CEO Age and Stock Price Crash Risk*
Panayiotis C. Andreou
1,2
, Christodoulos Louca
1,2
, and
Andreas P. Petrou
1
1
Cyprus University of Technology and
2
Durham University
Abstract
We show that firms with younger CEOs are more likely to experience stock price
crashes, including crashes caused by revelation of negative news in the form of
breaks in strings of consecutive earnings increases. Such strings are accompanied
by large increases in CEO compensation that do not dissipate with crashes. These
findings suggest that CEOs have financial incentives to hoard bad news earlier in
their career, which increases future crashes. This negative impact of CEO age effect
is strongest in the presence of managerial discretion. Overall, the findings highlight
the importance of CEO age for firm policies and outcomes.
JEL classification: G30, G02
Keywords: CEO age, Crash risk, Hoarding of bad news, Agency theory, Managerial discretion
1. Introduction
A considerable body of literature suggests that managers might hide bad operating perform-
ance news from investors when faced with adverse outcomes that affect negatively their
personal wealth (Gibbons and Murphy, 1992; Bliss and Rosen, 2001). However, if man-
agers withhold and accumulate negative information for an extended period, this eventually
leads to bad news stockpiling within the firm and to severe stock overvaluation. When
stockpiling reaches a critical threshold level, it becomes too costly for managers or even im-
possible to continue withholding the accumulated negative information (Baik, Farber, and
Lee, 2011). When revealed at one time in the market, the bad news will lead to a substantial
revision of investors’ expectations about the future prospects of the firm and, inevitably, to
a stock price crash (Jin and Myers, 2006).
* We are especially grateful to Vikrant Vig (Editor) and an anonymous referee for their constructive
suggestions and helpful comments. We also thank Demetris Koursaros, Neophytos Lambertides,
Hermes Niels, Panayiotis Theodossiou, Dimitris Tsouknidis, Photis Panayides, Christos Savva, and
participants at the BI Norwegian Business School, Cyprus University of Technology, and 2016
FINEST Summer Workshop for useful comments and suggestions. Andreas Procopiou provided ex-
cellent research assistance. All remaining errors are our own.
V
C
The Authors 2016. Published by Oxford University Press on behalf of the European Finance Association.
All rights reserved. For Permissions, please email: journals.permissions@oup.com
Review of Finance, 2017, 1287–1325
doi: 10.1093/rof/rfw056
Advance Access Publication Date: 24 October 2016
Downloaded from https://academic.oup.com/rof/article-abstract/21/3/1287/2444586
by Cyprus University of Technology user
on 21 November 2017
The literature generally ascribes stock price crashes to the failure of corporate govern-
ance control systems to alleviate agency problems (Hutton, Marcus, and Tehranian, 2009;
Kim, Li, and Zhang, 2011a; Callen and Fang, 2013; Andreou et al., 2016a; Kim and
Zhang, 2016). Despite this conceptual interest on agency problems, this literature focuses
on firm attributes and ignores agency problems that relate to CEO characteristics. In this
study, we suggest that pay-performance sensitivity creates incentives for bad news hoard-
ing. The incentives vary with CEO age and become a source of agency problems that leads
to the prediction that firms managed by younger CEOs are more likely to experience stock
price crashes.
The study draws motivation from prior literature suggesting that CEOs are highly con-
cerned about firm performance because performance directly affects their current and fu-
ture personal wealth through executive compensation packages (Gibbons and Murphy,
1992; Bliss and Rosen, 2001; Petrou and Procopiou, 2016). Thus, when the actions of
CEOs fail to deliver, concerns about their personal wealth can incentivize them to conceal
adverse operating outcomes from shareholders. However, the pay-performance sensitivity
of CEOs varies with CEO age. Younger CEOs could secure significant permanent increases
in compensation early in their career, which they can enjoy for a longer period.
Accordingly, younger CEOs might have more financial incentives to intentionally conceal
and accumulate adverse operating outcomes from investors, increasing in this respect the
probability of experiencing a stock price crash in the future.
