Journal of Derivatives Accounting, Vol. 1, No.

1 (2004) 2946 c JDA

STOCK OPTIONS AND MANAGERIAL INCENTIVES TO INVEST

TOM NOHEL University of Michigan, USA STEVEN TODD Department of Finance, School of Business Administration, Loyola University, Chicago, 820 N. Michigan Avenue Chicago, IL 60611, USA stodd@luc.edu

We examine the eect of stock options on managerial incentives to invest. Our chief innovation is a model wherein rm value and executive decisions are endogenous. Numerical solutions to our model show that managerial incentives to invest are multi-dimensional and highly sensitive to option strike prices, the managers wealth, degree of diversication, risk aversion, and career concerns. We show that under- and over-investment problems can be large and economically signicant, with hurdle rates ranging from more than 20 percentage points below to more than 35 percentage points above shareholders required rate of return. Finally, rm value is not a strictly increasing function of a managers incentive compensation or conventional pay performance sensitivity metrics. Stronger managerial incentives to invest can benet or harm a rm. Keywords : Executive stock options; incentives; investment; risk aversion; career concerns.

1. Introduction Researchers looking to examine the incentive eects of executive stock options are confronted with two immediate challenges. First, standard BlackScholes risk-neutral valuation techniques are inaccurate because executive stock options are not traded, and they are frequently held by managers with undiversied wealth (including human capital) facing short-sale and liquidity constraints. Second, option values are inextricably linked to executive actions and rm value. If options did not aect managerial behavior, and if managerial behavior did not aect rm value, then why would rms choose to grant options? The fact that options are a large and growing portion of US executive pay (Murphy, 1999), that

rms routinely re-price stock options (Brenner et al., 2000), that largely as a result of stock option pay, US executives, on average, hold a claim on more than 3% of their rms shares (Conyon and Murphy, 2000) all these facts suggest that the agency problems between managers and shareholders are large and that the benets of stock option compensation are potentially great. This paper examines the eect of stock options on managerial incentives to invest. We extend the literature by studying this eect within a model that links executive actions and rm value. This is important. Managers have direct control over their rms nancing and investment decisions. And stock options are a mechanism to alter managerial behavior so that

Corresponding author. 29

30

Tom Nohel and Steven Todd

shareholder and executive interests are aligned to maximize rm value. The value and incentive eects of executive stock options depend on the parameters of the option contract, and the managers wealth, degree of diversication, and risk aversion (Lambert et al., 1991). We construct a model of managerial investment that links rm value to managerial behavior and managerial behavior to executive-specic risk and wealth parameters. One advantage of our approach is that we are able to describe managerial incentives to invest as hurdle rates, rather than partial derivatives of the managers utility function (Hall and Murphy, 2000) or comparative statics from a BlackScholes option pricing model (Johnson and Tian, 2000). These hurdle rates allow us to quantify the degree of under- or over-investment in simple terms. In our model, a risk-averse executive sets the investment strategy of his rm based on private information he possesses about project payos. We assume the executive maximizes the expected utility of his future wealth. His wealth includes cash, and incentive compensation in the form of shares and options on his rms shares; it also includes a measure of his human capital the value of his expected future compensation, which itself is a function of the managers perceived ability. We assume the value of the managers human capital rises and falls with the value of the rm he manages. Since the manager is risk averse, he overweighs the downside risk to the value of his human capital. When the manager is reluctant to put his human capital at risk we say he has career concerns. Career concerns create an agency problem between the executive and the risk-neutral owners of the rm: the executive ignores investors risk-return preferences and maximizes his own utility (see Holmstrom and Ricart i Costa, 1986; Milgrom and Roberts, 1992; Nohel and Todd, 2003). We analyze how the executives investment choices change as the parameters of his wealth and compensation vary. Our analysis produces some interesting results. First, we nd that executive incentives to invest are highly sensitive to the distribution of managerial wealth and the magnitude of

executive career concerns. Using reasonable values for these parameters, we show that underand over-investment problems can be large and economically signicant, with hurdle rates ranging from more than 20 percentage points below to more than 35 percentage points above shareholders required rate of return. Second, we nd that incentive alignment is possible, there are a variety of ways to achieve incentive alignment, and that depending on investment risks and executive-specic risk and wealth parameters, stock options can be a small or large component of executive pay. Third, contrary to Hall and Murphy (2000), we nd that managerial incentives to invest are highly sensitive to stock option strike prices. Even when options constitute only 10% of an executives wealth, a 10% change in the option strike price can move the managers hurdle rate up or down by more than 9 percentage points. In our model, at-the-money options maximize a managers incentive to take risk. However, in many situations, premium options do a better job of aligning executive and shareholder interests, a point made in Johnson and Tian (2000). The relation between option strike prices and managerial investment incentives has important implications for the issue of stock option re-pricing. Fourth, we argue that if incentive alignment were the primary rationale for the use of stock options by corporations, we would expect to see frequent option re-pricing both when stock prices fall and when they rise. In contrast, Brenner et al. (2000) nd that nearly all re-pricing occurs after a fall in stock prices. We would also expect corporations to monitor executive wealth, possibly restricting executive share holdings, which discourage risk-taking and counteract the desired eects of stock options. Finally, we nd that the comparative statics from risk-neutral option valuation models (e.g. BlackScholes) do not provide accurate guideposts for the magnitude of managerial incentives to invest. Our results indicate that a risk-averse manager who cannot diversify the risk associated with his stock option wealth may not prefer an increase in the value of his rms stock

Stock Options and Managerial Incentives to Invest

31

or the variance of his rms stock returns. Managerial incentives to increase rm value do not necessarily rise as the managers pay becomes more sensitive to performance. Our results are relevant to studies that examine the relation between executive compensation and corporate investment decisions, such as Aggarwal and Samwick (1999) and Datta et al. (2001). Our results can also inform studies of ownership structure and rm performance, such as Demsetz and Villalonga (2001), Himmelberg et al. (1999), and Ljungqvist and Habib (2003). The remainder of this paper is organized as follows. We develop a model of managerial investment in Section 2 and solve the model numerically for a variety of parameterizations in Section 3. We discuss payperformance sensitivity and rm value in Section 4 and conclude in Section 5. 2. A Model of Managerial Investment In this section, we model the investment behavior of a risk-averse manager whose wealth consists of cash and incentive compensation in the form of shares, and call options on his rms shares. We assume the manager maximizes the expected utility of his future wealth, which includes not only cash and incentive compensation, but also a measure of his human capital the expected value of his future compensation that is based on his perceived ability. 1 The manager sets his rms investment strategy based on private information he possesses about project payos. One salient feature of our model is that rm value is a function of managerial behavior and managerial behavior is a function of executive wealth. 2.1. Set-up and information structure of the model Consider a rm with I dollars in cash on its balance sheet at t = 0. The rms investment
1

