looking in the rearview mirror: the effect of managers ...download.xuebalib.com/1fe8d0sawnlx.pdf ·...

Looking in the Rearview Mirror: The Effectof Managers’ Professional Experience onCorporate Financial Policy

Amy DittmarRoss School of Business, University of Michigan

Ran DuchinFoster School of Business, University of Washington

We track the employment history of over 9,000 managers to study the effects of professionalexperiences on corporate policies. Our identification strategy exploits exogenous CEOturnovers and employment in other firms and in non-CEO roles. Firms run by CEOs whoexperienced distress have less debt, save more cash, and invest less than other firms, withstronger effects in poorly governed firms. Experience has a stronger influence when it ismore recent or occurs during salient periods in a manager’s career. We find similar effectsfor CFOs. The results suggest that policies vary with managers’experiences and throughoutmanagers’ careers. (JEL G30, G31, G32)

Received January 28, 2014; accepted May 17, 2015 by Editor David Denis.

A growing body of evidence suggests that managerial traits affect corporatepolicies even after controlling for the firm-, industry-, and market-leveleconomic indicators that influence these policies. However, we know relativelylittle about how managers’ decision-making develops throughout their careers.In this paper, we investigate the potential impact of professional, or work-related, experiences of managers on corporate financial and investment policies.

By studying the impact of professional experiences, we provide newevidence on how both CEOs and CFOs develop their managerial style andhow the timing, frequency, and saliency of their experiences affect theirsubsequent decision-making. We exploit the variation along these dimensionsacross professional experiences to test several experimentally well-establishedbehavioral phenomena that have not been previously studied in the context

We gratefully acknowledge the helpful comments from Espen Eckbo, Sandy Klasa, Camelia Kuhnen, andGeoffrey Tate as well as seminar participants at the 2014 Financial Intermediation Research Society Conference,the 2014 FSU SunTrust Conference, 2013 European FinanceAssociation Conference, the 2012 Pacific NorthwestFinance Conference, the 2012 Miami Behavioral Finance Conference, Australian National University, DrexelUniversity, Tulane University, the University ofArizona, the University of Maryland, the University of Michigan,the University of New South Wales, the University of Pennsylvania, the University of Sydney, and the Universityof Technology Sydney. Send correspondence to Amy Dittmar, Ross School of Business, University of Michigan,701 Tappan Avenue, Ann Arbor, Michigan, 48109; telephone: (734) 764-3108. E-mail: [email protected].

© The Author 2015. Published by Oxford University Press on behalf of The Society for Financial Studies.All rights reserved. For Permissions, please e-mail: [email protected]:10.1093/rfs/hhv051 Advance Access publication September 4, 2015

at Florida Atlantic U

niversity on February 25, 2016http://rfs.oxfordjournals.org/

Dow

nloaded from

http://rfs.oxfordjournals.org/

The Review of Financial Studies / v 29 n 3 2016

of real-world decisions made by corporate executives. We also examine howprofessional experiences attenuate or enhance the impact of previously studiedmanagerial traits and personal experiences on corporate policy.

The importance of experience in decision-making is demonstrated in thepsychology literature (Nisbett and Ross 1980). Studies show that experiencemay lead individuals to make decisions that differ from those based on expectedutility theory because they only have access to samples of past outcomes andnot the full outcome distributions (e.g., Hertwig, Barron, Weber, and Erev 2004;Hertwig and Erev 2009; and Hertwig 2012). In finance, a growing literatureexamines how individual experiences affect investor behavior (Vissing-Jorgensen 2004; Kaustia and Knupfer 2008; Greenwood and Nagel 2009;Chiang et al. 2011; Malmendier and Nagel 2011, Malmendier and Nagel Forth-coming). Malmendier and Tate (2005) and Malmendier, Tate, and Yan (2011)show that managers’ early experiences, such as growing up during the GreatDepression and military service, affect corporate leverage and investment.

We build on this literature and explore how professional experiences affectmanagers’ decisions. Our focus is on past experience with negative corporateoutcomes such as bankruptcy and financial difficulties or shocks, though we alsoexplore experience with positive outcomes in subsequent tests. Experiencingtroubles may alter risk preferences or expectations, and lead managers toimplement more conservative policies. This hypothesis is consistent with the“hot stove” effect, which implies a bias against risky alternatives to avoidactions that have led to poor outcomes (Denrell and March 2001). Indeed,Eckbo and Thornburn (2003) and Eckbo, Thornburn, and Wang (2012) showthat bankruptcy leads to poor outcomes for the CEO.

To determine the effect of managers’ professional experiences on corporatedecisions, we use data from ExecuComp and BoardEx to track the employmenthistory of approximately 5,200 CEOs and 4,000 CFOs. After excludingmanagers with incomplete employment histories, the average CEO in oursample has twenty-one years of employment data at four different firms. Weuse these data to determine if a manager was previously employed by a troubledfirm. To separate CEO effects from firm effects, we require that the professionalexperience have taken place at a different firm than the current firm and thatthe current firm itself did not experience difficulties.

We construct four measures of poor corporate outcomes. Based on eachmeasure, we define a Professional experience indicator that equals one if themanager was employed by a firm that experienced trouble during her tenure.To mitigate the concern that the CEO is chosen based on experience runningtroubled firms, our main measures consider only experience in roles other thanCEO, though we also explore experience as CEO in subsequent tests. Thefirst measure is based on bankruptcy filings. In our sample, 0.8% of the CEOspreviously worked at a firm that filed for bankruptcy. Because bankruptcy isrelatively infrequent and salient enough to affect a manager’s career directly,we construct three additional measures. These measures are based on adverse

566



Dow

nloaded from


Looking in the Rearview Mirror

shocks to a firm’s cash flows, stock returns, and credit ratings. Depending onthe measure employed, 6.4–11.5% of the CEOs in our sample experiencedtrouble in at least one year of prior employment. We also create a compositeindex equal to one if any of these measures equals one, with 23.8% of CEOsexperiencing difficulties using the index.

In panel regressions, we find that firms run by a CEO who was previouslyemployed at a troubled firm hold less debt and more cash, and invest less.To address the concern that these findings are driven by either observableor unobservable firm characteristics that are correlated with the CEO’sprofessional experience and the firm’s policies, we control for time-varyingfirm and manager characteristics and year and firm fixed effects. The effectsof professional experience are statistically significant and economicallymeaningful. For the average firm, past experience at a troubled firm is associatedwith a 7–13% reduction in debt (having 1.5–2.8 percentage points less debt-to-assets), a 5–12% increase in cash (holding 1.1–2.6 percentage points morecash-to-assets), and a 5–10% reduction in capital expenditure (investing 0.3–0.6percentage points less in capital expenditures-to-assets).

Our use of a firm fixed effects model controls for the possibility a managerwith professional experience in a distressed firm may join a firm that is alreadyconservative prior to her appointment. However, a remaining concern is thatCEO changes themselves might be triggered by a change in firm characteristicsthat affects corporate policy. To address this concern, we exploit the change tothe professional experience of the CEO around exogenous CEO turnovers; thatis, turnovers due to natural causes (death or illness), planned retirements, orscheduled succession plans, and a subset of successions by internal candidates,which are unlikely to be associated with managerial performance or a changein the firm’s conditions. It is important to note that although recent work byFee, Hadlock, and Pierce (2013) shows that, on average, exogenous turnoversare not accompanied by substantial changes in corporate policy, our empiricaldesign exploits cross-sectional variation in professional experience within thesubset of exogenous turnovers.1

In firm fixed effects and first differences models around exogenous CEOturnovers, our results indicate that CEOs with professional experience attroubled firms decrease debt-to-assets by 0.9–1.5 percentage points, increasecash-to-assets by 2.0–2.8 percentage points, and decrease capital expenditures-to-assets by 0.4–0.5 percentage points after they became CEOs.

The above evidence demonstrates the importance of managers’ professionalexperience and is consistent with prior studies that show the impact of personalexperiences, such as being born in the Great Depression era or serving inthe military, and other managerial traits, such as overconfidence, on corporatepolicy. To disentangle the effect of professional experience from personal

1 In untabulated tests, similar to Fee, Hadlock, and Pierce (2013), we find that, on average, exogenous CEOturnovers are not accompanied by significant changes in leverage, cash, or investment policies.

567



Dow

nloaded from



experience and other traits, we repeat our tests with measures of GreatDepression experience, military service, and overconfidence. We find thatprofessional experience continues to have a significant impact on corporatepolicy, and moreover, that the effect of professional experience is frequentlystronger than that of the other factors.

Because professional experiences occur throughout a manager’s career, theymay attenuate or enhance the impact of personal experiences or overconfidence.We therefore interact the manager’s professional experience with thesemeasures. We find that professional experience attenuates the impact ofoverconfidence and enhances the impact of Depression-era experience.

To further examine how professional experience influences managers’ stylethroughout their career, we exploit the variation in the timing, frequency,and saliency of professional experiences. Several theoretical and experimentalstudies suggest that these factors may influence the impact of experience. Westudy three hypotheses related to this literature.

First, the reinforcement learning hypothesis suggests that individuals willrepeat positive outcomes and employ similar outcomes more frequently (Erevand Roth 1998), consistent with Watson’s (1930) recency law, which states thatthe event observed most recently is more likely repeated. To test the recencyhypothesis, we compare recent and distant experiences and find that recentexperiences have a stronger impact on corporate decisions.

Second, the saliency hypothesis (Tversky and Kahneman 1973, 1974;Bordola, Gennaioli, and Shleifer 2012a,b) suggests that individuals infer thefrequency or probability of an event based on its saliency. According to thishypothesis, more salient events will have a stronger impact on individualbehavior. To examine the saliency hypothesis, we reestimate our measure ofprofessional experience using experience that occurred when the manager wasa CEO or top-five executive at another firm, rather than a lower-rank employee.We also examine the impact of repeated experiences. We find that experiencesduring these salient periods or that occur multiple times have an even strongereffect on corporate policy.

Third, we consider the pessimism-bias hypothesis, which predicts thatnegative outcomes have asymmetrically larger effects than positive outcomes.Consistent with this hypothesis, Kahneman and Tversky (1979) document theimpact of loss aversion, and Kuhnen (2015) shows in a lab experiment thatnegative outcomes have a stronger impact than positive outcomes and leadindividuals to become overly pessimistic about investment alternatives. Toexamine this hypothesis, we define an experience index that considers positiveprofessional experiences. We reestimate our analysis and find that positiveexperience has no impact on corporate policy.

The variation in professional experiences also allows us to disentangle theimpact of CEO and CFO effects. To do this, we recreate our measures ofexperience for the CFO and find that 0.6–11.1% of the CFOs in our sampleworked at a troubled firm (in roles other than CFO). When we investigate the

568



Dow

nloaded from



joint impact of CEO and CFO experience on corporate policy, we find that bothCEO and CFO experiences affect corporate financing policy. However, onlyCEO experience affects corporate investment policy, suggesting that CFOs donot exert significant influence on the firm’s investment decisions. Further, whenboth the CEO and the CFO experience distress, the effects of experience onleverage and cash policy are even stronger.

In our final analysis, we study the relation between professional experienceand firm value. This analysis seeks to distinguish between two possibleinterpretations. On the one hand, experience-driven conservatism can besuboptimal if CEOs who experienced distress become too risk averse oroverestimate the likelihood and implications of distress. On the other hand,these conservative policies may result from CEO learning or altering the viewof the average CEO who underestimates risk, as recently found by Ben-David,Graham, and Harvey (2013).

Our evidence is more consistent with the suboptimal interpretation. Wefind that firms that hire a CEO who experienced distress have a reductionof 12 basis points in return on assets (ROA) and 5.5% in Tobin’s q. However,abnormal announcement returns around the appointment of CEOs do not differsignificantly based on the CEO’s professional experience.

Consistent with this interpretation, we also find that the effect of aCEO’s professional experience is significantly more pronounced at firmswith weaker governance, as measured by the E-index (Bebchuk, Cohen, andFerrell 2009) of shareholder rights, the presence of share block holders, andboard independence. A decrease of one standard deviation in the qualityof corporate governance, as measured by a composite index of these threegovernance proxies, is associated with an increase of 18.6–29.8% in the effectof professional experience on corporate policy. Overall, these results providesuggestive, indirect evidence that distress leads managers to enact overlyconservative policies when they are not monitored by investors or the board.

Our paper contributes to the growing literature that studies the effects ofmanagers on corporate policies. Although existing evidence suggests thatmanagement style affects corporate policies (Bertrand and Schoar 2003),largely following endogenous CEO turnovers (Fee, Hadlock, and Pierce2013), we still know relatively little about the determinants of this styleand its evolution throughout a manager’s career. Our paper improves ourunderstanding of this process by studying the effect of professional experienceson financing and investment decisions.

1. Sample and Data

1.1 FirmsOur initial sample consists of 11,578 industrial firms in the CRSP/Compustatfile over 1980–2011. Industrial firms are defined as companies with SICcodes outside the ranges 4900–4949 (utilities) and 6000–6999 (financials).