We test these predictions using ExecuComp firms for the period 1995–2013. We meas-
ure firm-specific stock price crashes as the presence of an extreme negative firm-specific
weekly return (Hutton, Marcus, and Tehranian, 2009; Kim, Li, and Zhang, 2011a).
Controlling for other known determinants of stock price crashes, the results show that
firms managed by younger CEOs are more likely to experience a stock price crash. To in-
vestigate the mechanism underpinning this relationship, that is, the hoarding of bad news,
we focus on stock price crashes triggered by earnings announcements that break previous
years’ strings of consecutive earnings increases. Myers, Myers, and Skinner (2007) suggest
that breaks in strings of consecutive earnings increases emanate from stockpiling of nega-
tive news, particularly when the break occurs after a longer string. Thus, breaks in earnings
strings that trigger stock price crashes is a manifestation of agency risk pertaining to the
practice of bad news hoarding. In addition, the severity of agency risk is positively related
to the length of the string. Using these crashes, we still find that firms managed by younger
CEOs are more likely to experience a stock price crash, ascertaining that the mechanism of
stockpiling of negative information pertaining to adverse operating performance drives this
relationship. In corroboration, we find that the length of the string prior to the break is
more strongly associated with crashes when a younger CEO leads the company. Next, we
investigate CEOs’ pay-performance incentives by focusing on the evolution of CEO com-
pensation before (up to 3 years), during, and after (up to 1 year) stock price crashes.
Controlling for known determinants of CEO compensation, the results demonstrate large
increases in CEO compensation in periods of consecutive earnings increases. Interestingly,
CEO compensation does not revert to previous levels during and after the crash. These find-
ings imply that CEOs have strong financial incentives to generate strings of consecutive
earnings increases earlier in their career, resulting in a CEO agency problem that drives
stock price crashes.
To prevent moral hazard situations, agency theory identifies the board’s monitoring
role, among others, as a critical control system (Eisenhardt, 1989). Accordingly, we
1288 P. C. Andreou et al.
Downloaded from https://academic.oup.com/rof/article-abstract/21/3/1287/2444586
by Cyprus University of Technology user
on 21 November 2017
examine two organizational factors which compromise board monitoring and increase
managerial discretion, namely, the CEO duality in the governance structure (Jensen, 1993;
Dalton et al., 1998) and the degree of corporate diversification (Martin and Sayrak, 2003;
Ndofor, Wesley, and Priem, 2013). Our results show that these two factors strengthen the
relationship between younger CEOs and future crash risk. This finding raises important
considerations for the competence of the board to effectively monitor and control self-
interested young CEOs.
Our results are robust to alternative measures of stock price crash risk, such as the nega-
tive coefficient of skewness of firm-specific weekly returns (Chen, Hong, and Stein, 2001)
and the negative of the worst deviation of firm-specific weekly return (Bradshaw et al.,
2010). In addition, the results are robust to potential model misspecifications. Specifically,
a propensity score-matching analysis ensures that the CEO age effect is not driven by differ-
ences between firms managed by younger or older CEOs among observable: (i) firm charac-
teristics, such as firm size, growth, leverage, profitability, performance, and age, and (ii)
CEO characteristics, for instance tenure, turnover, retirement, in the money option hold-
ings and equity holdings.
We also consider a variety of alternative explanations. First, a reverse relationship run-
ning from crash risk to CEO age is likely to exist under two conditions: (i) stock price crash
risk relates to CEO turnover and firms hire younger CEOs, and (ii) stock price crash risk
exhibits persistence. However, we find no statistically significant difference in the age of
newly hired CEOs for firms that experience a stock price crash relative to firms that do not.
In addition, after examining firms that exhibit more difficulties in handling risk or inher-
ently risky firms, which may require more healthy, flexible, and energetic young CEOs, we
find no evidence that the age of newly hired CEOs is significantly different among firms
that experience a stock price crash and firms that do not. Hence, crash risk is unlikely to re-
late to the age of newly hired CEOs. Finally, as a complementary test of the reverse causal-
ity explanation, we re-run the main analysis and find qualitatively similar results after
excluding the first three years of CEO tenure, which are affected more by persistence in
crash risk, and thus potentially may cause a reverse relationship.
