managers when they might spend them unwisely.
II. An Alternative Stylized Firm: Alpha Omega Oil
While risk management may enhance shareholder value, it may also be used by poorly diversified managers who might have private interests in managing risk to maximize their own utility (see Stulz, 1984). Some empirical evidence has provided support for the proposition that corporate risk management choices might be the product of managers' risk aversion and their exposure to the success of the firm, as provided by their compensation contracts and investments.(5) Other research looks at whether the form of compensation contracts - which in turn induce different risk management choices by managers - can be considered as optimal, in that they encourage managers to hedge when that would enhance firm value.(6)
The premise of this recent work is to discern whether risk management choices made by managers might be consistent with shareholder-value maximization. In the following discussion, I establish the possibility that risk management choices made by managers might exacerbate agency conflicts between them and shareholders, by insulating the firm from the capital markets.
To begin, it is useful to characterize the nature of the agency conflicts between managers and shareholders. Rather than ponder the plight of FSS's hypothetical Omega Oil, whose managers are all dedicated servants of the shareholders, consider the alleged situation of actual oil companies in the 1980s.(7) Oil company managers argued that continued investment in exploration and development (e&d) for new oil was warranted. Their critics countered that continued investment destroyed shareholder value, and was the result of excessive free cash flow available to managers, the poor judgment of entrenched managers, and ineffective corporate oversight and discipline?
As a counterexample to Omega Oil, consider the situation of another hypothetical oil company, "Alpha Omega Oil" (AOO), whose managers do not necessarily act in the best interests of shareholders. More concretely, suppose that the optimal investment program for Alpha Omega Oil (as judged by its shareholders) is represented by the dotted line schedule drawn by FSS for Omega Oil, but that the e&d program sought by managers is the schedule represented by the dark solid line in Figure 1. AOO's managers would prefer to spend all of the firm's cash flow on e&d when oil prices are high and maintain this spending program at no less than the existing levels when oil prices drop. AOO's managers are reluctant to reduce spending below current levels because cutting programs is not personally enjoyable.(9) However, when cash flow is strong, they would prefer to spend all of the firm's cash flow on e&d so that they can grow the assets under their control.
Were shareholders to have the decision rights in the firm, they would pursue an investment program similar to Omega Oil's. Yet the managers of AOO set the firm's investment and risk management strategy. By purchasing a put option on oil in advance, they can protect against a shortfall between the firm's supply of funds and their demand for funds. This would completely align the combination of the firm's internal funds plus option gains with the managers' desired investment program needs. As FSS argue, it would do so without forcing the firm to expend resources on the deadweight activity of raising external funds. More concretely, the managers would not have to make presentations to skeptical bankers, file detailed registration statements with the SEC, or endure grueling cross-country road show presentations. Furthermore, they would never risk being turned down for funds to continue their investment strategy, or being charged "high" rates by the capital markets.
This hedging is not costless. First, the firm would need to pay the premium for the purchase of the put on oil prices, which would partially depress spending in all states. However, the real cost of the hedging is the distortion to the investment process. While hedging makes the managers' lives easier and more predictable, it deprives the shareholders of AOO of a valuable set of allies - external providers of capital. Proponents of the capital markets as monitors argue that forcing firms (or more accurately, managers) to present their investment needs to bankers, potential buyers of debt, or potential equity holders serves to discipline managers (see Jensen and Meckling, 1976; Jensen, 1986; Rozeff, 1982; and Easterbrook, 1984). Forcing AOO's managers to explain to capital suppliers why the firm needs to maintain its investment program if oil prices drop may exert some extra restraint on managers. Were managers to petition the capital markets for funds to invest when investment opportunities are weak, they may either fail to receive "permission" in the form of new capital, or anticipating rejection, they may ex ante accept a lower level of investment reluctantly. It is true that potential investors must work to scrutinize the proposals put forth by management, and this costly state verification creates a deadweight cost that cashflow hedging seeks to avoid. However, while there are costs of having the external markets try to uncover what is going on in the firm, there are benefits as well: the external capital markets may sometimes refuse to fund poor investment projects.