opportunity set consists of two mutually exclusive projects. Each project requires an initial investment of I . One project is riskless and thus earns the risk-free rate, which we set equal to 0 for ease of exposition. The other project is risky and generates a payo that is uniformly distributed on the interval [VL , VH ], with VH I VL . The low end of the distribution, VL , is known and certain. The high end, VH , is uncertain but known to be uniformly distributed on the interval, [VH1 , VH2 ]. Assume the relationship among I, VL , VH1 , and VH2 is such that I = (VL + (VH1 + VH2 )/2)/2.2 Our model has three types of participants: investors, a manager, and an intermediary. Investors put their capital at risk in the rm. They select an intermediary to hire a manager and design the managers compensation. Think of this intermediary as either a compensation consultant or the rms board of directors. For ease of exposition, we refer to the intermediary as the board of directors, or simply the board. We assume that the board acts in the interest of shareholders and is otherwise passive. The manager makes the rms investment decisions. The timing in our model is as follows: at t = 0, investors know that the rm has I in cash on its balance sheet. Furthermore, at t = 0, investors are aware of the rms investment opportunities (including the distribution of project payos) and they elect a board of directors to hire a manager of unknown talent. The board xes a compensation budget for the manager and sets the terms of his compensation. At t = 0+, the board hires a manager and pays him. His pay consists of a cash salary, and incentive compensation in the form of shares and options on his rms shares. At t = 1, the manager receives a signal that identies VH . Based on this signal, the manager chooses between the rms two investment opportunities. The signal is never revealed and thus cannot be contracted on. At t = 2, the outcome of the investment decision is realized, the managers shares and options vest, and the

The compensation a manager can expect to earn over his career depends on his reputation. In our model, reputation is not something that the manager can cultivate, say, by working harder. 2 That is, ex-ante, both the riskless and the risky projects have zero NPV. This assumption is consistent with strong-form market eciency. However, as long as VH1 I VL , most of our results hold.

32

Tom Nohel and Steven Todd

rm is liquidated.3 Below is a time line specifying critical events in our model:

Investor select board and set comp. budget Designer hires mgr., reveals contract & inv. opps Manager gets private signal and invests Project payoffs determined, shares & options vest, and the firm is liquidated

|-------------------------------|----------------------------------|----------------------------------| t=0 t=0+ t=1 t=2

Investors and the managers dier in their preferences towards risk. We assume investors are risk neutral and the manager is risk averse in wealth. This assumption is consistent with Hemmer et al. (1996) who nd that executives exercise their stock options early. We further assume the managers utility function exhibits constant relative risk aversion as described below. 4 1 U (w) = w(1b) , b 0, b = 1 (1) 1b Here, w is the level of the managers wealth that he derives from the following sources: cash (including salary and non-rm wealth), incentive compensation (shares and options on his rms shares), and the change in the value of the managers human capital.5 Let K denote the value of the managers cash wealth. The managers incentive compensation includes n shares and m options with an exercise price of X . Here, n and m represent fractions of the shares outstanding in the managers rm. We restrict K , n, and m to be non-negative and we assume shares and options vest at t = 2. 6 We assume that the managers performance aects his expected future income, i.e. the value of his human capital. Since the manager is risk

averse, he overweighs the downside risk to the value of his human capital. These career concerns create an agency problem between the manager and the risk-neutral owners of the rm he manages (see also: Holmstrom and Ricart i Costa, 1986; Milgrom and Roberts, 1992; Nohel and Todd, 2003).7 Fee and Hadlock (2001) show that stock returns are used as a performance metric by the external labor market for top-ranked executives, and Jagannathan and Wang (1996) show that stock returns are linearly related to the return on human capital. Consistent with these results, we assume the value of the managers human capital is linear in rm value.8 Specically, we assume the change in the value of the managers human capital is proportional to the change in rm value from the time when the manager is hired (t = 0+), to the time when the rm is liquidated (t = 2). If is a non-negative constant, V is the rms liquidation value, and V 0+ is the value of the rm immediately after the manager is hired, then the change in the value of the managers human capital is (V V0+ ). Note that with this specication, the expected change in the value of the managers human capital is zero. Intuitively, we interpret the change in the value of the managers human capital as the present value of all future compensation gains (losses) that are attributable to the managers performance. Investors are rational so the rms value changes as new information is released. Starting at an initial value of I (at time 0), the rms value adjusts sequentially to reect information

3 The distribution of project outcomes and the information structure of our model are similar to Ross (1977). The binary nature of the investment choice is common in the literature. See, for example, John and John (1993), Lambert (1986), and Holmstrom and Ricart i Costa (1986). 4 The conclusions of our paper are valid with any utility function that exhibits decreasing absolute risk aversion. We chose the specication above since it is common in the literature; see Lambert et al. (1991) and Hall and Murphy (2000). 5 In the interests of simplicity, we model the managers risky security holdings as cash equivalents. For a similar treatment, see Lambert et al. (1991) and Hall and Murphy (2000). 6 In the interests of simplicity, we sidestep the more general problem of an executive who has a portfolio of options with varying strike prices. 7 We side-step the common assumption of managerial dis-utility of eort. The notion that a manager can positively aect output by working harder but prefers not to is inconsistent with the behavior of top-level executives. Kaplan (1984) points out that contemporary managers often work too hard rather than too little. It is hard to conceive of managers shirking, given continual monitoring by the nancial press and securities analysts. Allen (2001) suggests that instead of modeling eort aversion, research in this area should focus on implicit contracts and reputation. He considers the nancial industry, but his remarks apply to any situation where, as he says, the principal does not have the expertise that the agent does. 8 This mirrors Graham (1999) who assumes that analysts wages increase linearly with their reputations.

