The positive role of overconfidence and Optimism in Investment Policy b Simon gervais J. B. Heaton errance Deans 4 September 2002 Y *This paper is an updated version of a previous working paper, Capital Budgeting in the Presence of Managerial Overconfidence and Optimism, by the same authors. Financial support by the Rodney L. white Center for Financial Research is gratefully acknowledged. The authors would like to thank Andrew Abel, Jonathan Berk, Domenico Cuoco, David Denis, Janice Eberly, Robert Goldstein, Peter Swan, and seminar participants at the 2000 meetings of the European Finance Association, the 2001 meetings of the American Finance Association, and the Wharton School for their comments and suggestions. Heaton acknowledges that the opinions expressed here are his own, and do not reflect the position of Bartlit Beck Herman Palenchar scott or its attorneys. All remaining errors are of course the authors'responsibility. fInance Department, Wharton School, University of Pennsylvania, Steinberg Hall-Dietrich Hall, Suite 2300, Philadelphia, PA 19104-6367, gervais @wharton. upenn. edu, (215)898-2370 f Adjunct Associate Professor of Finance, Duke University Fuqua School of Business and Duke Global Capital Markets Center, and Associate, Bartlit Beck Herman Palenchar Scott, 54 West Hubbard, Suite 300, Chicago, IL 60610, jb. heaton obartlit-beck com, (312)494-4425 sHaas School of Business, 545 Student Services#1900, University of California at Berkeley, Berkeley CA 94720-1900, odean @haas. berkeley. edu, (510)642-6767
The Positive Role of Overconfidence and Optimism in Investment Policy∗ by Simon Gervais† J. B. Heaton‡ Terrance Odean§ 4 September 2002 ∗This paper is an updated version of a previous working paper, “Capital Budgeting in the Presence of Managerial Overconfidence and Optimism,” by the same authors. Financial support by the Rodney L. White Center for Financial Research is gratefully acknowledged. The authors would like to thank Andrew Abel, Jonathan Berk, Domenico Cuoco, David Denis, Janice Eberly, Robert Goldstein, Peter Swan, and seminar participants at the 2000 meetings of the European Finance Association, the 2001 meetings of the American Finance Association, and the Wharton School for their comments and suggestions. Heaton acknowledges that the opinions expressed here are his own, and do not reflect the position of Bartlit Beck Herman Palenchar & Scott or its attorneys. All remaining errors are of course the authors’ responsibility. †Finance Department, Wharton School, University of Pennsylvania, Steinberg Hall - Dietrich Hall, Suite 2300, Philadelphia, PA 19104-6367, gervais@wharton.upenn.edu, (215) 898-2370. ‡Adjunct Associate Professor of Finance, Duke University Fuqua School of Business and Duke Global Capital Markets Center, and Associate, Bartlit Beck Herman Palenchar & Scott, 54 West Hubbard, Suite 300, Chicago, IL 60610, jb.heaton@bartlit-beck.com, (312) 494-4425. §Haas School of Business, 545 Student Services #1900, University of California at Berkeley, Berkeley, CA 94720-1900, odean@haas.berkeley.edu, (510) 642-6767
The positive role of overconfidence and Optimism in Investment Policy Abstract managers with those of rational managers. We reach the surprising conclusion that managE We use a simple capital budgeting problem to contrast the decisions of overconfident, optimis overconfidence and optimism can increase the value of the firm. Risk-averse rational managers will postpone the decision to exercise real options longer than is in the best interest of shareholders Overconfident managers underestimate the risk of potential projects and are therefore less likely t postpone the decision to undertake. Optimistic managers, too, undertake projects quickly. Thus moderately overconfident or optimistic managers make decisions that are in the better interest of shareholders than do rational managers. Overly overconfident or optimistic managers may be too eager to undertake projects. This tendency can sometimes be controlled by increasing hurdle rates for risky projects. While compensation contracts that increase the convexity of manager payoffs can be used to realign the decisions of a rational manager with those of shareholders, it is less expensive to simply hire a moderately overconfident manager. The gains from overconfidence and optimism will at times be sufficient that shareholders actually prefer an overconfident, optimistic manager with less ability to a rational manager with greater ability JEL classification codes: G31 L21
The Positive Role of Overconfidence and Optimism in Investment Policy Abstract We use a simple capital budgeting problem to contrast the decisions of overconfident, optimistic managers with those of rational managers. We reach the surprising conclusion that managerial overconfidence and optimism can increase the value of the firm. Risk-averse rational managers will postpone the decision to exercise real options longer than is in the best interest of shareholders. Overconfident managers underestimate the risk of potential projects and are therefore less likely to postpone the decision to undertake. Optimistic managers, too, undertake projects quickly. Thus moderately overconfident or optimistic managers make decisions that are in the better interest of shareholders than do rational managers. Overly overconfident or optimistic managers may be too eager to undertake projects. This tendency can sometimes be controlled by increasing hurdle rates for risky projects. While compensation contracts that increase the convexity of manager payoffs can be used to realign the decisions of a rational manager with those of shareholders, it is less expensive to simply hire a moderately overconfident manager. The gains from overconfidence and optimism will at times be sufficient that shareholders actually prefer an overconfident, optimistic manager with less ability to a rational manager with greater ability. JEL classification codes: G31, L21