Thus, managers may adopt cash-flow hedging so as to avoid scrutiny and potential rejection by the capital markets. By hedging, AOO's managers will have more than enough cash to carry out their investment program, a program much larger than the optimal program that its virtuous twin, Omega Oil, would pursue. AOO's shareholders (and the market for corporate control) must work to force the managers of AOO to scale back the investment program, perhaps by disgorging cash.(10) Risk management is transformed into a vehicle to deliver excess free cash flow to AOO's managers and, in so doing, heightens the agency conflict between managers and shareholders.
The behavior described above can be modeled more formally, and an Appendix to this paper (which is available upon request) analyzes a model of firm behavior that captures the situation described above. (11) The stylized three-period model, closely patterned after work done by Hart (1993) and Hart and Moore (1996), can be represented by the time line in Table 1. Utility-maximizing managers make three decisions: hedging, financing, and investment. Risk-neutral providers of external financing also make a decision: whether or not to supply additional capital to the firm, if it is needed. The key results of the model are driven by three sets of parameters: the native risk of the project (its returns in the low and high states of the world (L and H), as well as the probabilities of these states), the deadweight cost of raising finance represented as a fraction of the gross funds raised (R), and the severity of the agency conflict represented by the fraction of the value of the completed project "consumed" by managers in the form of private benefits (M).(12)
A simple numerical example can demonstrate when this disagreement might be most severe. At t = 0, the firm has an asset-in-place which will generate either $20 or $180 with equal probability at t = 1. At t = 0, the firm can commit to swap these cash flows for a certain cash flow of $100. At t = 1, the firm will be offered an investment opportunity requiring a $100 investment that will earn an excess return of 5%. If the firm cannot raise the full $100, it can only invest in projects that earn a zero excess return. If it has more than $100, it earns zero excess returns on investments beyond the $100. At t = 1, if the firm does not have sufficient internal funds to complete the project, it can attempt to raise money from risk-neutral investors and pay a return of 8%, the deadweight costs of financing. However, these investors will only fund the project if the firm will have enough funds at t = 2 to repay the balance plus deadweight costs.
Consider our two hypothetical firms. In the first case (Omega Oil), managers can be trusted to act in the best interests of shareholders, and make investment, financing, and hedging decisions to maximize shareholder value. In the second (Alpha Omega Oil), the manager seeks to maximize his or her private utility and can benefit privately by consuming a fraction of the total value of the completed project value, but none of the funds not invested in the project. The logic for this setup is that funds invested in the positive-NPV project are fully under the manager's control, and he or she can siphon off a portion of their value for private use. These actions are observable but not verifiable (i.e., the shareholders and other outside capital suppliers can observe them, but not write contracts in advance conditioned upon them nor sue after the fact to recover this private consumption by the manager). This private consumption by the manager takes place prior to t = 2, so that the manager is a super-senior claimant upon the firm's assets.(13) I parameterize the [TABULAR DATA FOR TABLE 1 OMITTED] degree of agency costs in the second firm by the fraction of the funds invested in the positive-NPV project that are privately consumed by the manager. For this example, I set this private consumption fraction at 20% to reflect large, but not unrealistic, differences in the incentives of managers and shareholders.(14) Thus, if the project is fully funded (at $100), the manager will privately consume $20 of the project's final cash flows before funds can be distributed to others.
Table 2 represents the payoffs to the shareholders and managers of these two firms, and their optimal choices. Firm 1, Omega Oil, faces only deadweight costs of external financing, and for it, shareholder value is maximized by hedging, which allows the firm to avoid costly external financing. However, Firm 2, Alpha Omega Oil, also has to contend with agency conflicts between managers and shareholders. Here, shareholders would prefer that the firm not hedge, because hedging would allow the continuation of a pet project which managers prefer, but which creates no value for shareholders net of the managerial consumption.
This simple example illustrates the rather intuitive result that arises from the more formal model: if deadweight costs are large and agency conflicts small, then both managers and shareholders would prefer that the firm hedge. However, when agency conflicts become large, shareholders and managers can reach contrary conclusions about hedging. In particular, managers and shareholders would disagree when agency conflicts are large (i.e., the managers consume more than the gross return to the project: M [greater than] R) and when capital market constraints would be [TABULAR DATA FOR TABLE 2 OMITTED] binding without hedging (i.e., the capital markets would have turned down a t = 1 request for additional financing). In these circumstances, managers prefer that the firm had hedged, and shareholders prefer that the firm had not hedged.