Stock Options and Managerial Incentives to Invest

33

about the hiring of the manager (time 0+), the managers investment decision (time 1), and the nal project outcome (time 2). The price adjustment from time 0 to time 0+ reects investors valuation of the marginal benet/cost of hiring the manager, and is consistent with investor reactions to awards of incentive pay, as documented in Defusco et al. (1990) and Yermack (1997). Note that since the rms value is I just prior to the manager being hired, options with strike prices above (below) I are out-of-themoney (in-the-money). The stock price adjusts to V0+ only after the hiring of the manager is made public. Due to the binary nature of the investment choice in our model, an investment strategy is H , where dened as a cut-o point. Consider any V H VH1 . [V H ] denotes the strategy VH2 V where the manager invests in the risky project H ; othwhenever he observes a signal VH V erwise, he invests in the riskless project. Thus, H ], if the manager follows investment strategy [ V he chooses the risky project with probability H )/(VH2 VH1 ), and he chooses the risk(VH2 V H )/(VH2 less project with probability 1(V H2 V H VH1 )/(VH2 VH1 ). By denition, the VH1 ) = (V riskless project has zero NPV. In contrast, the opportunity to take the risky project has NPV H )/2 + VL ]/2 I .9 equal to [(VH2 + V Therefore, if the manager follows investment H ], the value of the rm after the manstrategy [V ager is hired, V0+ , as a function of the managers H ], is given by investment strategy, [V H ) = I + (VH2 VH ) V0+ (V (VH2 VH1 ) H )/2 + VL (VH2 + V I 2

return that exceeds the risk-free rate of zero. This implies that their preferred investment strategy H = 2I VL . is to set V The risk-averse manager with career concerns ignores investors preferences on investment strategy and maximizes his own utility. Herein lies the agency problem. Compensate the manager in cash and he is too conservative, because a risky project entails human capital risk without sucient reward in some states. Substituting shares for cash makes matters worse. In fact, in our model, the change in the value of a managers human capital is equivalent to the change in the value of a long forward contract on the managers rms shares. When compensation is linear in rm value, the combination of the managers risk aversion and career concerns compels the manager to follow an overly conservative investment H ], where V H > 2I VL . The board strategy [V of directors is faced with the problem of nding ways to overcome the managers excessive conservatism by incorporating compensation elements whose values are convex in rm value. 10 2.2. Solution to the model The manager develops his investment strategy H ] by solving for the value of the signal that [V makes him indierent between the two projects. Given wealth and compensation (K, n, m, X ) and a signal VH , the manager can invest in the risky project and derive expected utility equal to E [U (K, n, m, X, VH )] 1 1 = 1 b (VH VL )
Min(X,VH )

(2)

VL VH

(nV + K + (V V0+ ))(1b) dV (nV + m(V X ) + K

Equation (2) ignores the cost of the managers compensation. Risk-neutral shareholders strictly prefer the risky project as long as it oers an expected
9

+
Min(X,VH )

+(V V0+ ))(1b) dV

(3)

Note, this diers from the ex-ante NPV because, here, it is known that the manager gets a signal about V H . One possible solution to this problem is for shareholders to insure the managers human capital with a bonusthat is equivalent to a short forward contract on the rms shares. However, such a scheme is not incentive compatible because it asks the manager to return money to his rm when the value of his human capital increases. Since the managers performance is observable, the manager can simply quit his current rm and capture the increased value of his human capital at another rm.
10

34

Tom Nohel and Steven Todd

In this expression, (V V0+ ) measures the change in the value of the managers human capital.11 Alternatively, the manager can invest in the riskless project and derive (expected) utility equal to U (K, n, m, X, VH ) 1 = [nI + m Max(0, I X ) 1b + (I V0+ ) + K ]1b

3. Wealth, Incentive Compensation, and Managerial Investment In this section, we report solutions to Eq. (5). We are interested in comparing the incentive eects of various divisions of wealth and incentive compensation (K, n, m, X ). We follow Johnson and Tian (2000) and Hall and Murphy (2000) and compare schemes of equal cost from the perspective of risk-neutral shareholders. 12 This approach is broadly consistent with Bebchuk et al. (2002) who argue that executives have power to inuence their own compensation. As an alternative, an optimal contracting approach presumes shareholder value is maximized, with managers being guaranteed a reservation utility. Such an approach is considered in Nohel and Todd (2003). We set the managers coecient of relative risk aversion, b = 3, consistent with studies by Friend and Blume (1975), Litzenberger and Ronn (1986), Lambert et al. (1991) and Hall and Murphy (2000). The change in the value of the managers human capital is given by (V V 0+ ). Since we are comparing compensation schemes of equal cost, and since the managers utility function exhibits constant relative risk aversion, only the ratio of to wealth matters. We consider a range of ratios, /W , from 0% to 0.6%. With this range of ratios, career concerns run the gamut from slight to extreme.13 We set X = I (rm value at time 0) initially, consistent with the near uniform practice of granting options at-the-money (Murphy, 1999).14 We consider three risky projects, each requiring an initial investment of I = $100: a lowrisk project with (VL , VH1 , VH2 ) = (75, 100, 150);

(4)

Here, the term (1 V0+ ) measures the change in the value of the managers human capital. H ], The manager sets his investment rule [ V by solving for the value of VH that equates (3) and (4); this is the value of VH that solves (5): 1 1b
VL VH

1 (VH VL )
Min(X,VH )

(nV + K + (V V0+ ))(1b) dV (nV + m(V X ) + K

+
Min(X,VH )

+(V V0+ ))(1b) dV = 1 [nI + mMax(0, I X ) 1b +(I V0+ ) + K ]1b

(5)

denote the value of V that solves (5). If Let VH H the manager acts in his own interest, he follows ]. investment strategy [VH
11

We set the integral limits equal to Min(X, VH ) rather than X . When the option strike price is greater than the value of the managers signal, the options are worthless and the manager makes his investment decision based only on the value of his cash and shares. 12 In determining the cost of a compensation plan, we assume that the board values shares and options under shareH = 2I VL . Therefore, we compute n using the formula for rm value given holders preferred investment strategy, V in (2), with VH set equal to 2I VL . We determine m in an analogous way using the call option formula derived in Appendix A. For a similar treatment see Gavish and Kalay (1983). 13 Other things being equal, we would expect younger (or more recently tenured) managers with less accumulated wealth to have higher career concerns (see Gibbons and Murphy, 1992). In our numerical solutions below, we consider a range of risky project cash ows. For the median CEO with eective shareholdings equal to 1.48% of his rms shares (Conyon and Murphy, 2000), the range of career concerns we consider covers a gain/loss in human capital equal to +37% to 34% of total share wealth. For a CEO with smaller (larger) eective shareholdings or other wealth, the potential gain/loss in human capital covered by this range of career concerns can be considerably larger (smaller). 14 Again, note that until the manager is hired, the rm value is I . Thus, striking the managers options at-the-money means setting X = I .