569



Dow

nloaded from



Table 1Firm-level summary statistics

Variable Mean Median Std. Dev. N Obs. Mean full Median fullsample sample

Leverage 0.210 0.157 0.251 35,377 0.232 0.178Cash holdings 0.221 0.125 0.239 35,227 0.218 0.123Capital expenditure 0.061 0.041 0.065 35,047 0.071 0.046Tangibility 0.263 0.199 0.219 35,539 0.299 0.233Market-to-book 2.237 1.590 1.911 35,601 2.088 1.446Profitability 0.011 0.036 0.260 35,588 −0.051 0.029Size 5.695 5.602 1.965 35,601 5.218 5.063Cash flow 0.017 0.070 0.212 35,088 −0.004 0.057Industry cash flow volatility 0.086 0.074 0.048 35,229 0.081 0.067

This table presents summary statistics for firm-level variables used in the analyses. The sample comprisesindustrial firms in the Compustat/CRSP file from 1980 to 2011, with available information about the CEO’sprior employment for the last 10 years or more (without gaps) prior to joining the firm. Leverage is total debtdivided by total assets. Cash holdings are cash and short-term investments divided by total assets. Tangibility ismeasured as net property, plant, and equipment, divided by total assets. Market-to-book (or Tobin’s q) is measuredas the book value of total assets minus book value of equity plus market value of equity divided by total assets.Profitability is net income divided by total assets. Size is the natural logarithm of the book value of total assets.Cash flow is measured as earnings less interest and taxes, divided by total assets. Industry cash flow volatility isthe 10-year rolling window of median volatility of cash flow across the 48 Fama-French industries. The last twocolumns compare the average values in our sample to the average values in the full sample of industrial firms inthe Execucomp/BoardEx sample from 1980 to 2011.

We exclude firms that are not incorporated in the United States and those thatdo not have securities assigned a CRSP security code of 10 or 11. Because weare interested in CEOs’ professional experiences we exclude firms whose CEOis missing from ExecuComp and BoardEx. We find 5,498 firms and 52,017firm-year observations for which the CEO has nonmissing data on previousemployment in at least one firm that appears on Compustat.

Next, we exclude from our sample CEOs with relatively short observableemployment histories of less than 10 years before the start of their currentemployment. We also exclude CEOs whose employment history over the prior10 years before the start of current employment is incomplete (one or moreyears are missing). We impose these sample screens because the length of theobserved employment history and the gaps in the data may be nonrandom,and potentially correlated with CEO attributes such as tenure or age, and withfirm attributes such as size, industry, and IPO cohort. However, to ensure thisrestriction is not driving our findings, we repeat our tests keeping all managerswith any employment history and get similar results. To separate CEO effectsfrom firm effects, we also require that the firms in our sample did not experiencedifficulties according to any of our measures, described in Section 1.3. Afterimposing these screens, our final sample includes between 29,226 and 35,601firm-year observations, depending on the measure of difficulties used.

Table 1 presents summary statistics. We winsorize all variables at the 1stand 99th percentiles to lessen the influence of outliers. The main variables ofinterest are a firm’s: (i) leverage, defined as the ratio of short-term and long-term debt to book assets; (ii) cash holdings, defined as the ratio of cash andshort-term investments to book assets; and (iii) capital expenditure, defined as

570



Dow

nloaded from



the ratio of capital expenditures to book assets. Table 1 shows that leverageratios have a pooled mean of 21.0%, cash ratios have a pooled mean of 22.1%,and capital expenditures have a pooled mean of 6.1%. The average firm has afirm size (log of assets) of 5.7. Table 1 also shows that the average firm has acash flow-to-assets ratio of 1.7%, a market-to-book ratio of 2.2, and a ratio offixed assets to assets (tangibility) of 26.3%.

Table 1 also compares our sample firms and the sample of public industrialfirms in the Compustat universe. Not surprisingly, the firms in our sampletend to be larger, since their managers are more visible and have longer workexperience. We also find that the firms in our sample tend to be more profitable(and therefore less likely distressed) compared with the average Compustatfirm. Thus, we believe that our sample firms are a good laboratory to test whethera CEO’s professional experience of distress, rather than a firm’s distress, leadsto the implementation of more conservative corporate policies.

1.2 ManagersOur sample of executives consists of 9,133 individuals. This group includes5,178 CEOs and 3,955 CFOs who served at our sample firms between 1980and 2011. To collect employment information on CEOs and CFOs, we useboth ExecuComp and BoardEx. For each executive in our sample, we collectall available information on her employment history, including the identity ofprevious employers, dates of employment, and the role title. We then match theprior employers to Compustat firms and use Compustat data to construct ourmeasures of professional experience at troubled firms.

Panel A of Table 2 shows summary statistics for our sample of managers.An average CEO is 52.8 years old, worked at her firm for 7.3 years, and hasover 50% of compensation as equity-based compensation. The vast majority(97.8%) of CEOs are male. An average CFO is slightly younger (47.1 yearsold), worked at her firm for 6.8 years, and has 43% of compensation as equitybased compensation. Also, 2.4% of CFOs are female. Further, 33.3% of theCEOs and 46.5% of the CFOs have an MBA degree. Panel A also confirmsthat the managers in our sample are not significantly different from the generalpopulation of managers at public firms. The only exception is CEO age—theCEOs in our sample are, on average, 2.4 years older.

1.3 Measures of past professional experienceWe study how work-related experiences throughout managers’ professionallives affect managers’ careers and corporate policies. Our primary focus ison realizations of poor outcomes—that is, experiences of bankruptcy anddistress.2 Our analysis is motivated by the psychology literature, which showsthat individual experiences influence decision-making (Nisbett and Ross 1980).

2 In later tests, we compare the impact of negative and positive professional experiences.

571



Dow

nloaded from



Table 2Managers

Panel A: Summary statistics

Variable Mean Median Std. dev. N. Obs. Mean full t-statistic forsample diff. in means

CEOsAge 52.783 53.000 8.123 35,464 50.414 2.318Female 0.022 0.000 0.147 35,601 0.020 0.511Tenure 7.271 5.000 7.064 35,601 7.336 0.194Equity-based compensation 0.512 0.407 0.942 35,581 0.485 1.227MBA degree 0.333 0.000 0.471 35,543 0.322 0.283

CFOsAge 47.069 47.000 7.123 29,555 44.773 0.671Female 0.024 0.000 0.152 29,961 0.025 0.135Tenure 6.848 5.000 6.907 29,214 6.682 0.448Equity-based compensation 0.428 0.366 0.785 28,937 0.399 0.705MBA degree 0.465 0.000 0.499 29,476 0.431 0.931

Panel B: Frequency of professional experience

Indicator CEO N. Obs. CFO N. Obs.

Professional experience (bankruptcy) 0.8% 35,601 0.6% 29,961Professional experience (bond ratings) 6.4% 31,205 6.1% 26,138Professional experience (cash flow shocks) 11.5% 33,366 11.1% 27,612Professional experience (stock returns) 11.2% 33,925 10.8% 28,003Professional experience (composite index) 23.8% 29,226 22.0% 25,242

Panel C: Correlation between measures of CEO professional experience

Bankruptcy Bond Cash flow Stock Compositeratings shocks returns index

Bankruptcy 1.000Bond ratings 0.562 1.000Cash flow shocks 0.106 0.426 1.000Stock returns 0.172 0.338 0.331 1.000Composite index 0.185 0.636 0.692 0.670 1.000

(continued)

More specifically, we build on the “hot stove” effect, studied by March (1996),Denrell and March (2001), and Denrell (2007), which implies a bias againstrisky alternatives to avoid actions that have led to poor outcomes. Our mainhypothesis, therefore, suggests that managers that experienced poor outcomesof bankruptcy or distress in the past subsequently implement more conservativecorporate policies.

In contrast to prior studies, we focus on professional experiences rather thanpersonal experiences such as military service and growing up during the GreatDepression. We do so because professional experiences are typically morefrequent and recent, and therefore may exert greater influence on decision-making. Further, they occur in a similar corporate setting and thus likelycomprise relevant experiences for shaping the CEO’s management style.Finally, they can occur throughout a CEO’s career, thus implying that herdecision-making may change and evolve over time. These attributes allowus to test several experimentally well-established behavioral hypotheses in thecontext of real-world decisions made by corporate executives. Specifically, we

572



Dow

nloaded from



Table 2Continued

Panel D: Additional details about employment history and professional experience

Variable Mean Median Std. dev. N. Obs.

All CEOsEmployment history: N years 21.358 20.000 5.629 35,601Employment history: N firms 3.871 3.000 3.880 35,601

CEOs with composite index = 1Employment history: N years 21.028 20.000 5.586 6,956Employment history: N firms 3.691 3.000 3.705 6,956N professional experiences 1.226 1.000 1.683 6,956N years since last experience 5.249 5.000 6.528 6,956First job experience 0.284 0.000 0.426 6,956

All CFOsEmployment history: N years 17.582 15.000 4.994 29,961Employment history: N firms 2.619 2.000 3.127 29,961

CFOs with composite index = 1Employment history: N years 17.264 15.000 4.774 5,553Employment history: N firms 2.586 2.000 3.037 5,553N professional experiences 1.188 1.000 1.532 5,553N years since last experience 5.048 6.000 6.104 5,553First job experience 0.236 0.000 0.410 5,553

This table presents information about the CEOs and CFOs in our sample. Panel A provides summary statisticsabout managers’ age, gender, tenure, equity-based compensation, and education. Panel B describes the measuresof managers’ professional experience at troubled firms. Panel C provides the sample-wide correlations betweenmeasures of CEO professional experience. Panel D provides additional details about managers’ employmenthistory and professional experiences. Professional experience (bankruptcy) is an indicator equal to 1 if themanager worked at a firm that filed for chapter 11. Professional experience (bond ratings) is an indicator equalto 1 if the manager worked at a firm that belonged to the lowest decile of Compustat firms based on annualchanges in credit ratings, provided that the rating was downgraded. Professional experience (cash flow shocks)is an indicator equal to 1 if the manager worked at a firm that belonged to the lowest decile of Compustatfirms based on annual changes in operating cash flow, provided that the change in cash flow was negative.Professional experience (stock returns) is an indicator equal to 1 if the manager worked at a firm that belongedto the lowest decile of Compustat firms based on annual stock returns, provided that the stock return wasnegative. Professional experience (composite index) is the maximum of the four professional experience variables:Professional experience (bankruptcy), Professional experience (bond ratings), Professional experience (cash flowshocks), Professional experience (stock returns). In all cases, we exclude past employment as the CEO of otherfirms. All variable definitions are given in the Appendix.

examine the impact of repetition in reinforcing an outcome, and of timing indetermining the saliency of an event; we also compare the impact of positiveversus negative experiences in influencing corporate policy. These tests arefeasible only because professional experiences occur throughout a manager’scareer, whereas prior research focuses on one particular and important personalexperience that likely occurs at a similar time for all affected managers.

To measure CEOs’professional experience, we track the employment historyof the CEO using data from ExecuComp and BoardEx to determine if the CEOwas previously employed at a troubled firm. We restrict our attention to previousemployment at other firms to disentangle CEO effects from firm effects. Tomitigate the concern that the CEO is chosen based on her experience in runningtroubled firms, we focus on professional experience in non-CEO roles, thoughin subsequent tests we consider experience as CEO of other firms to test thesaliency hypothesis. To further control for firm effects, our tests exclude firmsthat experienced difficulties themselves and control for firm-level, time-varying

573



Dow

nloaded from



determinants of corporate policies, as well as unobservable time-invariant firmeffects.

Our measures of professional experience are based on the full set ofinformation we have available for each manager.3 For robustness andcompleteness, we employ four measures of distress. The first measure is basedon bankruptcy filings. The bankruptcy data come from the Bankruptcy ResearchDatabase of Lynn LoPucki at UCLALaw School, which includes all bankruptcycases filed under Chapter 7 or Chapter 11 of the bankruptcy code, for firmsthat had assets worth $100 million or more (in 1980 dollars) and filed anannual report for a year ending not less than three years prior to the filingof the bankruptcy case. Managers who previously worked at a firm that filedfor bankruptcy during their employment are defined as having experiencedbankruptcy. A potential concern with the bankruptcy-based measure, however,is that bankruptcy filing is salient enough to exert a significant direct effecton a manager’s career that is unrelated to its impact on her decision-making.Moreover, as Panel B of Table 2 shows, bankruptcy experience is relativelyinfrequent, and only 0.8% of the CEOs in our sample experienced bankruptcyin the past. We therefore construct three additional measures.

The second measure is based on adverse shocks to a firm’s bond rating.We retrieve data on bond ratings for all industrial firms on Compustat, andsort all firm-year observations into annual deciles based on the change in bondratings. Each year, firms in the lowest decile are defined as distressed, as long astheir ratings were in fact downgraded.4 Related approaches for characterizingfinancial difficulties are used by Kashyap, Lamont, and Stein (1994), Gilchristand Himmelberg (1995), and Almeida, Campello, and Weisbach (2004). Theadvantage of this measure is that it gauges the market’s assessment of a firm’scredit quality. Managers who previously worked at a firm that was categorizedas distressed during their employment are defined as having past distressexperience. Panel B of Table 2 shows that 6.4% of the CEOs in our sampleexperienced difficulties according to this measure.

Our two remaining measures of experience focus on adverse shocks toa firm’s operating cash flows and stock returns, respectively. We define afirm’s operating cash flow as earnings before interest, taxes, depreciation, andamortization (EBITDA) divided by total book assets. We sort all industrial firmson Compustat into annual deciles based on the change in annual operating cashflow and categorize firms in the lowest decile each year as experiencing distress.Similarly, we calculate a firm’s annual stock return, sort all industrial firms onCompustat into annual deciles based on their stock returns, and categorize

3 We exclude any experience that occurs at a utility company, as this may be due to regulatory restraint rather thandistress.