The analysis would suggest that one cannot make a blanket statement as to whether hedging is good or bad. While proponents of agency-based models and information-asymmetry models sometimes try to portray them as being separate, both are at work here. From the perspective of shareholders, the "first best" solution is a function of the severity of the informational asymmetries (D), the agency problems with management (M), the underlying technology of the project (R), and the volatility of the firm's existing cash flow (H, L). Without agency problems, the normal prescriptions of value-maximizing cash-flow hedging apply: hedging improves shareholder welfare, because the failure to hedge forces the firm to bear deadweight costs of external financing, or to forego positive-NPV projects. Even with large agency costs but still a financeable project, the FSS cash-flow hedging prescription continues to hold. However, when the capital market's reluctance to fund a project is related to large agency costs, then shareholders would prefer to use the capital markets to discipline managers. In certain instances in which it looks as if the capital markets had "failed" by refusing to fund a project, in reality, the capital markets are the strongest ally of shareholders facing agency conflicts with managers, confirming the arguments made by many prior authors, including Jensen (1986) and Easterbrook (1984). This result has been previously noted, although in a different context. Jensen (1986) discusses how diversification, which allows firms to have unused borrowing power and large free cash flow, can set the stage for firms to "undertake low-benefit or even value-destroying mergers."(15) The arguments above suggest that other forms of risk management (such as hedging and insurance) can have similar adverse incentive effects.
III. But Should We Worry?
The arguments above are what some would charitably call a "possibility theorem," in that they establish the potential for managerial mischief. Ultimately, it is a factual or empirical matter whether this potential danger is material. Perhaps academics may someday be able to use a combination of field studies and statistical analysis to determine how material the concerns raised are, and more positively, to establish the benefits of risk management.
However, this does not mean that managers and those who oversee them can ignore the problem - or forgo the opportunities to manage risk - until academics are able to prove its relevance. From my experiences with managers, it seems that at least some risk management programs are explicitly designed to protect manager's "pet projects" (i.e., those from which the managers derive personal benefits, but which might be judged more critically by shareholders). A few scattered experiences, illustrated by the following stories, highlight the real nature of these concerns.(16)
In one firm, the division manager explained to me that while his parent company had a policy against hedging, he was able to hedge against fluctuations in major input costs without having to report it to his superiors. He purposely hid his risk management choices from senior managers. His reason for this decision was not well-articulated, but one can imagine that it related to either his compensation schedule or his ability to garner additional resources in the firm's capital budgeting system (or to avoid losing resources). In an interview in Risk Magazine, the head of corporate finance at a major UK retail chain expressed his opinion on the popularity of managerial-directed risk management programs: "Over the past few years I have become increasingly convinced that one of the greatest incentives to hedge is to meet preset budgets so that management can collect their bonuses."(17)
Recently, I interviewed a risk management consultant from a very large firm, and he fully concurred with the UK manager. He noted that in his practice, he found that the "risk" being managed by risk management programs is "career or employment risk." Managers, not boards, usually make risk management decisions. In making these decisions, they often seek to make certain that projects that will advance their careers are not derailed by sudden risks, and he judged this tendency to be "rampant," not just involving small companies but large ones as well. "Financial theory gives way to expediency," he said.
To the extent that the projects that managers seek to protect enhance shareholder value, one could ignore the transaction costs of the hedging programs. However, were it optimal for unhedged projects to be postponed or shelved, then risk management has the potential for perpetuating poor investment decisions. The glee of the manager who comprehended how cashflow hedging could free him of having to answer to his bankers was enlightening to me. In the FSS analysis, his reaction could be interpreted as that of a shareholder-value-maximizing agent, who saw a way to minimize the deadweight costs of monitoring imposed by his bankers. Seen in person, the reaction was more like that of someone who would be personally relieved not to be called upon to explain why the bank should continue to fund his business.