Stock Options and Managerial Incentives to Invest

35

a medium risk project with (VL , VH1 , VH2 ) = (50, 100, 200); and a high-risk project with (VL , VH1 , VH2 ) = (25, 100, 250). As a percentage of invested capital, these projects have volatilities equal to 14.4%, 28.9%, and 43.3%, respectively, and net present values equal to 3.125%, 6.25%, and 9.375%, respectively. We believe these projects are representative of the projects available to typical US rms. As Dimson and Marsh (2001) show, for the period 19551999, the geometric equity risk premium for US rms is 6.2%. Moreover, Ibbotson and Sinqueeld (2001) report that over the period 19262000, stock returns had an average standard deviation of 28%, with rms in the largest decile averaging 19.1% and rms in the smallest decile averaging 45.8%. Tables 13 summarize managerial investment strategies for various divisions of wealth and incentive compensation and for a range of

career concerns. To facilitate comparative analysis, we convert the managers investment strategy ] into a hurdle rate. In all tables, sharehold[VH ers preferred hurdle rate is 0.0%. Table 1 reports managerial hurdle rates for the low risk project; Tables 2 and 3 report managerial hurdle rates for the medium and high risk projects, respectively. We set bold numeric values in the tables where over-investment occurs. We nd that managerial incentives to invest are highly sensitive to the distribution of wealth and the magnitude of executive career concerns. For the medium-risk project (Table 2), if we x the managers wealth at 50% cash and 50% options, the range of hurdle rates is 25.00% (when the ratio of to wealth equals 0%) to +25.00% (when the magnitude of career concerns equals 0.6%). Paying the manager entirely in cash works well when career concerns are zero, but leads

Table 1. Managerial investment strategy (hurdle rates) for the low-risk project.
% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 Career concerns, /Wealth 0% 2.86 6.79 9.27 12.50 1.75 6.06 8.19 9.55 10.73 12.50 0.81 6.71 8.71 9.93 10.94 12.50 0.00 0.2% 4.01 4.93 6.89 8.16 2.70 4.23 6.33 7.59 8.55 9.42 1.61 4.29 6.30 7.49 8.35 9.04 0.67 0.4% 5.46 2.97 4.15 2.51 3.86 2.55 4.59 5.76 6.57 7.21 2.55 2.53 4.51 5.70 6.55 7.21 1.46 0.6% 7.37 0.68 0.46 12.50 5.31 0.82 2.78 3.80 4.39 4.70 3.70 0.93 2.87 4.05 4.88 5.51 2.39

Table 1 reports hurdle rates implied by the managers preferred investment strategy. Under shareholders preferred investment strategy, the manager sets the hurdle rate at 0.00%. We assume the risky project cash ows are uniformly distributed on the interval (75, VH ) with VH uniformly distributed on the interval (100, 150). We set the managers coecient of relative risk aversion equal to 3.0; measures the magnitude of the managers career concerns. We set bold numeric values in the table where over-investment occurs.

36

Tom Nohel and Steven Todd

Table 2. Managerial investment strategy (hurdle rates) for the medium-risk project.
% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 Career concerns, /Wealth 0% 14.63 4.81 13.40 25.00 8.00 5.79 11.63 15.81 19.59 25.00 3.39 9.13 14.19 17.60 20.51 25.00 0 0.2% 25.00 6.46 0.41 5.03 14.41 1.50 4.16 8.05 11.32 14.43 7.54 2.15 6.82 9.95 12.38 14.43 2.91 0.4% 25.00 25.00 25.00 25.00 25.00 11.90 6.03 2.41 0.33 2.72 14.29 4.81 0.11 3.05 5.46 7.46 7.08 0.6% 25.00 25.00 25.00 25.00 25.00 25.00 25.00 25.00 25.00 25.00 25.00 15.97 10.03 6.44 3.86 1.82 14.32

Table 2 reports hurdle rates implied by the managers preferred investment strategy. Under shareholders preferred investment strategy, the manager sets the hurdle rate at 0.00%. We assume the risky project cash ows are uniformly distributed on the interval (50, VH ) with VH uniformly distributed on the interval (100, 200). We set the managers coecient of relative risk aversion equal to 3.0; measures the magnitude of the managers career concerns. We set bold numeric values in the table where over-investment occurs.

to an ever increasing under-investment problem as the ratio of to wealth increases. The under-investment problem is exacerbated when shares are substituted for cash. For example, in Table 2, if we x the magnitude of career concerns at 0.4%, the managers hurdle rate rises from 7.08% (when his wealth consists of 100% cash) to 14.29% (when his wealth consists of 75% cash and 25% shares). Options, with their convex payos, can be used to overcome the managers reluctance to invest. For example, in Table 2, keeping the magnitude of career concerns xed at 0.4%, the manager lowers his hurdle rate to +0.11% when his wealth consists of 75% cash, 15% shares, and 10% options. With this mix of wealth, the managers interests are nearly aligned with those of the shareholders. With too much option wealth, the manager over-invests. In Table 2, too much option wealth can be as little as 5% or 10% of wealth.

For example, if we x the magnitude of career concerns at 0.2%, the manager sets his hurdle rate at 6.82% when his wealth consists of 75% cash, 15% shares and 10% options. Tables 13 show that problems of either over-investment or under-investment can be large and economically signicant. There are a variety of ways to achieve incentive alignment between managers and shareholders. Depending on investment risks and executive-specic risk and wealth parameters, stock options can be a small or large component of executive pay. Kole (1997) and Murphy (1999) nd considerable variation in compensation awards across industries. Tables 13 show that incentive alignment is not inconsistent with this result. Two other features of Tables 13 are worth noting. First, increasing the share of option wealth generally leads to increased risk taking on the part of the manager. However, when

Stock Options and Managerial Incentives to Invest

37

Table 3. Managerial investment strategy (hurdle rates) for the high-risk project.
% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 Career concerns, /Wealth 0% 37.50 11.10 10.68 37.50 20.93 0.59 10.52 19.11 26.95 37.50 7.96 8.93 17.53 23.74 29.20 37.50 0.00 0.2% 37.50 37.50 37.50 37.50 37.50 25.28 11.93 2.20 6.19 14.09 20.94 6.35 1.83 7.71 12.45 16.54 7.17 0.4% 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 22.50 14.06 7.84 2.73 21.48 0.6% 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50 37.50

Table 3 reports hurdle rates implied by the managers preferred investment strategy. Under shareholders preferred investment strategy, the manager sets the hurdle rate at 0.00%. We assume the risky project cash ows are uniformly distributed on the interval (25, VH ) with VH uniformly distributed on the interval (100, 250). We set the managers coecient of relative risk aversion equal to 3.0; measures the magnitude of the managers career concerns. We set bold numeric values in the table where over-investment occurs.

options comprise a large portion of compensation, the reverse eect is possible (see, for example, Table 1, when the ratio of to wealth equals 0.6% and cash wealth equals 25%). Second, xing the managers mix of wealth and increasing a projects risk leads to decreased risk taking on the part of the manager, unless the magnitude of career concerns is small. An increase in project risk has two conicting eects on an options value to the manager: a positive eect from the convexity of the options payo (wealth eect), and a negative eect from the concavity of the managers utility function (risk-aversion eect). In Tables 13, the wealth eect dominates when the ratio of to wealth is less than or equal to 0.2% and the risk-aversion eect dominates when this ratio is greater than 0.2%. Several studies (e.g. Guay, 1999; Rajgopal and Shevlin, 2002) nd that options are used to encourage risk taking, and more option pay is associated with greater risk taking. These studies

implicitly assume the extra risk taking is benecial to shareholders. Our analysis indicates it is quite conceivable that rms overshoot and encourage excessive risk taking. In summary, Tables 13 show that managerial incentives to take risk are multi-dimensional. Several factors have signicant inuence over managerial incentives to take risks. Moreover, the BlackScholes metrics that are so often used as guideposts to understand option-based incentives do not apply once rm value is endogenous and managerial risk aversion is factored in. 3.1. Option strike prices and managerial investment We next consider the eect of option strike prices on managerial incentives to invest. Table 4 xes the magnitude of career concerns at 0.2% and reports hurdle rates for the medium risk project. We consider a range of strike prices from $70 to $150, with $100 being at-the-money.