4 Firms whose rating did not change are not classified as distressed. To address missing ratings, we determine ifthe firm has had a rating in the previous year. If it did, and currently has debt outstanding and no rating, it isclassified as distressed.

574



Dow

nloaded from



firms in the lowest decile each year as experiencing distress. To ensure thatthese measures capture poor performance and therefore distress, we requirethat the firms labeled as distressed by this measure have negative stock returnsor a decline in their cash flows. In our sample, the average stock return (cashflow shock) experienced by firms in the lowest decile is −21.5% (−18.5%).5

We define managers who previously worked at a firm that was in the lowestdecile during their employment as having past experience of distress. Panel Bof Table 2 shows that 11.5% (11.2%) of the CEOs in our sample experienceddistress according to the cash flow–based (stock return–based) measure.

In addition, we also create a composite index that is equal to one if any ofthe above experience measures is equal to one. Panel B of Table 2 shows that23.8% of the CEOs in our sample experienced distress according to at least oneof the experience measures.

Panel C of Table 2 reports the sample-wide correlations between themeasures of professional experience. The estimates show that, as expected, allmeasures of distress are positively correlated. These measures, however, areonly imperfectly correlated, suggesting that they capture different dimensionsof distress and motivating the use of the composite index. The correlationbetween the measures ranges from 0.11 to 0.56, with bankruptcy and cash flowshocks having the lowest and bankruptcy and credit ratings having the highestcorrelations. This approach of using multiple variables to capture distress isconsistent with default prediction models such as Bharath and Shumway (2008).

Panel D of Table 2 provides additional details about the employment historyand professional experiences of CEOs. We observe 21.4 years of employment,in 3.9 firms, for the average CEO in our sample. Similarly, for CEOs thatexperienced distress according to any of our measures, we observe 21.0 yearsof employment in 3.7 firms. Focusing on the subset of CEOs that experienceddistress, the average CEO in this subsample experienced difficulties 1.2 times,thus repeated occurrences are less common. On average, the number of yearssince the last experience is 5.2 years. Further, 28.4% of the CEOs experienceddifficulties in their first position, reflecting a close to even distribution over theapproximately 3.7 firms in the manager’s employment history. In Section 3,we test the recency hypothesis by comparing the effects of recent and distantexperiences on a firm’s policy and the saliency hypothesis by comparing theimpact of repeated experiences.

Our main focus is on the professional experience of the CEO, since theultimate responsibility for the firm’s financial and investment strategies restswith the CEO. However, we also study the experience of the CFO, who mayassist the CEO with financing decisions. We therefore create each measure ofexperience for CFOs, described in Panels B and D of Table 2.

5 In untabulated tests, we also reconstruct these measures requiring a shock of at least −10% and find similarresults.

575



Dow

nloaded from



2. CEO Professional Experience and Corporate Policies

2.1 Does professional experience affect corporate policy?Table 3 presents results of panel regressions of leverage (PanelA), cash holdings(Panel B), and capital expenditures (Panel C) on the professional experienceof the CEO and a relevant set of firm-level determinants of each of thesepolicies. We test the hypothesis that the professional experience of the CEOaffects corporate policy against the null hypothesis that CEOs have no effect oncorporate policy, and firm characteristics are the only determinant of leverage,cash, and investment, regardless of who the CEO is. We therefore also controlfor unobservable firm characteristics and market-wide effects by including firmand year fixed effects. We cluster the standard errors at the firm level.

To distinguish the effect of professional experience, we control for otherCEO traits that may affect corporate policy. First, we control for CEO agebecause an older CEO has had more time to be exposed to different firmenvironments and thus may be more likely to have experienced distress.Further, Bertrand and Schoar (2003) show that CEO age has a significanteffect on corporate policies. Second, we control for the gender of the CEObecause prior studies show that men are more likely to take risk than females.Barber and Odean (2001) and Weber, Blais, and Betz (2002), for example,show that financial risk-taking differs by gender. More broadly, Byrnes, Miller,and Schafer (1999) and Eckel and Grossman (2008) provide a review of theliterature on differential risk taking by gender. Third, we control for CEOs’financial education using an MBA indicator that equals one if the CEO has anMBA degree. Financial literacy may lead managers to rely more on externalfinance in lieu of internal cash savings. Lastly, we control for CEO incentivesusing equity-based compensation, defined as the fraction of total compensationpaid through stock and option grant, which may contractually mitigate theeffects of CEO preferences or the influences of past experiences, and mayalso affect the incentives of the CEO to take risk. In subsequent tests, wewill also control for personal experience and other managerial traits such asoverconfidence.

Panel A of Table 3 presents the regression estimates for leverage, measuredas short-term plus long-term debt divided by total assets. Following Frank andGoyal (2009), our firm-level controls include size (log assets), the market-to-book ratio as a measure of investment opportunities, profitability, thetangibility of assets (the ratio of fixed to total assets), and industry leverage.6

To control for unobservable firm characteristics that may be correlated withthe CEO’s professional experience and with the firm’s leverage policy, theregressions also include firm fixed effects.

6 Frank and Goyal (2009) also control for inflation, which, in our empirical model, is absorbed by year fixedeffects.

576



Dow

nloaded from



Table 3CEOs’ professional experience and corporate policy

Panel A: Leverage

Measure of Professional Professional Professional Professional Compositeprofessional experience experience experience experience index ofexperience (bankruptcy) (bond (cash flow (stock professional

ratings) shocks) returns) experience

Model (1) (2) (3) (4) (5)

Professional experience −0.024∗∗ −0.028∗∗ −0.015∗∗∗ −0.017∗∗∗ −0.021∗∗∗[0.011] [0.013] [0.005] [0.005] [0.007]

Industry leverage 0.465∗∗∗ 0.460∗∗∗ 0.459∗∗∗ 0.462∗∗∗ 0.461∗∗∗[0.026] [0.026] [0.025] [0.025] [0.026]

Tangibility 0.139∗∗∗ 0.144∗∗∗ 0.140∗∗∗ 0.142∗∗∗ 0.144∗∗∗[0.021] [0.021] [0.021] [0.022] [0.021]

Market-to-book 0.003 0.004 0.003 0.004 0.004[0.004] [0.004] [0.004] [0.004] [0.003]

Profitability −0.163∗∗∗ −0.165∗∗∗ −0.166∗∗∗ −0.165∗∗∗ −0.163∗∗∗[0.046] [0.047] [0.048] [0.048] [0.046]

Size 0.017∗∗∗ 0.017∗∗∗ 0.017∗∗∗ 0.017∗∗∗ 0.017∗∗∗[0.005] [0.005] [0.005] [0.005] [0.005]

CEO age (/100) 0.001 0.009 0.009 0.010 0.007[0.036] [0.035] [0.038] [0.036] [0.036]

Female −0.009∗ −0.011∗∗ −0.010∗ −0.008∗ −0.009∗[0.005] [0.005] [0.006] [0.004] [0.005]

MBA degree 0.008∗∗ 0.009∗∗∗ 0.010∗∗∗ 0.008∗∗ 0.009∗∗∗[0.003] [0.003] [0.003] [0.003] [0.003]

Equity-based compensation 0.143∗∗∗ 0.144∗∗∗ 0.146∗∗∗ 0.149∗∗∗ 0.144∗∗∗[0.037] [0.038] [0.041] [0.042] [0.037]

Year fixed effects Yes Yes Yes Yes YesFirm fixed effects Yes Yes Yes Yes Yes

N. Obs. 34,954 30,712 32,683 33,018 29,226R2 0.629 0.627 0.624 0.631 0.635

Panel B: Cash holdings

Professional experience 0.011∗∗∗ 0.017∗∗ 0.026∗∗∗ 0.019∗∗∗ 0.018∗∗∗[0.004] [0.005] [0.008] [0.004] [0.006]

Industry cash flow volatility 0.192∗∗∗ 0.192∗∗∗ 0.191∗∗∗ 0.193∗∗∗ 0.191∗∗∗[0.060] [0.060] [0.060] [0.061] [0.060]

Cash flow 0.068∗∗∗ 0.067∗∗∗ 0.067∗∗∗ 0.068∗∗∗ 0.068∗∗∗[0.010] [0.010] [0.010] [0.010] [0.010]

Market-to-book 0.011∗∗∗ 0.011∗∗∗ 0.011∗∗∗ 0.012∗∗∗ 0.011∗∗∗[0.001] [0.001] [0.001] [0.001] [0.001]

Credit ratings −0.024 −0.024 −0.024 −0.024 −0.027[0.017] [0.017] [0.017] [0.017] [0.017]

Size −0.025∗∗∗ −0.025∗∗∗ −0.025∗∗∗ −0.025∗∗∗ −0.025∗∗∗[0.002] [0.002] [0.002] [0.002] [0.002]

CEO age (/100) −0.093∗∗∗ −0.095∗∗∗ −0.094∗∗∗ −0.097∗∗∗ −0.094∗∗∗[0.012] [0.012] [0.012] [0.010] [0.012]

Female 0.021∗∗ 0.021∗∗ 0.019∗∗ 0.020∗∗ 0.020∗∗[0.009] [0.009] [0.009] [0.008] [0.008]

MBA degree −0.002 −0.003 −0.003 −0.003 −0.003[0.002] [0.002] [0.002] [0.003] [0.003]

Equity-based compensation −0.012∗∗∗ −0.012∗∗∗ −0.012∗∗∗ −0.012∗∗∗ −0.012∗∗∗[0.003] [0.003] [0.003] [0.003] [0.003]


N. Obs. 34,954 30,712 32,683 33,018 29,226R2 0.842 0.842 0.843 0.843 0.846

(continued)

577



Dow

nloaded from



Table 3Continued

Panel C: Capital expenditures

Measure of Professional Professional Professional Professional Compositeprofessional experience experience experience experience index ofexperience (bankruptcy) (bond (cash flow (stock professional

ratings) shocks) returns) experience

Model (1) (2) (3) (4) (5)

Professional experience −0.006∗∗ −0.004∗∗ −0.003∗ −0.005∗∗∗ −0.005∗∗∗[0.003] [0.002] [0.002] [0.002] [0.002]

Market-to-book 0.002∗∗∗ 0.002∗∗∗ 0.002∗∗∗ 0.002∗∗∗ 0.002∗∗∗[0.001] [0.001] [0.001] [0.001] [0.001]

Cash flow 0.009∗∗ 0.009∗∗ 0.008∗∗ 0.009∗∗ 0.009∗∗[0.003] [0.003] [0.004] [0.004] [0.003]

CEO age (/100) 0.001 −0.003 −0.003 <0.001 −0.001[0.005] [0.005] [0.006] [0.005] [0.006]

Female −0.001 −0.001 −0.002 −0.001 −0.001[0.003] [0.003] [0.003] [0.003] [0.003]

MBA degree 0.003∗∗ 0.002∗∗ 0.002∗ 0.002∗∗ 0.002∗∗[0.001] [0.001] [0.001] [0.001] [0.001]

Equity-based compensation 0.003∗ 0.003∗ 0.003∗ 0.003∗ 0.003∗[0.002] [0.002] [0.002] [0.002] [0.002]


N. Obs. 34,954 30,712 32,683 33,018 29,226R2 0.662 0.660 0.663 0.661 0.665

This table presents evidence on the relation between the professional experience of the CEO and firm–levelfinancial policies. In panel A, the dependent variable is the ratio of short-term plus long-term debt to book assets.In panel B, the dependent variable is the ratio of cash reserves to book assets. In panel C, the dependent variableis the ratio of capital expenditure to book assets. The key variable of interest is Professional experience, definedas an indicator equal to 1 if the CEO worked at another firm that experienced difficulties. We use four measure ofdifficulties based on bankruptcy filings, bond ratings downgrades, adverse cash flow shocks, and adverse shocksto the firm’s annual stock return. We also calculate a composite index of Professional experience, defined as themaximum of these measures. All variable definitions are given in the Appendix. All the regressions include yearand firm fixed effects. The standard errors (in brackets) are heteroscedasticity consistent and clustered at the firmlevel. Significance levels are indicated as follows: * = 10%, ** = 5%, *** = 1%.

The results in Panel A indicate a negative relation between leverageand CEOs’ professional experience of distress, as captured by the variableProfessional experience. This relation is consistently negative and significantacross all measures of distress. The economic magnitudes are substantial andcomparable in size across all columns: experiencing distress is associated with a1.5- to 2.8-percentage-point decline in the firm’s leverage ratio. For a manageroverseeing a firm with mean characteristics, this effect is associated with areduction of 7–13% in debt.

An analysis of the other control variables shows that, consistent with Frankand Goyal (2009), there is a positive relation between leverage and size,tangibility, and industry leverage, and a negative relation between leverageand profitability. We also find that male CEOs, CEOs with an MBA degree,and CEOs with more equity-based compensation have higher leverage ratios.These findings are consistent with greater risk-taking by male CEOs and agreater reliance on external finance due to financial literacy following an MBAdegree. Finally, these results suggest that risk-taking incentives resulting froma CEO’s compensation structure predict debt.