The essence of the critique developed here is that risk management may lead managers to continue poor projects too long. The recent experience of firms in the mining industry may prove to be an example of this phenomenon, at least according to some observers. In late 1997, the price of gold reached a historic low, and not surprisingly, a number of mines considered shutting down operations.(18) At least one industry observer worried that some high-cost, but hedged, mines would continue to produce, rather than optimally shut down and settle their financial contracts. His challenge was plain: Companies "should not hide behind derivatives. If mines aren't profitable, they should be closed."(19) Specifically, suppose that gold prices were $320 an ounce, a firm's variable production costs were $350, and it had hedged at $400 an ounce. It could mine its own gold, deliver it at $400, and make a profit. However, the firm might earn larger profits were it to close its mine, buy the gold at the market price, and use that gold to settle the hedging contracts. However, this latter decision might be less acceptable to managers of the mine, so they might prefer to continue to operate at lower, but still positive, profits. Were the firm not hedged, it would have little choice but to shut-in production.
Appreciating this possible abuse of risk management, boards of directors, risk auditors, and other overseers of risk management programs must therefore address a number of critical questions, such as the following:
Should we ban all hedging? Not at all. Don't overreact. Overseers of risk management programs must recognize and analyze the many valid benefits of risk management. If information asymmetries are large, tax schedules highly convex, financial distress very costly, and agency problems slight, the benefits of risk management are likely to overwhelm its costs. All firms face risks, and it is better to consciously choose which risks the firm will bear rather than to ignore them or simply accept them. The process of risk management allows firms to make important strategic choices, and banning this process would likely damage firms.
When should we worry? If the agency conflicts between managers and shareholders are severe, the cost of eliminating regular trips to the capital markets may be substantial enough to tip the scales against risk management. Unfortunately, while the benefits of risk management may be large, the costs may be large as well. As information asymmetries between informed insiders and outsiders grow, the costs of external financing grow and the logic of cash-flow hedging becomes stronger. Yet, as these information asymmetries become more severe, it can become harder for outside shareholders to judge whether management's decisions are in perfect alignment with their own, and the potential for agency conflicts can increase.
If agency problems are large (that is, if managers are not "trustworthy" servants of shareholder interests), then there are many ways that managers can advance their interests at the expense of shareholders. Managerial entrenchment can take the form of golden parachute contracts, poison pills, or stilted capital budgeting and reporting systems, to name a few suspects. Overseers must be vigilant to the potential misuse of all of these management decisions. This paper suggests that they may also need to be careful about how risk management programs might be used in ways contrary to shareholder interest. While outsiders might know to look closely at blatant entrenchment devices, they would not necessarily think of "innocent" risk management policy in the same way. However, overseers need to be cautious about the potential misuse of risk management.
Do derivative dealers monitor firms like other capital market providers? When a firm borrows from a bank, issues a bond, or sells stock, the provider of funds is a claimant on the firm. As claimants, they have an incentive to monitor the firm. However, in many option transactions, such as when a firm buys puts, the firm is a claimant on the derivative dealer instead. For example, if a gold mining firm buys puts on gold, it pays cash upfront, and the dealer has the liability to deliver in the future. As holders of liabilities, the dealer does not need to worry as much about the creditworthiness of the firm.
With risk management contracts that deliver linear payouts, like swaps or forward contracts, the contract could either be an asset or a liability of the firm, and one might assume that this would give rise to monitoring. Yet, even some linear risk management contracts necessitate little ongoing monitoring. By entering into a futures contract, a firm must post initial margin, and mark its position to market on a daily basis by posting additional margin. At the beginning of any day, its counterparty has no economic exposure to the firm, as its position has zero value. While the counterparty needs to worry about the legal safety of its collateral and its intra-day exposure to the firm until the next posting of margin, it need not invest a great deal in monitoring the long-term decisions of the firm.(20) Other linear contracts that are not marked to market can give rise to monitoring by risk management vendors.(21)
What kind of questions should we ask managers? Recent derivative disasters have led firms and their boards to review corporate risk management strategies carefully. For example, one recent survey of industrial corporations found that 63% of all firms had conducted or planned to conduct audits of their derivative practices (see Smithson, 1996). Much of this oversight seems directed toward ferreting out mischievous rogue traders. An earnest CFO who justifies her firm's new risk management policy as being consistent with the latest academic research, and who presents the program as an adjunct to the firm's projected investment program, will not appear to be a rogue betting on the gyrations of the market.