38

Tom Nohel and Steven Todd

Table 4. Managerial investment strategy (hurdle rates) for the medium risk project for various option strike prices.
% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 70 25.00 25.00 25.00 25.00 14.41 16.79 20.19 25.00 25.00 25.00 7.54 7.90 8.42 9.11 9.96 10.98 2.91 90 25.00 25.00 25.00 25.00 14.41 11.63 11.59 13.43 17.34 24.49 7.54 4.53 2.93 2.08 1.72 1.71 2.91 Option strike price 100 25.00 6.46 0.41 5.03 14.41 1.50 4.16 8.05 11.32 14.43 7.54 2.15 6.82 9.95 12.38 14.43 2.91 110 25.00 8.86 3.91 1.06 14.41 3.03 1.72 5.05 7.91 10.71 7.54 0.84 4.71 7.31 9.34 11.08 2.91 130 25.00 12.98 9.92 6.00 14.41 5.91 2.73 0.42 1.67 3.82 7.54 1.85 0.56 2.20 3.50 4.64 2.91 150 25.00 16.68 15.15 12.57 14.41 8.85 6.99 5.54 4.13 2.69 7.54 4.88 3.76 2.96 2.31 1.74 2.91

Table 4 reports hurdle rates implied by the managers preferred investment strategy for a range of option strike prices. Under shareholders preferred investment strategy, the manager sets the hurdle rate at 0.00%. We assume the risky project cash ows are uniformly distributed on the interval (50, VH ) with VH uniformly distributed on the interval (100, 200). We set the managers coecient of relative risk aversion equal to 3.0. We x the ratio of , the magnitude of the managers career concerns, to wealth at 0.2%. We set bold numeric values in the table where over-investment occurs.

As in Tables 13, we follow Johnson and Tian (2000) and Hall and Murphy (2000) and compare schemes of equal cost from the perspective of risk-neutral shareholders. Managerial incentives to invest are highly sensitive to option strike prices. Within Table 4, hurdle rates vary from 14.43% to +25.00%. Fixing option wealth at only 10% (with shares at 15% and cash at 75%), the range of hurdle rates is 6.82% to +8.42%. Compared to options granted at-the-money, a 10% drop in the option strike price moves the managers hurdle rate up more than 9 percentage points. These results conict with Hall and Murphy (2000) who nd little variation in the incentives provided by stock options over a wide range of strike prices. One reason our results dier is because we link managerial behavior to rm value, whereas Hall and Murphy (2000) do not.15
15

We illustrate part of Table 4 in Fig. 1. Here we look at managerial investment when the ratio of to wealth is xed at 0.2% and 50% of the managers wealth is in cash. The gure displays hurdle rates for a range of option strike prices for dierent mixes of share and option wealth. Two features of the gure are worth noting. First, at-the-money options (X = I ) induce the most risk taking, although premium options may do a better job of aligning executive and shareholder interests. Intuitively, at-the-money options only have value under the risky project. Moreover, of those options that are worthless under the riskless project, at-the-money options are the most valuable. Second, risk-taking behavior is not symmetric around X = I . When X > I , the managers response to higher exercise prices is fairly muted; gradually, he increases his hurdle rate

This result is not driven by our modeling of the agency problem. Qualitatively similar results obtain if we ignore agency costs entirely (i.e. set = 0).

Stock Options and Managerial Incentives to Invest

39

25 20 15 Hurdle rate (%) 10 5 0 5 10 15 20 50 70 90 110 40 130 150

ns ( %)

Option str

ike price

Fig. 1. Managerial investment as a function of mix of wealth and option strike prices: /Wealth = 0.2%; cash = 50%; shares = (50 options)%.

Our analysis also indicates that the distribution of a managers wealth has a large impact on his incentives to invest. Thus, if incentive alignment were the primary rationale for stock option awards, we would expect to see companies monitor executive wealth and possibly place restrictions on share holdings, which tend to encourage risk avoidance and counteract the desired eects of stock options. We have no evidence of either practice, though some rms mandate minimal shareholdings for executives (see Core and Larcker, 2001). In contrast, Ofek and Yermack (2000) nd evidence that executives prefer to sell shares whenever possible. 4. Risk-Taking Incentives and Firm Value In this section we examine risk-taking incentives and rm value. Table 5 reports rm value under the managers investment strategy for various divisions of wealth and incentive compensation and for a range of career concerns. We present results for the medium-risk project only, and we compute rm value using Eq. (2). Equation (2) ignores the cost of the managers compensation, but this is not a problem because we are comparing rm values derived from compensation schemes of equal cost. Firm value is highly sensitive to the distribution of wealth and the magnitude of executive career concerns. If we x the magnitude of career concerns at 0.4%, rm value varies from $106.25 (when wealth consists of 50% cash, 10% shares, and 40% options) to $100.00 (when wealth consists of 25% cash and any mix of shares and options). If we x the managers wealth at 50% cash and 50% shares, rm value varies from $105.61 (when career concerns are nil) to $100.00 (when the magnitude of career concerns is 0.6%). Contrary to the implications of risk-neutral option valuation models, rm value is not an increasing function of a managers option wealth

(i.e. accepts fewer risky projects). However, when X < I , the managers investment behavior responds dramatically to changes in the options exercise price. The intuition is as follows: as X increases from I , the options are still worthless under the risk-free project, and they are incrementally worth less under the risky project. In contrast, as X falls below I , the options suddenly have value under the risk-free project, thus inducing more conservative behavior. The relation between managerial investment and option strike prices bears directly on the issue of re-pricing options. Our analysis indicates that if incentive alignment were the primary rationale for executive stock option awards, we would expect to see option re-pricing both when stock prices fall and when they rise. 16 In fact, since a managers investment behavior responds so dramatically to a rise in an options intrinsic value, we would expect to see more frequent option re-pricing when stock prices rise (and options resemble shares). Quite the reverse, Brenner et al. (2000) nd that more than 99% of all re-pricing events are associated with a reduction in the option strike price, following a decline in stock prices.
16

In drawing a parallel between the results of Table 5 and the re-pricing phenomenon, we are assuming that options are re-priced at constant cost. Hall and Murphy (2002) call this a BlackScholes re-pricing and cite General Dynamics as one of the rst companies to adopt such a practice in 1991. Brenner et al. (2000) nd that from an ex-ante standpoint, re-pricing produces a small benet to executives, generally less than 15% of the initial option award value. Callaghan et al. (2000) obtain similar estimates of the benet executives derive from option re-pricing events.