578



Dow

nloaded from



Panel B of Table 3 presents the results for a firm’s cash savings policy,measured as cash and short-term assets divided by total assets. The regressionsinclude firm-level proxies for the precautionary savings motive, the predomi-nant motivation to hold cash based on Keynes (1936) and Miller and Orr (1966).The empirical predictions of this theory suggest that firms with higher cashflow volatility, better investment opportunities, and lower credit ratings willhold more cash. Opler et al. (1999), Almeida, Campello, and Weisbach (2004),Bates, Kahle, and Stulz (2009), Lins, Servaes, and Tufano (2010), Campello etal. (2011), and others all find empirical support for the precautionary savingsmotive. We also control for the firm’s size because prior research (e.g., Opleret al. 1999) shows there are economies of scale in cash policy.

The empirical results in Panel B show a positive relation between cashsavings and CEOs’ professional experience. This relation is consistentlysignificant across all measures of distress. The economic magnitudes arenontrivial: professional experience of distress is associated with a 1.1- to 2.6-percentage-point increase in the firm’s cash savings. For a manager overseeinga firm with mean characteristics, this effect is associated with an increase of5.0–11.8% in cash holdings.

As expected, an analysis of the other control variables suggests that firms withhigher cash flow volatility, cash flows, and market-to-book ratios (our proxyfor investment opportunities), as well as smaller firms, hold more cash. Theseresults are consistent with precautionary savings motive and with previousresearch (e.g., Opler et al. 1999). We also find that younger CEOs, femaleCEOs, and CEOs with less equity-based compensation tend to hold more cash.Under the view that cash is negative debt, these results are uniformly consistentwith the findings in Panel A.

Panel C of Table 3 analyzes the effect of a CEO’s professional experienceon a firm’s investment policy, as measured by the ratio of capital expendituresto book assets. The regressions control for firm-level investment opportunities,as measured by the market-to-book ratio, and cash flows. The estimatessuggest that capital expenditures are negatively associated with professionalexperiences of distress. The effects are consistently negative and significantacross all measures. The economic magnitudes show that professionalexperience of distress is associated with a 0.3- to 0.6-percentage-point declinein the firm’s capital expenditures. For a manager overseeing a firm with meancharacteristics, this effect is associated with a reduction of 5–10% in capitalexpenditures. An analysis of the control variables indicates that firms withhigher investment opportunities and more cash flow invest more in capitalexpenditures. We also find modest evidence that CEOs with an MBA or CEOsthat have more equity-based compensation invest more.

Taken together, our evidence shows that a manager’s experience at distressedfirms captures a significant effect beyond the firm-, industry-, and market-level determinants of corporate policy, controlling for a wide range of CEOcharacteristics that may influence her incentives and preferences.

579



Dow

nloaded from



2.2 CEO turnoverThe above analysis uses the full sample of firms for which we havefull employment data and includes firm fixed effects to control for time-invariant unobservable characteristics that could be correlated with the CEO’sprofessional experience and the firm’s financial and investment policy. Tofurther capture the causal effect of the CEO’s professional experience oncorporate policy, we focus on CEO turnovers, a setting in which the firmexperiences a shock to the experience of its CEO. An important issue in thisanalysis is that some CEO turnovers may be driven by a change in the firm’sinvestment opportunities, the poor performance of the departing CEO, or otherpotential determinants of the firm’s corporate policy, which may confound ourtests. The main concern is that the CEO’s professional experience is irrelevantbecause any new CEO (regardless of her experience) would have implementedthe new conservative policies given the change in firm characteristics.

To mitigate this concern, we use a subset of CEO turnovers that are unlikely tobe associated with managerial performance or a change in the firm’s conditions.In particular, we focus on the CEO turnovers that meet one of the followingterms:

(i) The departing CEO dies, departs due to health-related reasons, or is atleast 60 years old.

(ii) The CEO turnover is part of the firm’s succession plan.

These turnovers occur either unexpectedly or as part of the firm’s managementsuccession plan, and hence are unlikely to be caused by changes in the firm’sconditions that may warrant a change in its corporate policies. To classifyCEO turnovers, we follow the approach of Huson, Parrino, and Starks (2001)and read the article in the Wall Street Journal and the firm’s press releaseassociated with the CEO change for the specific reasons given for the turnover.We also collect information on the CEO’s age at the time of the turnover fromBoardEx. We find that 67.3% of CEO turnovers in our sample satisfy thesecriteria, consistent with the frequency of voluntary CEO turnovers estimatedin the literature (e.g., Yermack 2006; Falato, Li, and Milbourn Forthcoming;Jenter and Kanaan 2015).

These criteria focus our analysis on exogenous CEO turnover. However, toensure that we capture truly exogenous turnovers we also employ a stricterdefinition. To do this, for all CEOs that meet one of these criteria, we collectadditional data on the details of their succession plan and refine our definitionof exogenous turnover to ensure that the CEO either (i) departed as part of asuccession where the date of departure was announced in public at least oneyear prior to her departure, (ii) departed due to health reasons (including death),or (iii) retired at 65 or later.7 By purging the announcement of the succession

7 In subsequent tests, we further refine our definition to require a retirement age of 70 years old and find similarresults.

580



Dow

nloaded from



plan’s date of departure at least one year back, we ensure that the turnover is nota result of changes in firm attributes during the year preceding the turnover.8

By increasing the retirement date to 65 or even 70, it becomes more likely thatthe manager departs due to exogenous, age-related considerations, rather thanfor company- or policy-related reasons.

Another way to mitigate this concern is to focus on the subset of internalturnovers, in which the new CEO was already an employee of the firm beforeshe was appointed as CEO. In this setting, the choice of the CEO is less likelyto reflect major changes due to changes in underlying conditions.

Table 4 reports estimates from the following two regression models for thethree sets of CEO turnovers (exogenous succession/health/age-related CEOturnovers, more strictly defined exogenous turnovers, and internal turnovers):(ii) firm fixed effects panel regressions that only include firms whose CEOsturned over during our sample period; and (ii) difference regressions thatcompare the financial policy two years prior to the turnover of the CEO and twoyears after the turnover. The dependent variable in the fixed effects model is thefirm’s debt, cash, or investment, deflated by book assets. To address the concernthat the effects are driven by changes in the denominator (book assets) ratherthan the numerator (corporate policy), we estimate the difference regressionsusing the percentage change in corporate policy, which is not deflated by bookassets.

In both regression models, we exclude the three-year window surroundingthe turnover (i.e., we exclude the turnover year and the year that immediatelyprecedes and immediately follows the turnover year) to mitigate the potentialeffects of the upheaval surrounding CEO turnover. For brevity, we present onlythe analysis using the composite index of professional experience. In columns1–3 of each panel, the key independent variable is the composite index of theprofessional experience of the newly appointed CEO. In columns 4–6, the keyindependent variable is the change in the index of professional experience ofthe CEO resulting from the turnover.9 As in the previous tables, all of theregressions include CEO-, firm-, industry-, and market-level determinants ofeach of the corporate policies, which are not presented to conserve space.Similar to the index of professional experience, columns 1–3 include the controlvariables in levels and columns 4–6 include them in differences from two yearsbefore the turnover to two years after the turnover. As before, the standarderrors are clustered at the firm level.

8 Based on the stricter definition of CEO turnovers, we identify a total of 537 turnovers. Of these 537 turnovers,128 are due to a succession plan whose date of departure is announced at least a year in advance, 96 are dueto health issues, and the remaining 313 are retirements at the age of 65 or older. In unreported tests, we obtainsimilar results after excluding succession plans altogether from the set of exogenous turnovers.

9 The change in the CEO’s professional experience equals 1 if the new CEO experienced difficulties and thedeparting CEO did not, 0 if both CEOs either experienced or did not experience difficulties, and −1 if the newCEO did not experience difficulties whereas the departing CEO did.

581



Dow

nloaded from



Table 4Exogenous CEO turnovers

Panel A: Exogenous CEO turnovers

Specification Firm fixed effects Changes around CEO turnovers

Dependent Leverage Cash Capital � �Cash �Capitalvariable holdings expenditure Leverage holdings expenditure

Model (1) (2) (3) (4) (5) (6)

Professional experience −0.009∗∗ 0.028∗∗∗ −0.005∗∗ −0.016∗∗ 0.025∗∗ −0.006[0.004] [0.009] [0.002] [0.007] [0.012] [0.004]

Controls Yes Yes Yes Yes Yes YesYear fixed effects Yes Yes Yes Yes Yes Yes

N. Obs. 8,462 8,462 8,462 976 976 976R2 0.149 0.811 0.623 0.136 0.048 0.019

Panel B: Exogenous CEO turnovers: Stricter definition

Professional experience −0.015∗∗∗ 0.020∗∗ −0.004∗∗ −0.016∗∗ 0.018∗∗∗ −0.005∗[0.006] [0.009] [0.002] [0.007] [0.006] [0.003]


N. Obs. 6,377 6,377 6,377 537 537 537R2 0.603 0.828 0.618 0.185 0.092 0.028

Panel C: Internal CEO turnovers

Professional experience −0.008∗∗ 0.026∗∗ −0.004∗∗ −0.014∗∗ 0.025∗∗∗ −0.009∗[0.004] [0.012] [0.002] [0.007] [0.006] [0.005]


N. Obs. 10,081 10,081 10,081 1,228 1,228 1,228R2 0.163 0.828 0.641 0.135 0.053 0.018

This table presents estimates from fixed effects and first-difference regressions surrounding exogenous CEOturnovers. Professional experience is measured by the composite Index of professional experience. Panel Acorresponds to exogenous turnovers in which the CEO departed as part of a succession plan, departed due tohealth reasons (including deaths), or retired at the age of 60 or older. Panel B uses a stricter definition of exogenousturnovers in which the CEO departed as part of a succession plan whose date of departure was announced at leastone year ahead, due to health reasons (including deaths), or retired at the age of 65 or later. Panel C includesinternal turnovers in which the new CEO came from inside the firm. All variable definitions are given in theAppendix. All the regressions include the same controls as in Table 3, which are not shown. The standard errors(in brackets) are heteroscedasticity consistent and clustered at the firm level. Significance levels are indicated asfollows: * = 10%, ** = 5%, *** = 1%.

The results across all panels of Table 4 show that when a new CEO thatexperienced distress is appointed as CEO, the firm reduces its debt, increases itscash savings, and cuts its investment in capital expenditures. These results holdacross both regression models and for all subsets of CEO turnovers, and theyare statistically significant at conventional levels in all cases except Column 6of Panel A.

The economic magnitude of the impact of professional experience in theturnover sample is similar across samples and models and is nontrivial: basedon the sample of exogenous CEO turnovers (Panel A) using the firm fixedeffects model, a newly appointed CEO who experienced distress reduces debtby 0.9 percentage points, increases cash holdings by 2.8 percentage points,and decreases capital expenditures by 0.5 percentage points. For a manager

582



Dow

nloaded from



overseeing a firm with mean characteristics, this effect is associated with areduction of 4% in debt, an increase of 13% in cash holdings, and a decreaseof 8% in investment.

Importantly, our empirical design and results exploit cross-sectional variationwithin the subset of exogenous CEO turnovers. This empirical design differsfrom Fee, Hadlock, and Pierce (2013), who show that, on average, exogenousCEO turnovers are not accompanied by significant changes in a firm’s financingor investment policies. In unreported tests, we follow Fee, Hadlock, and Pierce(2013) and compare policy changes surrounding exogenous CEO turnovers andpolicy changes in matching firms that did not undergo CEO turnover. Consistentwith Fee, Hadlock, and Pierce (2013), we find that corporate leverage, cash,and investment policies do not change, on average, following exogenous CEOturnovers.

Taken together, our findings indicate that although exogenous CEO turnoversare not accompanied, on average, by significant policy changes, variation in theCEO’s professional experience within the subset of exogenous turnovers doesaffect corporate policy. Specifically, our results imply that exogenous turnoversthat result in a change in the experience of the CEO lead to substantial changesin corporate policies relative to exogenous turnovers that do not result in achange in the experience of the CEO.

For the remainder of the paper, we estimate our tests in the sample of stricterexogenous CEO turnovers, using the firm fixed effects model. As noted above,this specification controls for time-invariant firm characteristics that might becorrelated with professional experience and corporate policy, as well as changesin firm characteristics that may trigger CEO turnover and be correlated withthe CEO’s professional experience and corporate policy.

2.3 Personal experience and overconfidenceThus far, the analysis controlled for CEO characteristics, including gender,education, compensation, and age. In addition to these traits, a number of papersexamine the impact of CEOs’ personal experiences on corporate decision-making. Malmendier and Tate (2005) and Malmendier, Tate, and Yan (2011)show that early personal experiences of managers, such as growing up duringthe Great Depression and military service, affect corporate leverage andinvestment. Additionally, several papers seek to measure CEO conservatism oroverconfidence directly, and explore its impact on corporate policy. Malmendierand Tate (2005), Cronqvist, Makhija, and Yonkers (2012), Cain and McKeon(Forthcoming), and Hutton, Jiang, and Kumar (2014) measure conservatism oroverconfidence based on the CEO’s option-exercising behavior, and whethershe has high levels of personal debt, holds a pilot’s license, or is a Republican.The general finding is that more conservative managers implement moreconservative corporate policies, and, conversely, overconfident managerspursue more aggressive policies.

583



Dow

nloaded from



To compare the impact of professional experience with factors consideredin prior work, Table 5 repeats the analysis controlling for a manager being inthe military or being born during the Depression era (Panel A), and for twomeasures of overconfidence (Panel B).