However, this paper reminds us that behind every well-conceived cash-flow hedging strategy is an investment strategy that boards and outside investors must understand. One of the most celebrated users of cash-flow hedging has described the firm's risk management strategy as one "that provides management with what amounts to an insurance policy against the potentially damaging effect of currency volatility on the company's ability to carry out its strategic plan."(22) At the most fundamental level, a critical question that overseers of risk management strategy must address is whether risk management policies perpetuate investment programs that the external capital markets would otherwise consider unwarranted. This would involve examining the capital budgeting processes at the firm, how projects are approved, whether interim go/no-go reviews are in place, and how the firm incorporates new information in investment decisions.
Risk management has the potential to help firms fashion the risks they are willing to bear and those risks that would be best handled by others. Like most beneficial innovations, it can sometimes be used incorrectly or in excess, producing unwelcome outcomes. This paper alerts us to one possible misuse, and challenges managers and academics to better understand when and where the costs of corporate risk management may be sizable.
1 See Smithson (1996) and Bodnar, Hayt, and Marston (1996). Berkman, Bradbury, and Magan (1997) report that 100% of all large firms in New Zealand use derivatives in the management of their finances.
2 David Shimko and Charles Smithson have each written monthly columns in Risk Magazine which discuss the benefits of hedging. For example, see Shimko's End User's Guide column (November 1995) entitled "Derivatives and the Bottom Line," or Smithson's piece in his Class Notes column, entitled "Theory and Practice" (September 1996).
3 For example, see the heated discussion of the form of risk management carried out by Metallgesellschaft by Culp and Miller (1995) and Mello and Parsons (1995).
4 This statement can be established by appealing to arguments first made by Modigliani and Miller (1958), whose work established the irrelevance of financing decisions.
5 Industry-level studies of gold mining, savings and loans, and the oil and gas industry by Tufano (1996), Schrand and Unal (1997), and Haushalter (1997), respectively, show empirical relationships between managerial compensation and corporate risk management choices. Berkman and Bradbury (1996) find only a modest relationship between management share ownership and the extent of derivatives activities at firms.
6 Guay (1997) shows that extant compensation contracts tend to provide managers with lower risk-taking incentives when the value to be created from hedging is likely to be greatest, so that the prior empirical result may be explicable as the conscious outcome of firms creating proper incentives for managers. Stulz (1996) also makes this point as a logical proposition.
7 Rock (1984, 1992) describes and analyzes the investment decisions facing Gulf Oil, and specifically the difference between outside investors' and managers' perspectives on the continuation of a massive exploration program. For a discussion of the agency problems in major industrial firms, see Jensen (1993).
8 For example, see Jensen (1986, 1993).
9 Jensen (1993) discusses the reluctance of corporate managers to downsize, illustrated by the US tire industry, and he argues that many corporate investments destroy shareholder value. His evidence suggests that AOO's investment policy may reflect actual policies at some firms.
10 Jensen (1993) discusses the difficulties of firms voluntarily disgorging cash, cutting back on investments, and exiting businesses.
11 In work done in parallel to and independent of this paper, Chang (1997) reaches a similar conclusion, although using a different model. He models the compensation contracts that induce managers to take optimal investment decisions, and the impact of hedging on these decisions. He does not explicitly model the tradeoff between the costs of external financing and opportunistic behavior by managers as is done here.
12 Under the model, if the project is not completed, the managers capture a smaller fraction of the value of the partially completed project. In managerial terms, for example, this would correspond to the managerial benefits of running a large factory being far greater than that of being responsible for a half-done and shelved project.
13 As in Hart (1993), one can think of the manager's private consumption as siphoning off the cash flow of the positive-NPV project. Unlike investments in the firm's project, investments held as riskless securities are observable and verifiable, thus the manager can not appropriate them for private use. We can think of liquid investments that are in the form of bank deposits whose balances can be audited, and thus managers are unable to consume them.