Opt io

40

Tom Nohel and Steven Todd

Table 5. Firm value, given the managers investment strategy.

% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 Career concerns, /Wealth 0% 104.11 106.02 104.45 100.00 105.61 105.92 104.90 103.75 102.41 100.00 106.14 105.42 104.24 103.15 102.04 100.00 106.25 0.2% 100.00 105.83 106.25 106.00 104.17 106.23 106.08 105.60 104.97 104.17 105.68 106.20 105.78 105.26 104.72 104.17 106.17 0.4% 100.00 100.00 100.00 100.00 100.00 104.83 105.89 106.19 106.25 106.18 104.21 106.02 106.25 106.16 105.95 105.69 105.75 0.6% 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 100.00 103.70 105.24 105.84 106.10 106.22 104.20

Table 5 reports rm value [based on Eq. (2) in the text] under the managers preferred investment strategy. At time 0, rm value is $100.00. We assume the risky project cash ows are uniformly distributed on the interval (50, VH ) with VH uniformly distributed on the interval (100, 200). We set the managers coecient of relative risk aversion equal to 3.0; measures the magnitude of the managers career concerns.

or the delta of his portfolio. For example, with the ratio of to wealth xed at 0.2%, rm value is $105.68 when the managers wealth consists of 75% cash and 25% shares. Firm value rises to $106.20 when the managers wealth consists of 75% cash, 20% shares, and 5% options, but falls to $104.17 when the managers wealth consists of 75% cash and 25% options. Stronger managerial incentives to invest can benet or harm a rm, depending on the rms investment opportunities and executive specic risk and wealth parameters.17 We illustrate sections of Table 5 in Figs. 24, where we x one component of wealth and show rm value as a function of career concerns and the mix of other wealth. In Fig. 2, cash represents 50% of managerial wealth. Here, we see that when career concerns are low, substituting shares
17

107 106 105 104 103 102 101 100

Firm Value ($)

10

20 30 Shares (%) 0.20%

40

50

alpha / Wealth = 0%

0.40%

0.60%

Fig. 2. Firm value as a function of career concerns and mix of wealth: cash = 50%; options = (50 shares)%.

for options results in an increase in rm value. In Fig. 3, shares represent 50% of managerial wealth. Here, we see that once options constitute about 10% of wealth, increasing option wealth by substituting options for cash results in a decrease in rm value. In Fig. 4, options represent 25% of managerial wealth. Here, we see that when career

Managerial risk aversion is not driving this result. If we set the magnitude of career concerns equal to zero and the managers coecient of relative risk aversion, b, equal to zero (i.e. if the manager is risk neutral), this result still holds. Firm value is not strictly increasing in managerial option wealth as long as rm value is endogenous. In contrast, risk-neutral valuation models treat rm value as exogenous and independent of managerial decisions.

Stock Options and Managerial Incentives to Invest

107 106 105 104 103 102 101 100 0 10 20 30 40 50 Options (%) alpha / Wealth = 0% 0.20% 0.40% 0.60%

41

Fig. 3. Firm value as a function of career concerns and mix of wealth: shares = 50%; cash = (50 options)%.

107 106 105 104 103 102 101 100 0 25 Cash (%) alpha / Wealth = 0% 0.20% 0.40% 0.60% 50 75

Fig. 4. Firm value as a function of career concerns and mix of wealth: options = 25%; shares = (75 cash)%.

concerns are high, substituting cash for shares results in an increase in rm value. In summary, Figs. 24 conrm that rm value is highly sensitive to the distribution of wealth and the magnitude of executive career concerns. 4.1. Payperformance sensitivity and rm value It is possible to assess the extent to which an executive participates in wealth creation by estimating payperformance sensitivity (PPS) measures. Jensen and Murphy (1990) examine PPS and estimate that the average executive experiences a wealth increase of $3.25 for every $1000 increase in his rms value. Such a low PPS challenges the principal agent paradigm, which predicts, in the case of risk-neutral managers, a one-for-one relation
18

between executive and shareholder gains. Low pay sensitivities, however, may provide adequate incentives if managers are risk-averse (Haubrich, 1994), or if rms are highly levered (John and John, 1993). Consistent with the principalagent paradigm, payperformance measures decrease, and salaries increase, as rm and CEO incomes become more volatile (Garen, 1994). Stock options, used only modestly in the time period that Jensen and Murphy study, bolster PPS measures (Hall and Liebman, 1998) and provide powerful incentive eects (Guay, 1999). Following Jensen and Murphy (1990), we dene PPS as the change in managerial compensation associated with a $1000 change in rm value, for all possible liquidation values of the rm. We describe the computation method for this measure in Appendix B. We also compute a PPS measure based on the BlackScholes option pricing model, as described in Guay (1999). 18 The former measure allows for variation in executive career concerns; the latter ignores career concerns. Table 6 reports payperformance measures for various divisions of wealth and incentive compensation and for a range of career concerns. Column 1 uses the PPS measure described by Guay (1999); the remaining columns use the PPS measure outlined in Appendix B. We see that xing the managers career concerns, there is a positive relation between option wealth and PPS. If we x the managers mix of wealth, PPS declines as career concerns increase. Figures 5 and 6 illustrate the relation between rm value (again, without netting out the cost of the managers compensation) and PPS. Both gures are based on the PPS measure outlined in Appendix B. In Fig. 5, we x cash wealth at 75%; in Fig. 6 we x cash wealth at 50%. The gures show that xing the managers career concerns, there is no simple monotonic relation between PPS and rm value. In fact, in a simple regression of the rm value observations from Table 5 on the PPS measures in Table 6,

We assume stock prices follow a lognormal process with expected return and volatility parameters equal to 6.25% and 28.9%, respectively (same as the medium-risk project). We assume options are granted at-the-money and expire in 2 years, with the risk-free rate equal to 0%.

Firm Value ($)

Firm Value ($)

42

Tom Nohel and Steven Todd

Table 6. Payperformance sensitivity under managers investment strategy.