Following Malmendier, Tate, and Yan (2011), we define an option-basedoverconfidence indicator that equals one for CEOs who, at any point duringthe sample period, hold an option even though the option is at least 40% inthe money. We define a second overconfidence indicator that compares thenumber of media articles using the terms (i) “confident” or “confidence” or(ii) “optimistic” or “optimism” to the number of past articles that portray theCEO as (iii) not “confident,” (iv) not “optimistic,” or (v) “reliable,” “cautious,”“conservative,” “practical,” “frugal,” or “steady.” We set the indicator equalto one if (i) + (ii) >(iii) + (iv) + (v). We address possible bias due todifferential coverage by controlling for the total number of articles in theselected publications.

For brevity, we present only the analysis using the composite index ofprofessional experience. The set of controls is similar to that in Table 3, withthe exception of the Great Depression analysis, which excludes the controlvariable CEO age since it is highly correlated with growing up during theGreat Depression.

Panels A and B show that professional experience has an important effecton corporate policy after controlling for CEOs’personal experiences and traits.The effect of professional experience remains both significant and of similarimportance as in the previous tables. Consistent with prior studies, Panel Ashows that growing up during the Depression or military experience influencesboth financial and investment policies. In all specifications, professionalexperience has a greater impact than personal experience. Panel B demonstratesthe importance of overconfidence in determining corporate policy.

Because professional experiences occur throughout a manager’s career, theymay attenuate or enhance the impact of her earlier personal experiences oroverconfidence. To test this, in Panel C we interact the manager’s professionalexperience with these measures. We find that professional experience attenuatesthe impact of overconfidence and enhances the impact of Depression-eraexperience. Based on column 1 of Panel C, professional experience decreasesleverage by 2.6 percentage points, depression experience decreases leverage byanother 0.6 percentage points, and their interaction further decreases leverageby 0.6 percentage points. As Panel C indicates, the interaction betweenprofessional experience and Depression-era experience also significantly influ-ences cash and investment policies, whereas the interaction between militaryservice and professional experience is insignificant. Finally, overconfidenceand professional experience have significant interactions on all three policies.These findings are consistent with two possible interpretations: (i) the CEO’soverconfidence weakens after experiencing distress, or (ii) the effect of the

584



Dow

nloaded from



Table 5Professional Experience, Personal Experience, and Overconfidence

Panel A: Personal experience

Dependent variable Leverage Cash Capital expenditure

Model (1) (2) (3) (4) (5) (6)

Professional experience −0.017∗∗∗ −0.016∗∗∗ 0.019∗∗∗ 0.019∗∗∗ −0.004∗∗ −0.004∗∗[0.005] [0.005] [0.006] [0.006] [0.002] [0.002]

Military experience 0.012∗∗ −0.010∗∗ 0.002∗[0.006] [0.004] [0.001]

Depression experience −0.011∗∗ 0.012∗∗ −0.002∗[0.005] [0.005] [0.001]

Controls Yes Yes Yes Yes Yes YesYear fixed effects Yes Yes Yes Yes Yes YesFirm fixed effects Yes Yes Yes Yes Yes Yes

N. Obs. 5,165 6,377 5,165 6,377 5,165 6,377R2 0.611 0.608 0.834 0.833 0.622 0.625

Panel B: Overconfidence


Overconfidence (options) 0.023∗∗ −0.010∗∗ 0.002∗[0.010] [0.004] [0.001]

Overconfidence (media) 0.017∗∗ −0.014∗∗ 0.002[0.007] [0.006] [0.002]


N. Obs. 3,858 4,251 3,858 4,251 3,858 4,251R2 0.594 0.597 0.818 0.824 0.617 0.622

Panel C: Interactions


Military experience 0.014∗∗ −0.008∗∗ 0.002∗[0.005] [0.003] [0.001]

Military experience × −0.001 0.003 −0.001Professional experience [0.003] [0.002] [0.002]

Depression experience −0.006∗∗ 0.009∗∗ −0.002∗[0.003] [0.004] [0.001]

Depression experience × −0.006∗∗ 0.005∗∗ −0.002∗Professional experience [0.003] [0.002] [0.001]

Overconfidence (options) 0.019∗∗ −0.016∗∗ 0.003∗∗[0.008] [0.007] [0.001]

Overconfidence (options) × −0.007∗∗ 0.005∗∗ −0.002∗Professional experience [0.003] [0.002] [0.001]


N. Obs. 5,165 3,858 5,165 3,858 5,165 3,858R2 0.612 0.598 0.836 0.820 0.629 0.617

This table presents evidence on the relation between the professional experience, personal experience, andoverconfidence of the CEO. Professional experience is measured by the composite Index of professional experience.Panels A and B estimate the effect of professional experience on corporate policy controlling for the CEO’s militaryexperience (an indicator that equals 1 if the CEO served in the military), Depression experience (an indicatorthat equals 1 if the CEO grew up during the Great Depression), and overconfidence (based on option exercisingbehavior and media coverage, see the Appendix for a detailed definition). Panel C estimates the interaction effectsof professional experience of distress and personal experience or overconfidence. All variable definitions are givenin the Appendix. The regressions are estimated in a sample of exogenous turnovers in which the CEO departed aspart of a succession plan (whose date of departure was announced at least one year ahead), departed due to healthreasons (including deaths), or retired at the age of 65 or later. All the regressions include year and firm fixed effects.The standard errors (in brackets) are heteroscedasticity consistent and clustered at the firm level. Significance levelsare indicated as follows: * = 10%, ** = 5%, *** = 1%.

585



Dow

nloaded from



CEO’s overconfidence, rather than her overconfidence itself, weakens whenshe experiences distress during her career.

3. Variation in Professional Experience

The above analysis shows that, on average, professional experience has acausal effect on the financial and investment policies of a firm, controllingfor personal experiences and managerial traits. Thus far, we treat allprofessional experiences equally. However, professional experiences can varyin their timing, frequency, and saliency. In contrast, existing studies examinepersonal experiences and managerial traits that typically do not vary alongthese dimensions. We therefore exploit the variation in these attributesacross professional experiences to test several experimentally well-establishedbehavioral phenomena that have not been studied before in the context ofreal-world decisions made by corporate executives.

3.1 Frequency and importance of professional experienceProfessional experiences can be frequent, be recent, and occur during salientperiods of managers’ careers. Several theoretical and experimental studiessuggest that these factors may influence the impact of experience. Tverskyand Kahneman (1973, 1974) and Bordola, Gennaioli, and Shleifer (2012 a,b)show that individuals infer the frequency or probability of an event based onits saliency.

In Panel A of Table 6, we redefine the index of professional experienceto include only experiences that occurred when the manager was the CEO(columns 1–3) or a top-five executive (columns 4–6) at another firm. In contrast,the earlier analysis in this paper excluded experience as a CEO, and usedexperience in all non-CEO roles, including lower-rank positions outside thetop-five executive suite. Compared with the earlier estimates (reported inTable 4), professional experience as the CEO or a top-five executive has a moredramatic impact on corporate policies. For example, professional experienceas a CEO (top-five executive) results in a 3.5- (2.7-) percentage-point decreasein debt, compared with a 1.5-percentage-point decline associated with non-CEO experience. These findings support the hypothesis that managers inferthe probability of an event based on its saliency and are thus more affected byevents that occur during important times in their career.

Another interpretation of these results is that the board of directors selectsCEOs with salient experiences of distress to implement more conservativecorporate policies. Under this view, the board is likely to have anticipatedthe consequences of salient distress experiences, which took place when theincoming CEO was a CEO or a top-five executive at a different firm. Thus, thedramatic impact of salient experiences on corporate policies might reflect notonly choices made by the CEO, but also the selection of the CEO by the boardof directors to implement certain corporate policies.

586



Dow

nloaded from



Table 6Variation in experience

Panel A: Experience as a top executive

Specification Experience as CEO Experience as a top-five executive

Dependent Leverage Cash Capital Leverage Cash Capitalvariable holdings expenditure holdings expenditure

Model (1) (2) (3) (4) (5) (6)

Professional experience −0.035∗∗ 0.027∗∗∗ −0.007∗∗ −0.027∗∗∗ 0.023∗∗∗ −0.006∗∗[0.011] [0.005] [0.003] [0.002] [0.005] [0.002]


N. Obs. 5,394 5,394 5,394 5,626 5,626 5,626R2 0.632 0.819 0.648 0.641 0.805 0.644

Panel B: Number of experiences

Specification Single experience Multiple experiences


Model (1) (2) (3) (4) (5) (6)

Professional experience −0.012∗∗∗ 0.017∗∗ −0.003∗∗ −0.023∗∗∗ 0.026∗∗∗ −0.005∗∗∗[0.004] [0.008] [0.001] [0.007] [0.008] [0.002]


N. Obs. 5,893 5,893 5,893 5,394 5,394 5,394R2 0.607 0.825 0.617 0.602 0.834 0.622

Panel C: Recency

Specification Recent professional Distant professionalexperience experience


Model (1) (2) (3) (4) (5) (6)

Professional experience −0.019∗∗∗ 0.024∗∗ −0.005∗∗ −0.012∗∗ 0.015∗∗ −0.003∗[0.006] [0.009] [0.002] [0.005] [0.007] [0.002]


N. Obs. 5,612 5,612 5,612 5,612 5,612 5,612R2 0.608 0.831 0.616 0.599 0.824 0.603

(continued)

We also examine the saliency hypothesis by investigating whether repeatedexperience influences decision-making more than if it occurs only once. Ourprimary measures of experience require a manager to have only one exposureto distress. However, many managers experience multiple years of distress. Inour sample, 18.2% of the managers who experienced distress experienced themmore than once. To determine whether repeated experiences have a strongereffect, Panel B of Table 6 compares the impact of having one year versusmultiple years of professional experience. Again, for brevity, we present only

587



Dow

nloaded from



Table 6Continued

Panel D: Positive experience

Specification Positive Distress followed by aexperience positive experience


Model (1) (2) (3) (4) (5) (6)

Professional experience 0.003 −0.001 <0.001 −0.013∗∗∗ 0.019∗∗∗ −0.004∗∗[0.008] [0.005] [<0.001] [0.003] [0.004] [0.002]


N. Obs. 6,377 6,377 6,377 5,208 5,208 5,208R2 0.609 0.802 0.598 0.641 0.805 0.644

This table presents evidence on the relation between the professional experience of the CEO and corporate policies(leverage, cash holdings, and capital expenditures). Professional experience is measured by the composite Indexof professional experience. Panel A extends the analysis to consider the effect of experiencing distress as theCEO or a top-five executive. Panel B considers the effect of the number of experiences. Panel C investigatesthe timing of the professional experience. Recent (Distant) experiences occurred more than six years (less thansix years) prior to the start of the CEO’s current position. (six years is the median number of years since aCEO experienced distress. Panel D considers the effect of extreme positive professional experience, definedanalogously to professional experience of distress. The regressions are estimated in a sample of exogenousturnovers in which the CEO departed as part of a succession plan (whose date of departure was announced atleast one year ahead), departed due to health reasons (including deaths), or retired at the age of 65 or later.All variable definitions are given in the Appendix. The standard errors (in brackets) are heteroskedasticityconsistent and clustered at the firm level. Significance levels are indicated as follows: * = 10%, ** = 5%,*** = 1%.

the results using the composite index but note that we obtain similar results forall the measures of distress.

We find that the impact of experience is stronger when the managers have hadrepeated experiences. The point estimates suggest that repeated experiencesare associated with a reduction of 2.3 percentage points in debt (comparedwith 1.2 percentage points for the single-experience sample), an increase of2.6 percentage points in cash savings (compared with 1.7 percentage pointsfor the single-experience sample), and a reduction of 0.5 percentage pointsin capital expenditures (compared with 0.3 percentage points for the single-experience sample). In unreported tests, we find that the differences betweenthese estimates are also statistically significant at the 5 percent level or better.This finding further supports the saliency hypothesis, and it is also related torecent work by Aktas, de Bodt, and Roll (2013), which shows that firms learnthrough repeated acquisitions and this learning depends on the time betweensuccessive transactions.

Next, we test whether more recent experience exerts a stronger impacton corporate policy. The evidence on reinforcement learning suggests thatindividuals will repeat positive outcomes and similar outcomes will be morefrequently employed (Erev and Roth 1998). This evidence is consistent withWatson’s (1930) recency law that the event that was observed most recently ismore likely repeated.

588



Dow

nloaded from



In Panel C, we test the impact of the recency of the experience. To dothis, we divide the sample into instances where the manager had her mostrecent experience of distress in the last six years (recent) and those where theexperience was more than six years ago (distant). We break the sample at sixyears because this is the median number of years since an experience occurs inthe sample. Columns 1–3 present the impact of recent experience, and columns4–6 present the impact of distant experience.

We find that experience with distress influences corporate policy regardlessof when it occurs.10 However, consistent with the recency hypothesis, we findthat the impact is stronger when the experience is more recent. Specifically,a recent experience results in a 1.9-percentage-point decrease in leverage, a2.4-percentage-point increase in cash, and a 0.5-percentage-point decreasein investment, whereas distant experiences result in a 1.2-percentage-pointdecrease in leverage, a 1.5-percentage-point increase in cash, and a 0.3-percentage-point decrease in investment.