14 An upper bound on the amount of agency costs that might be tolerated in a firm may be the control premium that is typically paid in hostile control contests. In their summary of the empirical work on takeovers, Jensen and Ruback (1983) find an average run-up of 20-30% of the targets' stock upon the announcement of tender offers and other takeovers.
15 Recent empirical evidence furnished by Shin and Stulz (1996) does not support the conjecture that there is no evidence that better investments receive more of diversified firms' cash flows.
16 These anonymous examples are drawn from my discussions with managers.
17 "To Hedge or Not to Hedge," Risk Magazine, June 1997.
18 See D.G. McNeil, Jr., "Gold Prices Plunge, and a Nation Quakes," New York Times, July 10, 1997, p. D1.
19 M. Nicholls, "Reality Strikes Gold Miners," Risk Magazine, September 1997, p. 7.
20 Technically, for futures contracts, the counterparty is an exchange. However, this discussion would also apply to OTC collateralized swap contracts, in which collateral is posted equal to the market value of the swap. Usually this collateral is posted less frequently than daily, so the counterparties may have intra-month exposure, which is sometimes measured using a form of value-at-risk. For an example, see Esty, Tufano, and Headley (1994).
21 We observe monitoring in long-term forward contracts, such as spot deferred contracts in the gold mining industry or prepaid forward contracts in the oil industry. For descriptions of these contract types, see Tufano and Serbin (1993) and Perold and Singh (1995). These long-term risk management contracts have a material element of borrowing, and thus their vendors have interests similar to those of debt holders. As a result, these risk management contracts often include covenants similar to those found in bank debt.
22 Lewent and Kearney (1990); emphasis not in original.
Berkman, H. and M.E. Bradbury, 1996, "Empirical Evidence on the Corporate Use of Derivatives," Financial Management (Summer), 5-13.
Berkman, H., M.E. Bradbury, and S. Magan, 1997, "An International Comparison of Derivatives Use," Financial Management (Winter), 69-73.
Bodnar, G.M., G.S. Hayt, and R.C. Marston, 1996, "1995 Wharton Survey of Derivatives Usage by US Non-Financial Firms," Financial Management (Winter), 113-133.
Chang, C., 1997, "Does Hedging Aggravate or Alleviate Agency Problems? A Managerial Theory of Risk Management," Carlson School of Management, University of Minnesota Unpublished Manuscript.
Culp, C. and M.H. Miller, 1995, "Metallgesellschaft and the Economics of Synthetic Storage," Journal of Applied Corporate Finance (Winter), 62-76.
Donaldson, G., 1961, Corporate Debt Capacity: A Study of Corporate Debt Policy and the Determination of Corporate Debt Capacity, Boston, MA, Harvard Business School Division of Research.
Easterbrook, F.H., 1984, "Two Agency-Cost Explanations of Dividends," American Economic Review (September), 650-659.
Esty, B.C., P. Tufano, and J.C. Headley, 1994, "Bane One Corporation: Asset and Liability Management," Journal of Applied Corporate Finance (Fall), 33-51.
Froot, K.A., D.S. Scharfstein, and J.C. Stein, 1993, "Risk Management: Coordinating Corporate Investment and Financing Policies," Journal of Finance (December), 1629-1658.
Froot, K.A., D.S. Scharfstein, and J.C. Stein, 1994, "A New Approach to Risk Management," Harvard Business Review (November-December), 91-102.
Gale, D. and M. Hellwig, 1985, "Incentive Compatible Debt Contracts I: The One-Period Problem," Review of Economics Studies (October), 647-664.
Graham, J.R. and C.W. Smith, Jr., 1997, "Tax Incentives to Hedge," Fuqua School of Business, Duke University Unpublished Manuscript.
Guay, W.R., 1997, "Compensation, Convexity, and the Incentive to Manage Risk: An Empirical Analysis," University of Rochester Unpublished Manuscript.
Hart, O.D., 1993, "Theories of Optimal Capital Structure: A Managerial Discretion Perspective," in Margaret Blair, Ed., The Deal Decade, Washington, D.C., Brookings Institution, 19-43.
Hart, O.D. and J.D. Moore, 1996, "Debt, Default, and Renegotiation," Harvard University Unpublished Manuscript.