% Share of managerial wealth as Cash 25 25 25 25 50 50 50 50 50 50 75 75 75 75 75 75 100 Shares 75 50 25 0 50 40 30 20 10 0 25 20 15 10 5 0 0 Options 0 25 50 75 0 10 20 30 40 50 0 5 10 15 20 25 0 Career concerns, /Wealth 0%** $7.06 17.54 28.02 38.50 4.71 8.90 13.09 17.28 21.47 25.67 2.35 4.45 6.55 8.64 10.74 12.83 0 0% $7.06 12.40 21.08 33.13 4.71 6.92 10.06 13.61 17.51 22.04 2.35 3.60 5.21 6.98 8.86 11.04 0 0.2% $7.06 9.78 15.41 23.21 4.71 6.28 8.86 11.91 15.32 19.08 2.35 3.30 4.66 6.18 7.82 9.54 0 0.4% $7.06 4.71 2.35 0 4.71 5.23 6.97 9.16 11.66 14.48 2.35 2.98 4.05 5.33 6.73 8.24 0 0.6% $7.06 4.71 2.35 0 4.71 3.76 2.82 1.88 0.94 0 2.35 2.40 3.08 3.99 5.04 6.20 0

Table 6 reports PPS measures (i.e. the change in the value of the managers compensation per $1000 change in shareholder wealth) under the managers preferred investment strategy. We assume the risky project cash ows are uniformly distributed on the interval (50, VH ) with VH uniformly distributed on the interval (100, 200). We set the managers coecient of relative risk aversion equal to 3.0; measures the magnitude of the managers career concerns. See Appendix B for a description of PPS. *These PPS measures are computed from the BlackScholes option pricing model, as described in Guay (1999).
107 106 105 104 103 102 101 100

Firm Value ($)

the coecient on the PPS measure is not statistically signicant.19

4.2. Robustness
0 2 4 6 8 10 12 PPS ($ change in manager's wealth per $1,000 change in shareholder wealth) alpha / Wealth = 0% 0.20% 0.40% 0.60%

Fig. 5. Firm value and PPS: cash = 75%.

107 106 105 104 103 102 101 100 0 Firm Value ($)

10

15

20

25

PPS ($ change in manager's wealth per $1,000 change in shareholder wealth) alpha / Wealth = 0% 0.20% 0.40% 0.60%

Fig. 6. Firm value and PPS: cash = 50%.

The results we present in Tables 16 and Figs. 16 are qualitatively robust to changes in the parameters of our model, including project cash ows, -to-wealth ratios, risk-aversion coefcients and risk-free rates. The conclusions of our paper are valid with any managerial utility function that exhibits decreasing absolute risk aversion. Our results do not depend on our assumption that the managers human capital is linear in rm value. Moreover, our results do not depend on our assumption that managerial career concerns are responsible for the

19 We do not mean to suggest that the rows of Tables 5 and 6 represent a reasonable division of managerial wealth at a cross-section of rms.

Stock Options and Managerial Incentives to Invest

43

agency problem between managers and shareholders. Similar results obtain as long as the managers indirect utility function is concave in rm value. 5. Conclusions This paper examines the eect of stock options on managerial incentives to invest. We develop a model in which rm value is endogenous: a manager aects rm value through his investment decisions, which depend on the managers risk and wealth parameters. We then solve the model numerically for a variety of parameterizations that are consistent with the marketplace. We nd that managerial incentives to invest are multi-dimensional and highly sensitive to option strike prices, the managers wealth, degree of diversication, risk aversion, and career concerns. Under- and over-investment problems can be large and economically signicant, with hurdle rates ranging from more than 20 percentage points below to more than 35 percentage points above shareholders required rate of return. There is no simple monotonic relation between PPS and rm value. Firm value is not a strictly increasing function of a managers incentive compensation. Stronger managerial incentives to invest can benet or harm a rm, depending on the rms investment opportunities and executive specic risk and wealth parameters. Moreover, increasing the amount of incentive pay does not necessarily increase the incentive to take risks, and may even decrease it. Our results should prove useful to researchers who study managerial behavior, specically the impact of managerial compensation and/or ownership structure on rm value. Acknowledgments We are indebted to Lu Hong for many helpful discussions and suggestions, as well as comments from seminar participants at the University of Pittsburgh. Our research has beneted from comments by discussants and participants at the following 2002 meetings: the Royal Economic Society (Warwick), the European Financial Management Association (London), the

APFA/PACAP/FMA (Tokyo) and the Financial Management Association European conference (Copenhagen). A previous version of this paper was titled Executive compensation, reputation and risk-taking incentives. Appendix A: Valuing a Call Option Based on the Investment H ] Strategy, [V Here we derive a formula for calculating the value of a call option, assuming the manager purH ]. If the manager sues the investment strategy [V takes the riskless project, the nal stock price (at t = 2) is known with certainty. A call option on the entire rm with exercise price X , expiring at t = 2, is worth Max(0, I X ). When the manager takes the risky project, the value of a call option on the rm (ignoring compensation costs) is Call Value = 1 VH V L
VH

Max(0, V X )dV
VL

(A.1) Thus, if the manager follows investment policy, H ], the call value at time 0 is [V 1 C0 = VH2 VH1 + 1 VH2 VH1
VH VL H V

Max(0, I X )dVH
VH1 VH2 H V

1 VH V L (A.2)

Max(0, V X )dV dVH

The rst integral represents the calls payo under the riskless project, while the second integral represents the calls expected payo under the risky project. To simplify expression (A.2), H and X > V H we consider the cases X < V separately. If X < VH , then, C0 = H VH1 V Max(0, I X ) VH2 VH1 1 + VH2 VH1 VH2 VH 1 (V X )dV dVH VH V L X H V (A.3)

44

Tom Nohel and Steven Todd

H , then, Alternatively, if X > V H VH1 V C0 = Max(0, I X ) VH2 VH1 1 + VH2 VH1 VH2 VH 1 (V X )dV dVH VH V L X X (A.4) These can be integrated to give the following valuation formulae for the call option at t = 0. If H , X <V C0 = H VH1 V Max(0, I X ) VH2 VH1 + 1 (VH2 VL 2(VH2 VH1 ) 2 )2 H VL (V 2 )2

changes by (m Max[V X , 0] + n V + K ) (m C0+ + n V0+ + K ) (B.1) Rearranging terms, the change in the value of the managers compensation, as a function of the change in rm value from t = 0+ to t = 2, i.e. as rm value changes from V0+ to V , is
Pay m (V0+ X C0+ ) + (n + m )Value; when V > X = m C0+ + n Value, when V X

(B.2)

H ) 2(X VL )(VH2 V VH2 VL + (X VL )2 ln H VL V H , If X > V

We dene PPS as the coecient of Value in (B.2), consistent with Jensen and Murphy (1990a,b). It is straightforward to show that PPS (A.5) is dened as PPS = n + [m Prob(risky) Prob(V > X | risky)] (B.3) Here, Prob(risky) is the probability that the manager pursues the risky project, and Prob(V > X | risky) is the probability that the risky project payo exceeds the option strike price, given the risky project is selected. References Aggarwal, RK and AA Samwick (1999). Empirebuilders and shirkers: investment, rm performance, and managerial incentives. Working paper. Dartmouth College, Hanover, NH. Allen, F (2001). Presidential address: do nancial institutions matter? Journal of Finance, 56, 11651176. Bebchuk, L, J Fried and D Walker (2002). Managerial power and rent extraction in the design of executive compensation. University of Chicago Law Review, 69, 751846. Brenner, M, R Sundaram and D Yermack (2000). Altering the terms of executive stock options. Journal of Financial Economics, 57, 103128. Callaghan, SR, PJ Saly and C Subramaniam (2000). The timing of option repricing. Working paper. Texas Christian University, Fort Worth, TX.