3.2 Negative versus positive professional experiencesNext, we consider the pessimism-bias hypothesis, which predicts that negativeoutcomes will have asymmetrically stronger effects than positive outcomes.Consistent with this hypothesis, Kahneman and Tversky (1979) document theimpact of loss aversion, and Kuhnen (2015) shows that negative outcomeslead individuals to become overly pessimistic about possible investmentalternatives. We therefore examine if negative experiences have a strongereffect than positive experiences. To determine whether this is the case, we createan alternative professional experience measure to capture positive experiences.Specifically, we sort firms based on changes in credit ratings, cash flows, andstock returns, and determine if a firm falls in the top (as opposed to bottom)decile of each measure. Similar to our definition of negative experience, wefurther require that the firm’s credit is indeed upgraded, and that it experiencespositive cash flow changes and stock returns. We then create a composite indexthat equals one if any of the three individual indicators equals one. Note thatthe positive experience index does not consider an indicator for bankruptcy asthere is no upside equivalent.

The results are reported in the first three columns of Table 6, Panel D. Wefind that positive professional experience does not affect corporate policy. Thisfinding is consistent with Kuhnen (2015) and shows that negative experienceshave a stronger impact than positive experiences, in support of the pessimism-bias hypothesis.

10 Examining distant experiences also may mitigate selection bias in that more distant experiences may have lessimpact on the choice of managers as CEO. This is similar to Schoar and Zuo (2013), who study the effect of theeconomic conditions when the CEO enters the labor market on her subsequent career.

589



Dow

nloaded from



We further compare the impact of negative and positive experiencesby examining how mixed experiences influence managers. Specifically, weexamine whether positive experiences that follow negative ones attenuate orreverse the impact of the negative experiences. These results are reported incolumns 4–6 of Panel D. We find that negative experiences have a dominanteffect, which is not significantly affected by subsequent positive experiences.The point estimates in columns 4–6 are not materially different from ourbaseline estimates in Table 4.

3.3 The professional experience of CFOsSo far the analysis has focused on the professional experience of the CEO.The results indicate that CEO experience affects corporate policies, and allowfor two interpretations: (i) CEOs directly determine corporate policies, or(ii) CFOs also determine corporate policies, but their decisions are positivelycorrelated with CEO traits. In this subsection, we distinguish between the twointerpretations by directly considering the effect of the professional experienceof the CFO. To measure these professional experiences, we use the samemethodology as in our main analysis.

To study the professional experience of CFOs, we recreate our measures ofprofessional experience at troubled firms for the 3,955 CFOs in our sample.As Table 2 shows, we find that 0.6–11.1% of the CFOs in our sample werepreviously employed by troubled firms.

In Table 7, we estimate fixed effect regressions in the sample of exogenousCEO turnovers, similar to those in Table 4. The regressions examine the impactof the professional experience of the CFO, controlling for the professionalexperience of the CEO, a set of firm-, industry-, and market-level controlssimilar to that in Table 3, and the characteristics of the CEO and the CFO.We also include the interaction term CEO experience × CFO experienceto test whether the effects strengthen when both the CEO and the CFOexperienced distress. As before, the standard errors are clustered at the firmlevel.

Columns 1–3 examine all exogenous CEO turnovers. In columns 4–6, wefocus on cases in which CEO and CFO turnovers are unrelated, by excludingany instances where CFO turnovers occur within two years of CEO turnovers.This empirical methodology allows us to exclude joint turnovers of the CEOand the CFO that might be triggered by an overall change in the firm’s attributesand therefore affect corporate policy regardless of the experience of the newCEO and CFO.

The empirical results in Table 7 indicate that the firm’s financial policies,leverage, and cash savings are affected by the professional experiences of boththe CEO and the CFO. These effects are statistically significant at the 5% levelor better, and the economic magnitudes imply that the effect of the CFO’sprofessional experience is as important as that of the CEO’s experience. Basedon the sample that excludes joint turnovers, a CFO (CEO) with experience

590



Dow

nloaded from



Table 7CFOs’ professional experience

Turnovers Exogenous CEO Excluding joint turnoversturnovers (within 2 years)

Dependent Leverage Cash Capital Leverage Cash Capitalvariable holdings holdings holdings expenditure

Model (1) (2) (3) (4) (5) (6)

CEO professional experience −0.019∗∗ 0.021∗∗ −0.004∗∗ −0.015∗∗∗ 0.017∗∗∗ −0.004∗∗[0.008] [0.008] [0.002] [0.004] [0.007] [0.002]

CFO professional experience −0.012∗∗ 0.020∗∗ −0.001 −0.008∗∗ 0.019∗∗ <0.001[0.005] [0.008] [0.003] [0.003] [0.008] [0.001]

CEO x CFO professional −0.005 0.010∗ −0.003 −0.006∗ 0.012∗∗ −0.003experience [0.005] [0.005] [0.002] [0.003] [0.005] [0.003]

CEO age (/100) 0.015 −0.082∗∗∗ −0.019∗∗∗ 0.009 −0.088∗∗∗ −0.020∗∗∗[0.034] [0.009] [0.002] [0.043] [0.017] [0.000]

CEO female −0.002 −0.003 −0.001 −0.007 0.006 −0.006[0.003] [0.011] [0.001] [0.005] [0.012] [0.007]

CEO MBA degree 0.018∗∗∗ −0.019∗∗ 0.010∗∗∗ 0.018∗∗∗ −0.014∗∗ 0.007∗[0.005] [0.009] [0.000] [0.005] [0.006] [0.004]

CEO equity-based 0.141∗∗∗ −0.012∗∗∗ 0.018∗∗∗ 0.141∗∗∗ −0.011∗∗ 0.015∗∗∗compensation [0.032] [0.005] [0.005] [0.038] [0.005] [0.001]

CFO age (/100) 0.010∗ −0.010∗∗∗ −0.011 0.004 −0.014∗∗∗ −0.007[0.006] [0.004] [0.008] [0.008] [0.003] [0.008]

CFO female −0.008∗∗∗ 0.011 −0.006∗∗∗ −0.009∗∗∗ 0.015∗∗ 0.002[0.003] [0.011] [0.002] [0.003] [0.007] [0.006]

CFO MBA degree 0.002 0.001 0.007 −0.003 −0.004∗∗∗ 0.004∗∗∗[0.007] [0.003] [0.005] [0.008] [0.001] [0.001]

CFO equity-based 0.081∗∗ −0.006∗∗∗ 0.010∗∗∗ 0.083∗ −0.005∗∗∗ 0.013∗∗∗compensation [0.035] [0.001] [0.004] [0.043] [0.000] [0.004]


N. Obs. 5,063 4,853 4,966 4,214 4,081 4,133R2 0.638 0.855 0.679 0.635 0.858 0.680

This table presents evidence on the relation between the professional experience of both the CEO and the CFOand corporate policies (leverage, cash holdings, and capital expenditures). Professional experience is measuredby the composite Index of professional experience. In Columns 1–3, we consider exogenous CEO turnoversand the full sample of CFO turnovers. In Columns 4–6, we exclude cases in which the CEO and the CFO turnover within two years of each other. All variable definitions are given in the Appendix. The standard errors (inbrackets) are heteroscedasticity consistent and clustered at the firm level. Significance levels are indicated asfollows: * = 10%, ** = 5%, *** = 1%.

replacing a CFO (CEO) without experience results in a 0.8% (1.5%) decreasein leverage and 1.9% (1.7%) increase in cash. In contrast, the firm’s investmentpolicy is unaffected by the professional experience of the CFO. The resultsin columns 3 and 6 suggest that only the professional experience of the CEOaffects the firm’s investment policy.

The interaction term CEO experience × CFO experience is statisticallysignificant at conventional levels in columns 2, 4, and 5. These findingsindicate that when both the CEO and the CFO undergo a similar distressexperience, there are compounding effects on financial policy, though theeconomic magnitudes are small.

Throughout these tests, we control for characteristics of both the CEO andthe CFO. We find that CFOs and CEOs with high levels of equity-based

591



Dow

nloaded from



compensation have higher leverage, less cash, and more investment. We alsocontinue to find that CEOs with an MBA hold more debt, have less cash, andinvest more, consistent with our earlier findings.

These findings suggest that the professional experiences of the CEO andthe CFO have distinct effects on the firm’s financial policies. Our evidence onCFOs’ professional experience complements recent studies that investigate theinfluence of CFOs on firms’financial policies. Using survey evidence on CFOs,Ben-David, Graham, and Harvey (2013) show that CFOs’ forecasts about thestock market and their own firm’s prospects are “miscalibrated,” and as a result,their firms follow more aggressive corporate policies.Although we also find thatCFOs affect corporate policies, our results indicate that CFOs who experienceddistress in other companies tend to implement less aggressive policies relative toother CFOs. Malmendier and Zheng (2012) show that both CEOs’ and CFOs’overconfidence influences corporate decision-making. We complement theirwork by studying the joint impact of CEOs’and CFOs’professional experienceon corporate policies.11

4. Robustness and Extensions

4.1 Robustness: Carrying policy from a previous firmOur findings show that past distress experience influences corporate policy.However, one potential concern with this interpretation is that the managermay not be reacting to the experience, but, rather, may be carrying the policyof her old firm to the new firm when she becomes the CEO. Under thisalternative view, the impact of the experience is not the primary determinantof the policy.

To address this concern, we examine the difference in the policy from themost recent firm at which the CEO experienced distress to the firm the CEOis currently managing. The policy in the old, distressed firm is measured asthe average policy over the period when the CEO worked at her old firm priorto the distress event. The policy in the new firm is measured as the averagepolicy over the tenure of the CEO at her new firm. The dependent variable isthe percentage change in the average policy (leverage, cash holdings, capitalexpenditure) from the old firm to the new firm, and the control variables aredefined analogously as the percentage changes in the averages at the old andnew firms. As before, the regressions are estimated in a sample of exogenousturnovers in which the CEO departed as part of a succession plan (whose dateof departure was announced at least one year ahead), departed due to healthreasons (including deaths), or retired at the age of 65 or later.

11 A related question is whether a CEO who experiences distress is more likely to appoint a CFO with a similarexperience. In untabulated tests, we examine but do not find evidence that a CEO who experienced distress ismore likely to appoint a CFO who also experienced distress.

592



Dow

nloaded from



The results are presented in Table 8. In the first three columns, we examineexperience in non-CEO roles, as we have in the majority of our tests. Wefind that a manager who experienced distress will implement a policy thathas less debt, more cash, and less investment than the most recent firm atwhich she was employed when she experienced distress. The differences arestatistically significant at the 1% level and economically important. At the newfirm, leverage is lower by 2.3 percentage points, cash is higher by 1.6 percentagepoints, and investment is lower by 0.6 percentage points.

In previous analyses, we found that the experience that occurred when themanager was a CEO at another firm is particularly influential. Columns 4–6therefore repeat the analysis using only experience that occurred when themanager was a CEO, with similar results. Finally, columns 7–9 report theresults for a modified professional experience index that excludes bankruptcyexperience. By excluding bankruptcy experience, we mitigate the concern thatthe effects are driven by bankruptcy events, which tend to be more extreme,potentially involve policy changes outside the control of the CEO, and mayaffect the CEO’s future career prospects. As columns 7–9 show, the results arevirtually unchanged when we exclude bankruptcy experience.12

Taken together, these findings indicate that the manager is not merelycarrying the policy of the previous firm to the new firm; rather, she isimplementing a new policy at her new firm, which is affected by her professionalexperience.

4.2 Professional experience, performance, and corporate governanceThe evidence thus far may be consistent with either an efficient or an inefficientexplanation. If CEOs who experienced distress in the past overestimate thelikelihood and adverse implications of distress, as implied by the “hot stove”effect described in Denrell and March (2001), they might be more conservativethan is optimal in order to hedge against distress. On the other hand, if CEOs,in general, are overconfident and underestimate risk, then the conservatismof CEOs who experienced distress may push firms’ policies closer to theiroptimum. The view that managers are in general overconfident is consistentwith the hubris hypothesis introduced by Roll (1986) and with recent evidenceprovided by Ben-David, Graham, and Harvey (2013).

In this section, we seek to distinguish between these hypotheses in twoways. First, we investigate the impact of professional experience on firmperformance and value. Second, we examine the relation between corporategovernance and the impact of professional experience on corporate policy. Ifprofessional experience fuels over-conservatism, it is likely to correspond tolower performance and value, and have a stronger effect in poorly governed

12 A remaining concern is that a CEO was already influenced by prior distress experiences that occurred prior toher most recent experience. To address this concern, in unreported tests, we repeat the analyses after excludingCEOs that have experienced distress at multiple firms, and find similar results.

593



Dow

nloaded from



Tabl

e8

Cha

nge

inpo

licy

acro

ssfir

ms

Spec

ifica

tion

Exp

erie

nce

asno

n-C

EO

Exp

erie

nce

asC

EO

Exp

erie

nce

excl

udin

gba

nkru

ptcy

Dep

ende

nt�

Lev

erag

e�

Cas

h�

Cap

ital

�L

ever

age

�C

ash

�C

apita

l�

Lev

erag

e�

Cas

h�

Cap

ital

vari

able

hold

ings

expe

nditu

reho

ldin

gsex

pend

iture

hold

ings

expe

nditu

re

Mod

el(1

)(2

)(3

)(4

)(5

)(6

)(7

)(8

)(9

)

Prof

essi

onal

expe

rien

ce−0

.023

∗∗∗

0.01

6∗∗∗

−0.0

06∗∗

∗−0

.014

∗∗0.

