Managerial Over-Optimism

Managerial Over-Optimism

Managerial over-optimism is the empirical centerpiece of Perspective 2 in behavioral corporate finance: managers (specifically CEOs) systematically over-estimate the prospects of their own firms, leading to over-investment, distorted financing, and elevated investment-cash-flow sensitivity. Malmendier and Tate ($2005$) build the empirical case using observable proxies derived from CEOs’ own option-exercise behavior, sidestepping the difficulty of measuring an unobservable cognitive state.

The Malmendier-Tate framework. A risk-averse, undiversified CEO holding in-the-money company stock options should rationally exercise the options well before expiration to diversify her concentrated wealth. Holding the options long past the rational-exercise threshold reveals overconfidence: the CEO must believe her firm’s stock will continue to appreciate enough to compensate for the foregone diversification. The same logic applies to net stock purchases — buying additional company stock when already over-exposed to firm-specific risk reveals overconfidence about the firm’s future.

Three CEO-overconfidence proxies.

1. Holder67. A CEO is Holder67 in year $t$ if she holds her vested options past the fifth year of the option vesting schedule when those options are at least $67\%$ in-the-money (stock price is at least $1.67$ times the strike price). The threshold is chosen to ensure that a rational CEO would have exercised by then for diversification reasons; failing to exercise reveals overconfidence.
2. Longholder. A CEO is Longholder if she ever holds an option until expiration regardless of moneyness. This is a stronger condition; it indicates extreme overconfidence in the firm’s persistent growth.
3. NetBuyer. A CEO is NetBuyer if she is a net buyer of her own firm’s stock during the first five years of her tenure. Net buying reveals overconfidence because the CEO is voluntarily increasing her firm-specific risk concentration.

Three predicted patterns and empirical confirmation.

1. Over-investment. Overconfident CEOs over-invest because they over-estimate the NPV of available projects. Tested by regressing investment on Q (Tobin’s Q) and the overconfidence proxies; overconfident CEOs invest more aggressively at any given Q.
2. Pecking-order financing. Overconfident CEOs perceive their securities as undervalued and therefore prefer internal funds over equity issuance. Investment-cash-flow sensitivity is correspondingly stronger for overconfident CEOs.
3. Investment-cash-flow sensitivity is stronger for equity-dependent firms with overconfident CEOs. The interaction of overconfidence with equity dependence is particularly sharp.

Why the proxies work. The proxies’ validity rests on a clean theoretical fact: the standard portfolio-choice problem for an undiversified, risk-averse agent says she should exercise in-the-money options and reduce concentrated firm-specific exposure. Holding past the rational-exercise threshold has only one straightforward explanation — overconfidence about the firm’s future returns. Direct survey-based overconfidence measures would be unreliable; the option-exercise behavior provides a revealed-preference proxy that is less subject to self-report bias.

Intuition

The empirical challenge facing Perspective 2 of behavioral corporate finance is that managerial cognitive biases are inherently unobservable. Asking a CEO whether she is overconfident produces unreliable answers — overconfident managers do not generally describe themselves that way. Malmendier and Tate’s contribution is to construct behavioral proxies for overconfidence that rely on observable financial-market actions whose rational interpretation requires an extreme belief about the firm’s future.

The key insight is that holding deeply in-the-money company stock options past the rational-exercise threshold is, for a risk-averse undiversified CEO, very hard to justify under standard portfolio theory. The CEO is already heavily exposed to her firm through her human capital (her job, her reputation, her future earnings). Holding additional firm-specific options on top of this exposure makes the overall portfolio dramatically under-diversified relative to the optimum. The only way to rationalize the holding pattern within standard theory is by assuming the CEO has private information that the stock will continue to rise. Malmendier-Tate’s empirical strategy is to argue that, in the absence of insider trading on such private information (which would be illegal and detectable), the holding pattern instead reveals over-confidence — the CEO believes she has private information when she does not.

Once the proxies are in hand, the predictions follow from a simple model. An overconfident CEO who over-estimates her firm’s value will over-estimate the NPV of investment opportunities. She will under-issue equity (which she perceives as undervalued by the market) and over-rely on internal cash flow. Her investment will be more sensitive to internal cash flow than that of a rational CEO would be — exactly the investment-cash-flow sensitivity that empirical corporate-finance researchers have documented for decades and that the standard rational benchmark struggles to explain.

Properties

(Holder67 Definition) Holding deeply in-the-money options past year $5$ reveals overconfidence.

The Holder67 indicator equals $1$ in the first year a CEO crosses the threshold of holding options that are at least $67\%$ in-the-money (stock price $\ge 1.67$ times strike) past their fifth year of vesting, and remains $1$ for all subsequent years. The $67\%$ threshold is calibrated from Hall-Murphy ($2002$) calibration of the rational-exercise frontier for CEO-style undiversified portfolios. Below the threshold, holding might still be rationalizable by realistic risk-aversion parameters; above the threshold, holding requires either overconfidence or insider information.

(Longholder Definition) Holding any option to expiration reveals stronger overconfidence.

The Longholder indicator equals $1$ if the CEO ever holds a stock option until its expiration date regardless of moneyness. Holding to expiration is the maximum possible delay; even an overconfident CEO would typically exercise some intermediate options. Longholders are thus a smaller, more extreme subset of overconfident CEOs. The two proxies (Holder67 and Longholder) capture overlapping but distinct populations.