Haushalter, G.D., 1997, "The Role of Corporate Hedging: Evidence from Oil and Gas Producers," University of Oregon Unpublished Manuscript.
Jensen, M.C., 1986, "Agency Costs of Free Cash Flow, Corporate Finance and Takeovers," American Economic Review (May), 323-329.
Jensen, M.C., 1993, "Presidential Address: The Modern Industrial Revolution, Exit, and the Failure of Internal Control Systems," Journal of Finance (July), 831-880.
Jensen, M.C. and W.H. Meckling, 1976, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure," Journal of Financial Economics (October), 305-360.
Jensen, M.C. and R.S. Ruback, 1983, "The Market for Corporate Control: The Scientific Evidence," Journal of Financial Economics (April), 5-50.
Lessard, D.R., 1991, "Global Competition and Corporate Finance in the 1990s," Journal of Applied Corporate Finance (Winter), 59-72.
Lewent, J.C. and A.J. Kearney, 1990, "Identifying, Measuring, and Hedging Currency Risk at Merck," Journal of Applied Corporate Finance (Winter), 19-28.
Mello A.S. and J.E. Parsons, 1995, "Maturity Structure of a Hedge Matters: Lessons from the Metallgesellschaft Debacle," Journal of Applied Corporate Finance (Spring), 106-120.
Modigliani, F. and M.H. Miller, 1958, "The Cost of Capital, Corporation Finance, and the Theory of Investment," American Economic Review (June), 261-297.
Perold, A.F. and K.J. Singh, 1995, "Bankers Trust Global Investment Bank," Harvard Business School Case 295010.
Rock, K.F., 1984, "Gulf Oil Corp. Takeover," Harvard Business School Case 285-053.
Rock, K.F., 1992, "Gulf Oil Corp. Takeover- Teaching Note," in W. Fruhan, et el, Instructors' Manual to Accompany Case Problems in Finance (Homewood IL, Irwin).
Rozeff, M.S., 1982, "Growth, Beta and Agency Costs as Determinants of Dividend Payout Ratios," Journal of Financial Research (Fall), 249-259.
Schrand, C.M. and H. Unal, 1997, "Hedging and Coordinated Risk Management: Evidence from Mutual-to-Stock Conversions," University of Maryland Unpublished Manuscript.
Shin, H. and R.M. Stulz, 1996, "Is the Internal Capital Market Efficient?" Unpublished Manuscript.
Smith, C.W. and R.M. Stulz, 1985, "The Determinants of Firms' Hedging Policies," Journal of Financial and Quantitative Analysis (December), 391-405.
Smithson, C.W. (Ed.), 1996, Managing Financial Risk 1996 Yearbook, New York, NY, CIBC Wood Gundy.
Stulz, R.M., 1984, "Optimal Hedging Policies," Journal of Financial and Quantitative Analysis (June), 127-140.
Stulz, R.M., 1990, "Managerial Discretion and Optimal Financing Policies," Journal of Financial Economics (July), 3-27.
Stulz, R.M., 1996, "Rethinking Risk Management," Journal of Applied Corporate Finance (Fall), 8-24.
Townsend, R.M., 1979, "Optimal Contracts and Competitive Markets with Costly State Verification," Journal of Economic Theory (October), 265-293.
Tufano, P., 1996, "Who Manages Risk? An Empirical Examination of Risk Management Practices in the Gold Mining Industry," Journal of Finance (September), 1097-1137.
Tufano, P. and J. Serbin, 1993, "American Barrick Resource Corporation: Managing Gold Price Risk," Harvard Business School Case 293-128.
I would like to thank Carliss Baldwin, Ben Esty, Ken Froot, Andre Perold, Richard Ruback, Rene Stulz, Markus Mullarkey, three anonymous referees, the Editors, and the participants in the HBS Financial Decisions and Control Seminar for their comments on this work. This research was funded by the Division of Research of the Harvard Business School and was conducted as part of the Global Financial Systems project.
Peter Tufano is an Associate Professor of Business Administration at Harvard Business School.
|Printer friendly Cite/link Email Feedback|
|Date:||Mar 22, 1998|
|Previous Article:||Valuing long-term commodity assets.|
|Next Article:||Market efficiency and the returns to technical analysis.|