1 Introduction A vast experimental literature finds that individuals are usually optimistic (i.e, they believe out comes favorable to themselves to be more likely than they actually are) and overconfident (i.e., they believe their knowledge is more precise than it actually is). Since optimism and overconfidence di rectly influence decision making, it is natural to ask how optimistic and overconfident managers will ffect the value of the firm. Are managerial optimism and overconfidence sufficiently detrimental to firm value that shareholders should actively avoid hiring optimistic and overconfident managers? What possible benefits might optimistic and overconfident managers bring to the firm? We use a simple model of capital budgeting to contrast the decisions of overconfident, optimistic managers with those of rational managers. We reach the surprising conclusion that managerial overconfidence and optimism can increase the value of the firm. Moderate overconfidence can align more closely with those of shareholders. However, extreme managerial overconfidence and optimism are detrimental to the firm. Our analysis starts with the observation that many capital budgeting decisions can be viewed as decisions whether or not to exercise real options(Dixit and Pindyck, 1994). Because of their greater risk aversion, rational managers will postpone the decision to exercise real options longer than is in the best interest of shareholders. As Treynor and Black(1976)write If the corporation undertakes a risky new venture, the stockholders may not be very concerned, because they can balance this new risk against other risks that they hold heir portfolios. The managers, however, do not have a portfolio of the corporation does badly because the new venture fails, they do not have any risks except the others taken by the same corporation to balance against it. They are hurt by a failure more than the stockholders, who also hold stock in other corporations, are hurt.” Since overconfident managers believe that the uncertainty about potential project is less than it actually is, they are less likely to postpone the decision to undertake the project. Thus moder ately overconfident managers make decisions that are in the better interest of shareholders than do rational managers. Overconfident managers also benefit the firm by expending more effort than rational managers, as they overestimate the value of that effort. Optimistic managers believe that the expected net present value of potential projects is greater than it actually is. Like overconfident managers, optimistic managers undertake projects more quickly than do rational managers. How
1 Introduction A vast experimental literature finds that individuals are usually optimistic (i.e., they believe outcomes favorable to themselves to be more likely than they actually are) and overconfident (i.e., they believe their knowledge is more precise than it actually is). Since optimism and overconfidence directly influence decision making, it is natural to ask how optimistic and overconfident managers will affect the value of the firm. Are managerial optimism and overconfidence sufficiently detrimental to firm value that shareholders should actively avoid hiring optimistic and overconfident managers? What possible benefits might optimistic and overconfident managers bring to the firm? We use a simple model of capital budgeting to contrast the decisions of overconfident, optimistic managers with those of rational managers. We reach the surprising conclusion that managerial overconfidence and optimism can increase the value of the firm. Moderate overconfidence can align managers’ preferences for risky projects more closely with those of shareholders. However, extreme managerial overconfidence and optimism are detrimental to the firm. Our analysis starts with the observation that many capital budgeting decisions can be viewed as decisions whether or not to exercise real options (Dixit and Pindyck, 1994). Because of their greater risk aversion, rational managers will postpone the decision to exercise real options longer than is in the best interest of shareholders. As Treynor and Black (1976) write: “If the corporation undertakes a risky new venture, the stockholders may not be very concerned, because they can balance this new risk against other risks that they hold in their portfolios. The managers, however, do not have a portfolio of employers. If the corporation does badly because the new venture fails, they do not have any risks except the others taken by the same corporation to balance against it. They are hurt by a failure more than the stockholders, who also hold stock in other corporations, are hurt.” Since overconfident managers believe that the uncertainty about potential project is less than it actually is, they are less likely to postpone the decision to undertake the project. Thus moderately overconfident managers make decisions that are in the better interest of shareholders than do rational managers. Overconfident managers also benefit the firm by expending more effort than rational managers, as they overestimate the value of that effort. Optimistic managers believe that the expected net present value of potential projects is greater than it actually is. Like overconfident managers, optimistic managers undertake projects more quickly than do rational managers. How- 1