019∗

∗∗−0

.006

∗∗−0

.020

∗∗∗

0.01

8∗∗

−0.0

05∗∗

[0.0

05]

[0.0

04]

[0.0

01]

[0.0

06]

[0.0

05]

[0.0

02]

[0.0

05]

[0.0

09]

[0.0

02]

Con

trol

sY

esY

esY

esY

esY

esY

esY

esY

esY

esN

.Obs

.53

753

753

748

348

348

348

348

348

3R

20.

086

0.10

50.

021

0.08

60.

099

0.01

90.

085

0.09

80.

020

Thi

sta

ble

exam

ines

the

chan

gein

polic

yfr

omth

em

ostr

ecen

tfirm

inw

hich

the

CE

Oex

peri

ence

ddi

stre

ssto

the

CE

O’s

curr

entp

ositi

on.T

hepo

licy

inth

eol

d,di

stre

ssed

firm

ism

easu

red

asth

eav

erag

epo

licy

prio

rto

the

dist

ress

even

twhe

nth

eC

EO

wor

ked

athe

rol

dfir

m.T

hepo

licy

inth

ene

wfir

mis

mea

sure

das

the

aver

age

polic

yov

erth

ete

nure

ofth

eC

EO

athe

rne

wfir

m.T

hede

pend

entv

aria

ble

isth

epe

rcen

tage

chan

gein

the

aver

age

polic

y(l

ever

age,

cash

hold

ings

,cap

itale

xpen

ditu

re)

from

the

old

firm

toth

ene

wfi

rm,a

ndth

eco

ntro

lvar

iabl

esar

ede

fined

anal

ogou

sly

asth

epe

rcen

tage

chan

ges

inth

eav

erag

esat

the

old

and

new

firm

s.P

rofe

ssio

nal

expe

rien

ceis

mea

sure

dby

the

com

posi

teIn

dex

ofpr

ofes

sion

alex

peri

ence

.T

here

gres

sion

sar

ees

timat

edin

asa

mpl

eof

exog

enou

stu

rnov

ers

inw

hich

the

CE

Ode

part

edas

part

ofa

succ

essi

onpl

an(w

hose

date

ofde

part

ure

was

anno

unce

dat

leas

ton

eye

arah

ead)

,de

part

eddu

eto

heal

thre

ason

s(i

nclu

ding

deat

hs),

orre

tired

atth

eag

eof

65or

late

r.A

llva

riab

lede

finiti

ons

are

give

nin

the

App

endi

x.C

olum

ns7–

9re

peat

the

anal

ysis

with

am

odifi

edpr

ofes

sion

alex

peri

ence

inde

xth

atex

clud

esba

nkru

ptcy

expe

rien

ce.T

hest

anda

rder

rors

(in

brac

kets

)ar

ehe

tero

sced

astic

ityco

nsis

tent

and

clus

tere

dat

the

firm

leve

l.Si

gnifi

canc

ele

vels

are

indi

cate

das

follo

ws:

*=

10%

,**

=5%

,***

=1%

.

594



Dow

nloaded from



Table 9CEOs’ professional experience and firm performance or value

Dependent Abnormal Abnormal Industry- Tobin’s ROAvariable returns— returns—Fama adjusted q (%)

market and French returnsmodel three-factor

model

Model (1) (2) (3) (4) (5)

Professional experience −0.003 −0.002 −0.002 −0.124∗∗∗ −0.120∗∗[0.004] [0.004] [0.003] [0.030] [0.054]

Year fixed effects Yes Yes Yes Yes YesFirm fixed effects No No No Yes Yes

N. Obs. 386 386 386 6,377 6,377R2 0.028 0.025 0.026 0.514 0.535

This table presents evidence on the relation between the professional experience of the CEO and: (i) theannouncement return surrounding the appointment of the CEO, and (ii) the value or performance of the firm asmeasured by Tobin’s q and ROA. ROA is measured as a percent. Professional experience is measured by thecomposite Index of professional experience. The regressions are estimated in a sample of exogenous turnoversin which the CEO departed as part of a succession plan (whose date of departure was announced at least oneyear ahead), departed due to health reasons (including deaths), or retired at the age of 65 or later. All variabledefinitions are given in the Appendix. The standard errors (in brackets) are heteroskedasticity consistent andclustered at the firm level. Significance levels are indicated as follows: * = 10%, ** = 5%, *** = 1%.

firms. On the other hand, if professional experience mitigates overconfidenceand the miscalibration of risk, it will correspond to higher performance andvalue, and have a stronger effect in well-governed firms.

In Table 9, we investigate the relation between the professional experience ofthe CEO and firm performance or value. We do so in a number of ways. First, wehand-collect data on the announcement dates of CEO turnovers for the sample ofexogenous turnovers. We are able to collect these data for 386 turnovers, whichcompose 72% of the exogenous turnovers in our sample. We then calculateabnormal announcement returns around these exogenous turnovers and regressthem on the professional experience index and year dummies. We calculate one-day abnormal returns using the market model (column 1), the Fama and Frenchthree-factor model (column 2), and relative to the value-weighted industryportfolio, where industry is defined based on the Fama and French 48 industryclassification (column 3). Across all three columns, we do not find a significanteffect of the CEO’s professional experience on the returns surrounding herappointment announcement.

In columns 4 and 5 of Table 9, we estimate firm fixed effects regressionsexplaining the firm’s Tobin’s q (column 4) and ROA (column 5). We findthat both Tobin’s q and ROA are negatively related to the CEO’s professionalexperience. These effects are statistically significant at the 5% level or better andare economically important. The estimates imply that a CEO who experienceddistress corresponds to a decline of 12 basis points in ROA and 5.5% inTobin’s q.

While these results are inconclusive, they provide suggestive evidencethat CEOs with distress experience reduce firm value by enacting overly

595



Dow

nloaded from



conservative leverage, cash, and investment policies, which lead to tax shieldlosses and forgoing positive net present value investments. These effects,however, are not anticipated by investors when these managers are appointedas CEOs.

In Table 10, we seek to provide additional evidence on the implicationsof professional experience for efficiency by studying the effects of corporategovernance. We use a number of corporate governance measures to gauge theseverity of the firm’s agency problems. In particular, we include the E-Index(Bebchuk, Cohen, and Ferrell 2009) of antitakeover provisions, where a higherindex level indicates weaker governance. We also consider large shareholdermonitoring. We define a blockholder indicator that equals one if an institutionalinvestor holds 5% or more of the firm’s outstanding shares and zero otherwise.Finally, we also consider the impact of the board of directors using boardindependence, where less independent boards represent weaker governance.We measure board independence as the ratio of the number of independentdirectors to the total number of directors.

In addition, we calculate a corporate governance index that combinesthe three governance measures. To maintain consistency across the threemeasures, we use the negative of the E-index, such that across all threemeasures, higher values correspond to better corporate governance. We definethe index as the average of a firm’s percentile ranking in the sample accordingto each measure. We then scale the index to range from 0 (low) to 1(high).

Table 10 presents the results of the governance regressions, estimated, asbefore, in the sample of exogenous turnovers. The dependent variable is oneof the firm’s policies (leverage, cash savings, and investment). For brevity, wereport the results for the composite index of professional experience, but theresults persist across the individual measures of professional experiences. Theindependent variable of interest is the interaction term between the compositeindex of professional experience and corporate governance. This term captureswhether the association between professional experiences and corporate policyvaries with the quality of corporate governance. Other independent variablesinclude the index of professional experience, corporate governance, and thesame set of controls as in our main analysis. As before, we cluster the standarderrors at the firm level. Due to data availability, our sample size decreases whenwe use different governance measures.

The results in Table 10 indicate that professional experience remainssignificant with similar economic magnitudes after controlling for governance.Based on the interaction term in columns 1–3, which correspond to thecomposite index of corporate governance, the impact of professional experienceon corporate policy is stronger in poorly governed firms. These effects arestatistically significant at the 10% level or better, and are economicallymeaningful. A decrease of one standard deviation in the quality of corporategovernance, as measured by a composite index of these three governance

596



Dow

nloaded from



Tabl

e10

CE

Os’

prof

essi

onal

expe

rien

cean

dco

rpor

ate

gove

rnan

ce

Dep

ende

ntL

ever

age

Cas

hC

apita

lL

ever

age

Cas

hC

apita

lL

ever

age

Cas

hC

apita

lL

ever

age

Cas

hC

apita

lva

riab

leho

ldin

gsex

pend

iture

hold

ings

expe

nditu

reho

ldin

gsex

pend

iture

hold

ings

expe

nditu

re

Mea

sure

ofgo

vern

ance

Gov

erna

nce

inde

xE

-ind

exB

lock

hold

erB

oard

inde

pend

ence

Mod

el(1

)(2

)(3

)(4

)(5

)(6

)(7

)(8

)(9

)(1

0)(1

1)(1

2)

Prof

essi

onal

expe

rien

ce−0

.021

∗∗∗

0.01

7∗∗∗

−0.0

04∗∗

−0.0

19∗∗

0.01

6∗∗∗

−0.0

04∗∗

−0.0

20∗∗

∗0.

018∗

∗∗−0

.004

∗∗−0

.024

∗∗∗

0.02

0∗∗∗

−0.0

05∗∗

[0.0

06]

[0.0

05]

[0.0

02]

[0.0

09]

[0.0

05]

[0.0

02]

[0.0

06]

[0.0

05]

[0.0

02]

[0.0

07]

[0.0

04]

[0.0

02]

Gov

erna

nce

−0.0

060.

014∗

∗0.

001

0.00

1−0

.002

∗0.

001

0.00

60.

009∗

−0.0

01−0

.014

0.02

3∗0.

002

[0.0

06]

[0.0

06]

[0.0

04]

[0.0

01]

[0.0

01]

[0.0

01]

[0.0

05]

[0.0

05]

[0.0

02]

[0.0

09]

[0.0

12]

[0.0

08]

Prof

essi

onal

expe

rien

ce0.

010∗

∗∗−0

.013

∗∗∗

0.00

2∗−0

.002

0.00

4∗−0

.001

0.00

6∗−0

.011

∗∗0.

001

0.01

1∗−0

.020

∗∗0.

006

×G

over

nanc

e[0

.003

][0

.004

][0

.001

][0

.003

][0

.002

][0

.002

][0

.003

][0

.005

][0

.003

][0

.005

][0

.008

][0

.009

]

Con

trol

sY

esY

esY

esY

esY

esY

esY

esY

esY

esY

esY

esY

esY

ear

fixed

effe

cts

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Firm

fixed

effe

cts

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

N.O

bs.

5,20

35,

203

5,20

34,

755

4,75

54,

755

4,92

64,

926

4,92

65,

203

5,20

35,

203

R2

0.61

10.

854

0.63

30.

583

0.84

10.

611

0.59

20.

837

0.61

50.

599

0.84

60.

624

Thi

sta

ble

pres

ents

evid

ence

onth

ere

latio

nbe

twee

nco

rpor

ate

gove

rnan

ce,t

hepr

ofes

sion

alex

peri

ence

ofth

eC

EO

,and

firm

-lev

elfin

anci

alpo

licie

s(l

ever

age,

cash

hold

ings

,and

capi

tal

expe

nditu

res)

.P

rofe

ssio

nal

expe

rien

ceis

mea

sure

dby

the

com

posi

teIn

dex

ofpr

ofes

sion

alex

peri

ence

.T

heke

yva

riab

leof

inte

rest

isth

ein

tera

ctio

nte

rm:

Pro

fess

iona

lex

peri

ence

×G

over

nanc

e.T

here

gres

sion

sar

ees

timat

edin

asa

mpl

eof

exog

enou

stu

rnov

ers

inw

hich

the

CE

Ode

part

edas

part

ofa

succ

essi

onpl

an(w

hose

date

ofde

part

ure

was

anno

unce

dat

leas

ton

eye

arah

ead)

,dep

arte

ddu

eto

heal

thre

ason

s(i

nclu

ding

deat

hs),

orre

tired

atth

eag

eof

65or

late

r.A

llva

riab

lede

finiti

ons

are

give

nin

the

App

endi

x.T

hest

anda

rder

rors

(in

brac

kets

)ar

ehe

tero

sced

astic

ityco

nsis

tent

and

clus

tere

dat

the

firm

leve

l.Si

gnifi

canc

ele

vels

are

indi

cate

das

follo

ws:

*=

10%

,**

=5%

,***

=1%

.

597



Dow

nloaded from



proxies, is associated with an increase of 18.6–29.8% in the effect ofprofessional experience.

We find directionally similar effects when we consider each of the governancemeasures separately. The effects, however, are statistically weaker comparedwith those of the governance index, likely because our tests lack power dueto the persistence of the individual governance measures across time coupledwith the inclusion of firm fixed effects in our tests.

The effects of corporate governance provide indirect evidence consistentwith a suboptimal explanation. Under this scenario, distress leads managers toenact overly conservative policies when they are not monitored by the investorsand board of directors. These policies may result in value losses due to the loss ofvalue-increasing debt tax shields, forgoing valuable investment opportunities,and holding excessive cash reserves. These effects also suggest that the changesin corporate policies are unlikely to be driven only by the selection of theCEO by the board of directors. If this were the case, we would expect to findstronger effects in well-governed firms, where the CEO and the board worktogether and the board has more control over CEO succession and corporatepolicies.