(NetBuyer Definition) Voluntarily buying additional own-stock reveals overconfidence.

The NetBuyer indicator equals $1$ if the CEO is a net buyer of her firm’s stock during her first five years of tenure. Voluntary purchase of own-firm stock is an even stronger signal of overconfidence because it actively increases concentrated firm-specific exposure rather than passively delaying its reduction. NetBuyers are less correlated with Holder67 and Longholder than the latter two are with each other; using all three proxies together provides a robustness check.

(Result 1 — Over-Investment) Overconfident CEOs invest more at any given $Q$.

Malmendier and Tate ($2005$) regress firm investment on Tobin’s $Q$, on cash flow, and on overconfidence-proxy interactions. Overconfident CEOs invest substantially more than non-overconfident CEOs at any given $Q$ — the Holder67 indicator increases the investment-Q sensitivity by roughly $30\%$. The pattern is consistent with overconfident managers over-estimating project NPVs and pursuing too many marginal investments.

(Result 2 — Pecking-Order Financing) Overconfident CEOs prefer internal funds and avoid equity issuance.

The pecking-order theory (Myers, $1984$) predicts that managers prefer internal funds, then debt, then equity — a ranking that arises from asymmetric-information considerations. An overconfident CEO has an even stronger version of the same preference: she perceives her own firm’s securities as undervalued (more so than the market believes) and therefore views debt and especially equity issuance as costly. Empirically, overconfident CEOs in the Malmendier-Tate sample issue equity less frequently than non-overconfident CEOs and rely more heavily on internal cash flow.

(Result 3 — Investment-Cash-Flow Sensitivity) Sensitivity is stronger for overconfident CEOs of equity-dependent firms.

The most important Malmendier-Tate empirical result: investment-cash-flow sensitivity (the elasticity of investment with respect to internal cash flow, controlling for $Q$) is substantially higher for overconfident CEOs, and the effect is particularly strong for firms that are equity-dependent. The double interaction (overconfidence × equity dependence) confirms the pecking-order mechanism: overconfident CEOs rely on cash flow because they refuse to issue what they perceive as undervalued equity, and equity-dependent firms have no debt-capacity escape valve. The result distinguishes the overconfidence story from classical asymmetric-information stories that would predict similar sensitivity for all CEOs.

(Why Insider Information is Ruled Out) CEOs are not on average abnormally good predictors.

The natural alternative interpretation of Holder67 — that CEOs have private information justifying their persistent holding — predicts that CEOs who hold past the threshold should subsequently outperform the market. Malmendier-Tate test this and reject: the average return on shares held past the threshold is not abnormally high. Some CEOs are right, but on average they are not, indicating that the holding pattern reflects optimism rather than information.

Examples

A CEO holding deeply in-the-money options to retirement

A technology-company CEO receives stock options each year as part of her compensation, vesting over four years. By her seventh year of tenure, her cumulative vested options are worth \50$millionincashifexercised—wellabove$67%$in-the-money.Thestandardportfolio-choicerulesayssheshouldexercisetodiversify(herhumancapitalplusoptionswouldotherwiserepresentmorethan$80%$ofhernetworthinfirm-specificrisk).Sheinsteadholdsalloptionsthroughyear$10$. By the Holder67 criterion, she is classified as overconfident. Malmendier-Tate’s empirical regression then predicts that her firm should exhibit higher investment-cash-flow sensitivity, less equity issuance, and more aggressive investment than otherwise-comparable firms with non-overconfident CEOs.

The pecking order in a Holder67 firm

An overconfident CEO faces a marginal positive-NPV project requiring \100$millioninfinancing.Thefirmhas$$30$millionininternalcash;itcouldborrow$$50$milliongivenitscurrentdebtcapacity;itcouldissue$$20$millioninequity.TheCEObelievesherstockisundervaluedbythemarketby$30%$—equityissuancewoulddiluteexistingshareholdersbywhatsheperceivesastoomuch.Shefundstheprojectentirelyfromcashanddebt,financingonly$$80$millionofthe$$100$millionrequired.Theremaining$$20$ million in NPV-positive investment is forgone. The pecking-order pattern arises endogenously from the overconfident CEO’s belief that her equity is too cheap to issue.

Why insider information fails as an alternative

Suppose a non-overconfident CEO holds her options past year $5$ because she has favorable insider information about an upcoming product launch. Her shares should subsequently outperform the market significantly when the launch is announced. If many CEOs were doing this, Malmendier-Tate would observe that Holder67-classified firms have abnormally high subsequent returns. The actual data show the opposite: returns to Holder67 firms are not abnormally high (and are slightly lower than benchmark in some sub-samples). The information-based alternative explanation therefore fails empirically; the holding pattern reflects an optimism that is not validated by subsequent performance.

See Also

• 8.13. Behavioral Corporate Finance — parent hub.
• 8.13.6. Managerial Overconfidence and Miscalibration — sibling: a different bias (variance underestimation) with related corporate-decision implications.
• 8.13.3. Investment Policy and Equity Dependence — sibling: investment-sensitivity pattern that overconfidence interacts with.
• 8.13.7. Single Firm-Wide Discount Rate — sibling: another bounded-rationality pattern in CEO investment decisions.
• 8.13.1. Market Timing of Security Issuance — Perspective-1 counterpart that overconfidence makes harder for the firm to follow.
• 5. Corporate Finance — broader corporate-finance hub; the rational benchmark this pattern departs from.