ever, unlike overconfident managers, optimistic managers will sometimes undertake projects that actually have negative expected net present values. This error may be mitigated by raising hurdle rates While compensation contracts that increase the convexity of manager payoffs can be used to realign the decisions of a rational manager with those of shareholders, it may be less expensive to simply hire an overconfident, optimistic manager. The gains from overconfidence and optimism will at times be sufficient that shareholders actually prefer an overconfident, optimistic manager with less ability to a rational manager with greater ability. Extreme overconfidence or optimism is, however, always detrimental to the firm. Extremely overconfident or optimistic managers will perceive too little risk or too little chance of failure. They will greatly underestimate of delaying a project or greatly overestimate the likelihood of success. When such individuals in charge of a firm's capital budgeting decisions, they will destroy that firm's value in the long run Our research helps to explain a puzzle in corporate finance. If rational individuals make better decisions than those influenced by behavioral biases, such as overconfidence, why are many CEO verconfident(Audia, Locke and Smith, 2000; Malmendier and Tate, 2001)? This puzzle can be viewed from the perspective of the individual manager and that of the firm. Gervais and Odean(2001)demonstrate, in the context of investors, that the human tendency to take too much credit for success and attribute too little credit to chance can cause successful people to become erconfident. Thus, in a corporate setting, managers who successfully climb the corporate ladder to become CEOs are likely to also become overconfident. In the current paper, we address the puzzle of CEO overconfidence from the perspective of the firm. We show that it can be in the best interest of shareholders to hire managers(e. g, CEOs) who are overconfident Our paper proceeds as follows. Section 2 reviews some of the literature on optimism and verconfidence. Section 3 introduces a simple capital budgeting problem that is used throughout the paper to analyze the effects of behavioral biases on the value of the firm. The same section presents the first-best solution, which serves as a benchmark for later sections. Section 4 formally introduces the concepts of overconfidence and optimism, and shows how these individual traits can ffect the value of the firm when a manager's sole intention is to maximize firm value. The principal agent nature of the relationship between firm owners and managers is analyzed in section 5. This section shows how contracting interacts with manager biases to solve the firms agency problems section 6. we show how our basic model can be extended to accommodate other forces that are
ever, unlike overconfident managers, optimistic managers will sometimes undertake projects that actually have negative expected net present values. This error may be mitigated by raising hurdle rates. While compensation contracts that increase the convexity of manager payoffs can be used to realign the decisions of a rational manager with those of shareholders, it may be less expensive to simply hire an overconfident, optimistic manager. The gains from overconfidence and optimism will at times be sufficient that shareholders actually prefer an overconfident, optimistic manager with less ability to a rational manager with greater ability. Extreme overconfidence or optimism is, however, always detrimental to the firm. Extremely overconfident or optimistic managers will perceive too little risk or too little chance of failure. They will greatly underestimate the option value of delaying a project or greatly overestimate the likelihood of success. When such individuals are put in charge of a firm’s capital budgeting decisions, they will destroy that firm’s value in the long run. Our research helps to explain a puzzle in corporate finance. If rational individuals make better decisions than those influenced by behavioral biases, such as overconfidence, why are many CEOs overconfident (Audia, Locke and Smith, 2000; Malmendier and Tate, 2001)? This puzzle can be viewed from the perspective of the individual manager and that of the firm. Gervais and Odean (2001) demonstrate, in the context of investors, that the human tendency to take too much credit for success and attribute too little credit to chance can cause successful people to become overconfident. Thus, in a corporate setting, managers who successfully climb the corporate ladder to become CEOs are likely to also become overconfident. In the current paper, we address the puzzle of CEO overconfidence from the perspective of the firm. We show that it can be in the best interest of shareholders to hire managers (e.g., CEOs) who are overconfident. Our paper proceeds as follows. Section 2 reviews some of the literature on optimism and overconfidence. Section 3 introduces a simple capital budgeting problem that is used throughout the paper to analyze the effects of behavioral biases on the value of the firm. The same section presents the first-best solution, which serves as a benchmark for later sections. Section 4 formally introduces the concepts of overconfidence and optimism, and shows how these individual traits can affect the value of the firm when a manager’s sole intention is to maximize firm value. The principalagent nature of the relationship between firm owners and managers is analyzed in section 5. This section shows how contracting interacts with manager biases to solve the firm’s agency problems. In section 6, we show how our basic model can be extended to accommodate other forces that are 2