5. Conclusion

We know relatively little about how managers’ professional experience affectscorporate policy. In this paper, we examine how prior employment at troubledfirms affects managers’ financial and investment decisions. Our findingsindicate that firms operated by CEOs who experienced distress at another firmbehave more conservatively: they have less debt, save more cash, and spendless on capital expenditures.

Existing evidence focuses on early-life and personal experiences, whereasour paper is the first to study the role of more recent professional experiencesthroughout the manager’s career. This setting is an important source ofinfluence on managers’ decision-making because of the time proximity ofthese experiences and their greater degree of relevance to the type of decision-making required from corporate managers. In contrast to early-life experiences,professional experiences vary in their timing, frequency, and saliency, andtherefore allow us to test whether these factors, which have been found tobe important in laboratory experiments, also have an impact on real-worldmanagerial decision-making. Our results indicate that ongoing professionalexperiments—their timing, frequency, and saliency—are key determinants ofmanagers’ style.

Overall, our findings provide a possible explanation, rooted in the psychologyliterature, for the differences in management style across corporate executiveswho go through different experiences. They also suggest that management styleis not time-invariant or predetermined early in a manager’s life.

598



Dow

nloaded from



Appendix: Variable Definitions

Note: Compustat data items are given in parentheses.

A Firm-level variables

Blockholder is an indicator equal to one if an institutional investor holds 5% or more of the firm’soutstanding shares and zero otherwise.Board independence is the ratio of independent directors to total directors.Cash holdings is cash plus short-term investments (che) divided by total assets (at).Cash flow is earnings (ebitda) less interest and taxes (txt+xint), divided by total assets (at).Capital expenditure is capital expenditure (capx) divided by total assets (at).Leverage is total debt (debt in current liabilities [dlc] plus long-term debt [dltt]) divided by totalassets [at]).E-Index is an alternative antitakeover index to the G-Index, which is based on a subsample ofrelevant variables shown by Bebchuk, Cohen, and Ferrell (2009) to affect shareholder value.Industry cash flow volatility is the ten-year rolling window median volatility of cash flow/assetsacross the 48 Fama-French industries.Industry Leverage is the average leverage ratio in a firm’s industry, where industries are definedusing the Fama-French 48 industry classification.Market-to-book (or Tobin’s q) is the market value of assets, defined as total assets (at) minus bookequity (ceq) plus market value of equity (csho*prcc), divided by total assets (at).Profitability is net income (ni) divided by total assets (at).Size is the natural logarithm of the book value of total assets (at).Tangibility is net property, plant, and equipment (ppent) divided by total assets (at).�X is (Xt −Xt−1)/Xt−1for each variableX (e.g., cash holdings, leverage, capital expenditure).

B Manager-level variables

Age is the number of years since the manager was born.Depression Experience is an indicator equal to one if the manager grew up in the decade leadingto the Great Depression (CEOs born between 1920 and 1929).Military Experience is an indicator equal to one if the manager served in the military.Equity-based Compensation is the fraction of the manager’s total compensation paid in stock andoptions grants.Female is an indicator equal to one if the manager is a woman.MBA degree is an indicator equal to one if the manager holds an MBA degree.Overconfidence (options) is an indicator equal to 1 for managers who, at any point during thesample period, hold an option even though the option is at least 40% in the moneyOverconfidence (media) is an indicator that compares the number of media articles using the terms(a) “confident” or “confidence” or (b) “optimistic” or “optimism” to the number of past articlesthat portray the CEO as (c) not “confident,” (d) not “optimistic,” or (e) “reliable,” “cautious,”“conservative,” “practical,” “frugal,” or “steady”. We set the indicator equal to 1 if (a) + (b) >(c) +(d) + (e). We address possible bias due to differential coverage by controlling for the total numberof articles in the selected publications.Professional experience (bankruptcy) is an indicator equal to one if the manager worked at afirm that filed for chapter 11, excluding past employment as the CEO of the other firm.Professional experience (bond ratings) is an indicator equal to one if the manager worked at afirm that belonged to the lowest decile of Compustat firms based on annual credit ratings, excludingpast employment as the CEO of the other firm.Professional experience (cash flow shocks) is an indicator equal to one if the manager worked ata firm that belonged to the lowest decile of Compustat firms based on annual changes in operatingcash flows, excluding past employment as the CEO of the other firm.

599



Dow

nloaded from



Professional experience (stock returns) is an indicator equal to one if the manager worked at afirm that belonged to the lowest decile of Compustat firms based on annual stock returns, excludingpast employment as the CEO of the other firm.Professional experience (composite index) is the maximum of the five Experience variables:Professional experience (bankruptcy), Professional experience (bond ratings), Professionalexperience (Hadlock and Pierce), Professional experience (cash flow shocks), Professionalexperience (stock returns) , excluding past employment as the CEO of the other firm.Tenure is the number of years that the manager has been with the company.

References

Aktas, N., E. de Bodt, and R. Roll. 2013. Learning from repetitive acquisitions: Evidence from the time betweendeals. Journal of Financial Economics 108:99–117.

Almeida, H., M. Campello, and M. Weisbach. 2004. The cash flow sensitivity of cash. Journal of Finance59:1777–804.

Barber, B., and T. Odean. 2001. Boys will be boys: Gender, overconfidence, and common stock investment.Quarterly Journal of Economics 116:261–92.

Bates, T., K. Kahle, and R. Stulz. 2009. Why do U.S. firms hold so much more cash than they used to? Journalof Finance 64:1985–2021.

Bebchuk, L., A. Cohen, and A. Ferrell. 2009. What matters in corporate governance? Review of Financial Studies22:783–827.

Ben-David, I., J. Graham, and C. Harvey. 2013. Managerial miscalibration. Quarterly Journal of Economics128:1547–84.

Bertrand, M., and A. Schoar. 2003. Managing with style: The effect of managers on firm policies. QuarterlyJournal of Economics 118:1169–208.

Bharath, S., and T. Shumway. 2008. Forecasting default with the Merton distance to default model. Review ofFinancial Studies 21:1339–69.

Bordola, P., N. Gennaioli, and A. Shleifer. 2012a. Saliency theory of choice under risk. Quarterly Journal ofEconomics 27:1243–85.

Bordola, P., N. Gennaioli, andA. Shleifer. 2012b. Saliency and experimental tests of endowment effects. AmericanEconomic Review Paper and Proceedings 102:47–52.

Byrnes, J., D. Miller, and W. Schafer. 1999. Gender differences and risk-taking: A meta-analysis. PsychologicalBulletin 125:367–83.

Cain, M., and S. McKeon. Forthcoming. CEO personal risk taking and corporate policies. Journal of Financialand Quantitative Analysis.

Campello, M., E. Giambona, J. R. Graham, and C. R. Harvey. 2011. Liquidity management and corporateinvestment during a financial crisis. Review of Financial Studies 24:1944–79.

Chiang, Y.-M., D. Hirshleifer, Y. Qian, and A. E. Sherman. 2011. Do investors learn from experience? Evidencefrom frequent IPO investors Review of Financial Studies 24:156–89.

Cronqvist, H., A. Makhija, and S. Yonkers. 2012. Behavioral consistency in corporate finance: CEO personaland corporate leverage. Journal of Financial Economics 103:20–40.

Denrell, J. 2007. Adaptive learning and risk taking. Psychology Review 114:177–87.

Denrell, J., and J. March. 2001. Adaptation as information restriction: The hot stove effect. Organization Science12:523–38.

Eckbo, B. E., and K. S. Thorburn. 2003. Control benefits and CEO discipline in automatic bankruptcy auctions.Journal of Financial Economics 69:227–58.

600



Dow

nloaded from



Eckbo, B. E., K. S. Thorburn, and W. Wang. 2012. How costly is corporate bankruptcy for top executives?Working Paper.

Eckel, C., and P. Grossman. 2008. Differences in the economic decisions of men and women: experimentalevidence. In Handbook of Experimental Economic Results, vol. 1, ed. Charles Plott and Vernon Smith, 509–19.New York: Elsevier.

Erev, I., and A. Roth. 1998. Predicting how people play games: Reinforcement learning in experimental gameswith unique, mixed strategy equilibria. American Economic Review 88:848–81.

Falato, A., D. Li, and T. Milbourn. Forthcoming. Which skills matter in market for CEOs? Evidence from payfor CEOs’ credentials. Management Science.

Fee, C. E., C. J. Hadlock, and J. R. Pierce. 2013. Managers with and without style: Evidence using exogenousvariation. Review of Financial Studies 26:567–601.

Frank, M., and V. Goyal. 2009. Capital structure decisions: Which factors are reliably important. FinancialManagement 38:1–37.

Gilchrist, S., and C. Himmelberg. 1995. Evidence on the role of cash flow for investment. Journal of MonetaryEconomics 36:541–72.

Greenwood, R., and S. Nagel. 2009. Inexperienced investors and bubbles. Journal of Financial Economics93:239–58.

Hertwig, R. 2012. The psychology and rationality of decisions from experience. Synthese 187:269–92.

Hertwig, R., G. Barron, E. Weber, and I. Erev. 2004. Decisions from experience and the effect of rare events inrisky choice. Psychological Science 15:534–39.

Hertwig, R., and I. Erev. 2009. The description-experience gap in risky choice. Trends in Cognitive Sciences13:517–23.

Huson, M., R. Parrino, and L. Starks. 2001. Internal monitoring mechanisms and CEO Turnover: A long-termperspective. Journal of Finance 56:2265–97.

Hutton, I., D. Jiang, and A. Kumar. 2014. Corporate policies and Republican managers. Journal of Financialand Quantitative Analysis 49:1279–310.

Jenter, D., and F. Kanaan. 2015. CEO turnover and relative performance evaluation. Journal of Finance70:2155–84.

Kahneman, D., and A. Tversky. 1979. Prospect theory: Analysis of decisions under risk. Econometrica47:263–91.

Kashyap, A., O. Lamont, and J. Stein. 1994. Credit conditions and the cyclical behavior of inventories. QuarterlyJournal of Economics 109:565–92.

Kaustia, M., and S. Knupfer. 2008. Do investors overweight personal experience? Evidence from IPOsubscriptions. Journal of Finance 63:2679–702.

Keynes, J. M. 1936. The general theory of employment. In Interest and money. London: Harcourt Brace.

Kuhnen, C. 2015. Asymmetric learning from financial information. Journal of Finance 70:2029–62.

Lins, K., H. Servaes, and P. Tufano. 2010. What drives corporate liquidity? An international survey of cashsavings and lines of credit. Journal of Financial Economics 98:160–76.

Malmendier, U., and S. Nagel. 2011. Depression babies: Do macroeconomic experiences affect risk-taking?Quarterly Journal of Economics 126:373–416.

———. Forthcoming. Learning from inflation experiences. Quarterly Journal of Economics. Advance Accesspublished October 1, 2015, 10.1093/qje/qjv037.

Malmendier, U., and G. Tate. 2005. CEO overconfidence and corporate investment. Journal of Finance60:2661–700.

601



Dow

nloaded from



Malmendier, U., G. Tate, and J. Yan. 2011. Overconfidence and early-life experiences: The Effect of managerialtraits on corporate financial policies. Journal of Finance 66:1687–733.

Malmendier, U., and H. Zheng. 2012, Managerial duties and managerial biases. Working Paper, University ofCalifornia at Berkeley.

March, J. 1996. Learning to be risk averse. Psychology Review 103:309–19.

Miller, M., and D. Orr. 1966. A model of the demand for money by firms. Quarterly Journal of Economics80:413–35.

Nisbett, R., and L. Ross. 1980. Human inference: Strategies and shortcomings of social judgment. EnglewoodCliffs, N.J.: Prentice Hall.

Opler, T., L. Pinkowitz, R. Stulz and R. Williamson. 1999. The determinants and implications of corporate cashholdings Journal of Financial Economics 52:3–46.

Roll, R. 1986. The hubris hypothesis of corporate takeovers. Journal of Business 59:197–216.

Schoar, A., and L. Zuo. 2013. Shaped by booms and busts: How the economy impacts CEO careers andmanagement styles. Working Paper.

Tversky, A., and D. Kahneman. 1973. Availability: A heuristic for judging frequency and probability. CognitivePsychology 4:207–32.

———. 1974. Judgment under uncertainty: Heuristics and biases. Science 185:1124–31.

Vissing-Jorgensen, A. 2004. Perspectives on behavioral finance: Does “irrationality” disappear with wealth?Evidence from expectations and actions. NBER Macroeconomics Annual, 139–208.

Watson, J. 1930. Behaviorism. New York: W. W. Norton and Company.

Weber, E., A. Blais, and N. Betz. 2002. A domain-specific risk attitude scale: Measuring risk perceptions andrisk behaviors. Journal of Behavioral Decision-Making 15:263–90.

Yermack, D. 2006. Golden handshakes: Separation pay for retired and dismissed CEOs, Journal of Accountingand Economics 41:237–56.

602



Dow

nloaded from


本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页文献云下载图书馆入口外文数据库大全疑难文献辅助工具

http://www.xuebalib.com/cloud/

http://www.xuebalib.com/

http://www.xuebalib.com/cloud/


http://www.xuebalib.com/vip.html

http://www.xuebalib.com/db.php

http://www.xuebalib.com/zixun/2014-08-15/44.html


looking in the rearview mirror: the effect of managers ...download.xuebalib.com/1fe8d0sawnlx.pdf ·...

Documents