Second, physiological and psychological characteristics of the CEO and heterogeneous
abilities change with age, and some of these characteristics might provoke stock price
crashes. Such characteristics include the effects of ability, power, overconfidence, youthful
creativeness, and inexperience with corporate communications. Controlling for CEO dem-
onstrated ability, power, and overconfidence, the results remain unaltered. Youthful cre-
ativeness and inexperience with corporate communication are more problematic to control
directly because it is difficult to measure them precisely; nevertheless, we can observe their
consequences, and hence, we can design appropriate tests to examine their merit as alterna-
tive explanations of the CEO age effect. More specifically, youthful creativeness associated
with younger CEOs experimenting with novel strategies should predict fat tails generally,
not only one-sided exposure to crashes. In contrast to such an explanation, we find no rela-
tionship between CEO age and the probability of a positive jump in the firm-specific
weekly returns. Thus, CEO age appears to predict only negative jumps, that is, stock price
crashes. Similarly, inexperience of younger CEOs in corporate communication could lead
them to portray optimistic earnings expectations to analysts. In response, younger CEOs
might hoard bad news to meet or beat analyst earnings forecasts, increasing in this respect
future stock price crash risk. Excluding crashes that likely result from setting inappropriate
earnings expectations from the main analysis does not affect the CEO age effect.
CEO Age and Stock Price Crash Risk 1289
Downloaded from https://academic.oup.com/rof/article-abstract/21/3/1287/2444586
by Cyprus University of Technology user
on 21 November 2017
Finally, we examine the possibility that the CEO age effect reflects unobservable habit-
ual CEO characteristics (Graham, Li, and Qiu, 2012) that affect disproportionately
younger CEOs. Specifically, such characteristics can have implications for stock price
crashes and can lead to CEO turnover, particularly younger CEOs who are less reputable,
creating a sample selection bias that affects mostly younger CEOs. Nevertheless, this ex-
planation does not gain support because we still find CEO age effect for the subsample of
firms with CEOs that avoid turnover for at least 5 years. In this subsample, habitual CEO
characteristics should affect a firm’s crash risk similarly over a long period.
This study contributes to the literature on stock price crashes by showing that compen-
sation incentives create CEO-level variation in agency problems that increase the likelihood
of firms with younger CEOs to experience future stock price crashes due to hoarding of bad
news. Prior literature finds that crash risk relates to accounting opacity (Hutton, Marcus,
and Tehranian, 2009), tax avoidance (Kim, Li, and Zhang, 2011b), accounting conserva-
tism (Kim and Zhang, 2016), equity-based compensation (Kim, Li, and Zhang, 2011a),
and inefficient governance (Callen and Fang, 2013; Andreou et al., 2016a). However, what
motivates managers to conceal bad news largely remained unexplored in the literature.
This study’s main contribution fills this gap by providing novel evidence that CEOs have fi-
nancial incentives to pursue bad news hoarding activities earlier in their career, which sub-
sequently lead to stock price crashes.
In addition, the study contributes to the emerging literature that links heterogeneous
CEO characteristics to firm policies and outcomes (Bertrand and Schoar, 2003). In this
vein, recent studies find that CEO age significantly affects corporate investments. For in-
stance, Yim (2013) finds that financial incentives motivate younger CEOs to make more ac-
quisitions, whereas Serfling (2014) provides evidence that older CEOs invest less in
research and development, make more diversifying acquisitions and maintain lower operat-
ing leverage, resulting in lower firm risk. Our perspective is different and links CEO age to
future stock price crashes. This perspective has important implications for corporate gov-
ernance policies by raising concerns about the role of boards in monitoring and incentiviz-
ing CEOs. Specifically, the findings of our study should probe boards to devise appropriate
governance mechanisms that combat agency problems that emerge from CEO age.
The rest of the study is organized as follows. Section 2 develops our hypotheses and out-
lines the testable predictions. Section 3 describes the research design. Section 4 presents the
empirical results. Section 5 presents the robustness analysis results. Section 6 presents re-
sults on alternative explanations of the findings. Finally, Section 7 concludes the study.