likely to play a role in capital budgeting problems. Finally, section 7 discusses our findings and concludes. All the proofs are contained in the appendix 2 elated work 2.1 perimental Studies For this paper, we define optimism to be the belief that favorable future events are more likely than they actually are. Researchers find that, generally, individuals are unrealistically optimistic about future events. They expect good things to happen to themselves more often than to their peers (Weinstein, 1980; Kunda, 1987). For example, Ito(1990)reports that foreign exchange companies are more optimistic about how exchange rate moves will affect their firm than how they will affect others. People overestimate their ability to do well on tasks and these overestimates increase when the task is perceived to be controllable(Weinstein, 1980) and when it is of personal importance (Frank, 1935). March and Shapira(1987) find that managers tend to believe that outcomes are largely controllable and that projects under their supervision are less risky than is actually the case. Finally, optimism is most severe among more intelligent individuals(Klaczynski and Fauth 1996)and, we expect, most top managers are intelligent For this paper, we define overconfidence to be the belief that the precision of one's information greater than it actually is, that is, one puts more weight on one's information than is warranted Studies of the calibration of subjective probabilities find that individuals do tend to overestimate the precision of their information(Alpert and Raiffa, 1982; Fischhoff, Slovic and Lichtenstein, 1977).2 Such overconfidence has been observed in many professional fields. Clinical psychologist (Oskamp, 1965), physicians and nurses( Christensen-Szalanski and Bushyhead, 1981; Baumann, er a ) investment bankers(Stael von Holstein, 1972), engineers(Kidd 1970), entrepreneurs(Cooper, Woo and Dunkelberg, 1988), lawyers(Wagenaar and Keren, 1986) negotiators(Neale and Bazerman, 1990), and managers(Russo and Schoemaker, 1992)have all been observed to exhibit overconfidence in their judgments The best established finding in the calibration literature is that people tend to be overconfident in answering questions of moderate to extreme difficulty(Fischhoff, Slovic and Lichtenstein, 1977; Over two years, the Japan Center for International Finance conducted a bi-monthly survey of foreign exchange experts in 44 companies. Each was asked for point estimates of future yen /dollar exchange rates. The experts in import-oriented companies expected the yen to appreciate(which would favor their company), while those in export- oriented companies expected the yen to fall(which would favor their company). People are even unrealistically optimistic about pure chance events(Marks, 1951; Irwin, 1953: Langer and Roth, 1975) See Lichtenstein, Fischhoff, and Phillips(1982)for a review of the calibration literature
likely to play a role in capital budgeting problems. Finally, section 7 discusses our findings and concludes. All the proofs are contained in the appendix. 2 Related Work 2.1 Experimental Studies For this paper, we define optimism to be the belief that favorable future events are more likely than they actually are. Researchers find that, generally, individuals are unrealistically optimistic about future events. They expect good things to happen to themselves more often than to their peers (Weinstein, 1980; Kunda, 1987). For example, Ito (1990) reports that foreign exchange companies are more optimistic about how exchange rate moves will affect their firm than how they will affect others.1 People overestimate their ability to do well on tasks and these overestimates increase when the task is perceived to be controllable (Weinstein, 1980) and when it is of personal importance (Frank, 1935). March and Shapira (1987) find that managers tend to believe that outcomes are largely controllable and that projects under their supervision are less risky than is actually the case. Finally, optimism is most severe among more intelligent individuals (Klaczynski and Fauth, 1996) and, we expect, most top managers are intelligent. For this paper, we define overconfidence to be the belief that the precision of one’s information is greater than it actually is, that is, one puts more weight on one’s information than is warranted. Studies of the calibration of subjective probabilities find that individuals do tend to overestimate the precision of their information (Alpert and Raiffa, 1982; Fischhoff, Slovic and Lichtenstein, 1977).2 Such overconfidence has been observed in many professional fields. Clinical psychologists (Oskamp, 1965), physicians and nurses (Christensen-Szalanski and Bushyhead, 1981; Baumann, Deber and Thompson, 1991), investment bankers (Sta¨el von Holstein, 1972), engineers (Kidd, 1970), entrepreneurs (Cooper, Woo and Dunkelberg, 1988), lawyers (Wagenaar and Keren, 1986), negotiators (Neale and Bazerman, 1990), and managers (Russo and Schoemaker, 1992) have all been observed to exhibit overconfidence in their judgments. The best established finding in the calibration literature is that people tend to be overconfident in answering questions of moderate to extreme difficulty (Fischhoff, Slovic and Lichtenstein, 1977; 1Over two years, the Japan Center for International Finance conducted a bi-monthly survey of foreign exchange experts in 44 companies. Each was asked for point estimates of future yen/dollar exchange rates. The experts in import-oriented companies expected the yen to appreciate (which would favor their company), while those in exportoriented companies expected the yen to fall (which would favor their company). People are even unrealistically optimistic about pure chance events (Marks, 1951; Irwin, 1953; Langer and Roth, 1975). 2See Lichtenstein, Fischhoff, and Phillips (1982) for a review of the calibration literature. 3