H VH1 V Max(0, I X ) C0 = VH2 VH1 1 (VH2 VL )2 (X VL )2 + 2(VH2 VH1 ) 2 2 2(X VL )(VH2 X ) VH2 VL + (X VL )2 ln (A.6) X VL Appendix B: PayPerformance Sensitivity Following Jensen and Murphy (1990), we dene PPS as the change in compensation value associated with a $1000 change in rm value for all possible liquidation values of the rm, given a xed compensation package. The value of cash compensation is uncorrelated with changes in rm value, while the value of share and option compensation is positively correlated with changes in rm value. Thus, PPS is zero for cash compensation and positive for share- and option-based compensation. In our model, the rms value is V0+ once the manager is hired. As rm value moves from V0+ to V (its liquidation value), the value of the managers incentive compensation

Stock Options and Managerial Incentives to Invest

45

Conyon, MJ and K Murphy (2000). The prince and the pauper? CEO pay in the United States and United Kingdom. The Economic Journal, 100, F640F671. Core, JE and DF Larcker (2001). Performance consequences of mandatory increases in CEO stock ownership. Journal of Financial Economics, 64, 317340. Datta, S, M Iskandar-Datta and K Raman (2001). Executive compensation and corporate acquisition decisions. Journal of Finance, 56, 22992336. DeFusco, R, R Johnson and T Zorn (1990). The eect of executive stock option plans on stockholders and bondholders. Journal of Finance, 45, 617627. Demsetz, H and B Villalonga (2001). Ownership structure and corporate performance. Journal of Corporate Finance, 7, 209233. Dimson, E and P Marsh (2001). U.K. nancial market returns, 19552000. Journal of Business, 74, 131. Fee, CE and CJ Hadlock (2001). Raids, rewards, and reputations in the market for CEO talent. Review of Financial Studies, 16, 13151357. Friend, I and M Blume (1975). The demand for risky assets. American Economic Review, December, 901922. Garen, J (1994). Executive compensation and principal-agent theory. Journal of Political Economy, 102, 11751199. Gavish, B and A Kalay (1983). On the asset substitution problem. Journal of Financial and Quantitative Analysis, 18, 2130. Gibbons, R, K Murphy (1992). Optimal incentive contracts in the presence of career concerns. Journal of Political Economy, 100, 468505. Graham, J (1999). Herding among investment newsletters: theory and evidence. Journal of Finance, 54, 237268. Guay, W (1999). The sensitivity of CEO wealth to equity risk: an analysis of the magnitude and determinants. Journal of Financial Economics, 53, 4371. Hall, B and J Leibman (1998). Are CEOs really paid like bureaucrats? Quarterly Journal of Economics, 113, 653691. Hall, B and K Murphy (2000). Optimal exercise prices for executive stock options. American Economic Review, 90, 209214.

Hall, B, and K Murphy (2002). Stock options for undiversied executives. Journal of Accounting and Economics, 33, 342. Haubrich, J (1994). Risk aversion, performance pay, and the principalagent problem. Journal of Political Economy, 102, 258276. Hemmer, T, S Matsunaga and T Shevlin (1996). The inuence of risk diversication on the early exercise of employee stock options by executive ocers. Journal of Accounting and Economics, 21, 4568. Himmelberg, CP, RG Hubbard and D Palia (1999). Understanding the determinants of managerial ownership and the link between ownership and performance. Journal of Financial Economics, 53, 353384. Holmstrom, B and J Ricart i Costa (1986). Managerial incentives and capital management. Quarterly Journal of Economics, 101, 835860. Ibbotson, R and R Sinqueeld (2001). Stocks, Bonds, Bills and Ination (SBBI) 2000 Yearbook. Chicago: Ibbotson Associates. Jagannathan, R and Z Wang (1996). The conditional CAPM and the cross-section of expected returns. Journal of Finance, 51, 353. Jensen, M and K Murphy (1990). Performance pay and top management incentives. Journal of Political Economy, 98, 225264. John, K and T John (1993). Top management compensation and capital structure. Journal of Finance, 48, 949974. Johnson, SA and YS Tian (2000). The value and incentive eects of non-traditional executive stock option plans. Journal of Financial Economics, 57, 334. Kaplan, R (1984). The evolution of management accounting. The Accounting Review, 59, 390418. Kole, S (1997). The complexity of compensation contracts. Journal of Financial Economics, 43, 79104. Lambert, R (1986). Executive eort and the selection of risky projects. Rand Journal of Economics, 17, 7788. Lambert, R, D Larcker and R Verrechia (1991). Portfolio considerations in valuing executive compensation. Journal of Accounting Research, 29, 129149.

46

Tom Nohel and Steven Todd

Litzenberger, R and E Ronn (1986). A utilitybased model of common stock price movements. Journal of Finance, March, 6792. Ljungqvist, AP and MA Habib (2003). Firm value and managerial incentives: a stochastic frontier approach. Journal of Business, forthcoming. Milgrom, P and J Roberts (1992). Economics, Organization and Management. Englewood Clis, NJ: Prentice Hall. Murphy, KJ (1999). Executive compensation. In O Ashenfelter and D Card (eds.), Handbook of Labor Economics, Vol. 3. Amsterdam: North-Holland. Nohel, T and S Todd (2003). Compensation for managers with career concerns: the role of

stock options in optimal contracts. Journal of Corporate Finance, forthcoming. Ofek, E and D Yermack (2000). Taking stock: equity-based compensation and the evolution of managerial ownership. Journal of Finance, 55, 13671384. Rajgopal, S and T Shevlin (2002). Empirical evidence on the relation between stock option compensation and risk taking. Journal of Accounting and Economics, 33, 145171. Ross, S (1977). The determination of nancial structure: the incentive-signaling approach. Bell Journal of Economics, 8, 2340. Yermack, D (1997). Good timing: CEO stock option awards and company news announcements. Journal of Finance, 52, 449476.