2. Hypotheses Development
2.1 CEO Age and Crash Risk
Gibbons and Murphy (1992) argue that the “labor market uses a worker’s current output
to update its belief about the worker’s ability and then base future wages on these updated
beliefs”. Accordingly, superior performance affects a manager’s value in the labor market
and results in future compensation increases. Because of that relationship, younger CEOs
should have strong financial incentives to deliver superior (or to hide poor) performance to
gain early rises in compensation, which they will enjoy for a longer period. Consistent with
this argument, Yim (2013) finds that younger CEOs are more likely to pursue acquisitions
and that CEOs are rewarded as much as $300,000 in additional annual compensation for
each sizable acquisition they make. Similarly, Boschen et al. (2003) show that excess
1290 P. C. Andreou et al.
Downloaded from https://academic.oup.com/rof/article-abstract/21/3/1287/2444586
by Cyprus University of Technology user
on 21 November 2017
performance has a positive effect on the cumulative financial gain of CEOs. Such evidence
suggests that younger CEOs might be more sensitive about firm performance and that simi-
lar performance achievements have more wealth-related value for younger CEOs.
Drawing on agency theory (Jensen and Meckling, 1976), we suggest that different levels
of CEO pay for performance sensitivity, which depend upon CEOs’ ages, should create dif-
ferent responses to adverse operating outcomes. For instance, disclosure of negative infor-
mation about performance should harm the personal wealth of younger CEOs more
because the labor market will use this information to update beliefs about their abilities
and set a corresponding (lower) level of compensation (Gibbons and Murphy, 1992),
which, when accumulated across a CEO’s career, is more costly for younger CEOs.
Therefore, these CEOs have more incentives to hide negative information to avoid personal
wealth consequences, hoping that poor current performance will be offset by stronger fu-
ture performance. Hiding and accumulating bad news, however, is unsustainable in the
long run; eventually, bad news will spill out in the market when strong future performance
does not materialize (Jin and Myers, 2006; Bleck and Liu, 2007). Investors’ response to un-
expected bad news is fierce, leading to an abrupt downward revision of their expectations
about the firm’s long-term prospects, which triggers a stock price crash (Jin and Myers,
2006; Callen and Fang, 2015). The abovementioned discussion leads us to the following
hypothesis:
Hypothesis 1. Firms managed by younger CEOs are associated with higher levels of future stock
price crash risk.
2.2 The Moderating Effect of Management Discretion
CEOs are more prone to engage in moral hazard situations when they have discretion,
which they might use to compromise the effectiveness of the boards’ monitoring function
(Finkelstein and Hambrick, 1989; Ocasio, 1994). Such opportunities emerge in the pres-
ence of two organizational characteristics: the existence of CEO duality in the governance
structure and the degree of corporate diversification.
A CEO-Chair can acquire significant influence over the board, thereby weakening the
board’s ability to effectively monitor and control management decisions (Hambrick and
Finkelstein, 1987; Jensen, 1993; Dalton et al., 1998). This influence can be achieved in a
number of ways. First, CEO-Chairs, who nominate board directors, can select directors
who are loyal to them (Westphal and Zajac, 1995). Second, the duality structure can enable
CEOs to root themselves in the organization by creating norms of not questioning manage-
ment effectiveness (Finkelstein and D’aveni, 1994). Finally, these CEOs might control the
board’s distribution of attention to organizational matters, purposely discouraging ad-
equate attention to monitoring (Tuggle et al., 2010). Consequently, when the CEO-Chair
position is held by younger CEOs who are more sensitive to adverse changes in firm per-
formance, it is more likely to suppress the board’s monitoring function to facilitate hoard-
ing of bad news from shareholders. Effectively, such behavior makes firms more prone to
future stock price crash risk. Consequently, we expect that:
Hypothesis 2. The relationship between CEO age and future stock price crash risk is stronger in
the presence of a CEO-Chair position.
Likewise, in diversified firms, there is greater organizational complexity, which can com-
promise in many ways the effectiveness of board monitoring (McKendall and Wagner,
CEO Age and Stock Price Crash Risk 1291
Downloaded from https://academic.oup.com/rof/article-abstract/21/3/1287/2444586
by Cyprus University of Technology user
on 21 November 2017
