Management Risk Incentives and the Readability of Corporate Disclosures

• Author: Bidisha Chakrabarty,Ananth Seetharaman,Zane Swanson,Xu (Frank) Wang
• DOI: http://doi.org/10.1111/fima.12202
• Date: 01 September 2018
• Published date: 01 September 2018

Management Risk Incentives and the

Readability of Corporate Disclosures

Bidisha Chakrabarty, Ananth Seetharaman, Zane Swanson,

and Xu (Frank) Wang∗

Managers with higher risk incentives (greateroptions vega) issue less readable disclosures. Firms

in the top quartile of vega file annual reports that are about 15.4% more voluminous than those

in the bottom quartile. The effect of vega on obfuscation remains after controlling for firm risk,

operating complexities, accounting and auditor choices, chief executive officer changes, and an

exogenous shockto option compensation. This effect is tempered by higher institutional ownership,

lower management entrenchment, and greater analyst following. Obfuscation benefits managers

by increasing return volatility (option value) and allowing greater earnings management. These

findings document a new link between options and disclosure transparency.

“One way to hide a log is to put it in the woods.”1

We examine the role of stock-options-based compensation on the readability of corporate

annual reports (Form 10-Ks). Economic theory posits that the asymmetric payoff of stock options

incentivizes managers to overcome their risk aversion and invest in high-risk/high-return projects

(Harris and Raviv, 1978; Amihud and Lev, 1981), and empirical research shows that managers

invest in riskier projects whengiven more options (Guay, 1999; Coles, Daniel, and Naveen,2006).

The underlying assumption is that managers respond to option compensation by investingin high-

risk (expected) positive net present value projects that are consistent with the firm’s strategy and

increase shareholder wealth.

However, theory (Bolton, Scheinkman, and Xiong, 2006) and empirical studies (Dong, Wang,

and Xie, 2010; Kim, Li, and Zhang, 2011) document that options-compensated managers may

also take actions that, although increasing return volatility (and consequently their options value),

do not promote shareholder interests. It is reasonable to surmise that if their actions are motivated

by self-interest, or do not align with the firm’s strategy, managers may hide information to mask

their choices. Howevernondisclosure of material information imposes large litigation costs (Field,

Lowry, and Shu, 2005). Thus, we ask the question: Do options-compensated managers obfuscate

their firms’ disclosures? We investigate this issue by examining the effect of options-based

Wethank Raghu Rau (Editor), an anonymous referee, Daniel Chi, Scott Duellman, YongtaeKim, Vikrant Vig,and seminar

participants at Saint Louis University for helpful comments.

∗Bidisha Chakrabarty is the Edward Jones Professor of Finance at Saint Louis University in St. Louis, MO. Ananth

Seetharaman is the Don and Donna Millican Endowed Professor and Chairman of the Department of Accounting at

the University of North Texas in Denton, TX. Zane Swanson is an Accounting Full Professor at University of Central

Oklahoma in Edmond, OK. Xu (Frank) Wang is an Associate Professor of Accounting at Saint Louis University in St.

Louis, MO.

1Bethany McLean, “Why Enron WentBust,” Fortune (December 24, 2001), p. 58, quoting Congressman John Dingell’s

reaction to Enron’sannual report. “One way to hide a log is to put it in the woods,” said Michigan Democrat John Dingell,

who is calling for a congressional investigation. “What we’re looking at here is an example of superbly complexf inancial

reports. They didn’t haveto lie. All they had to do was to obfuscate it with sheer complexity—although they probably lied

too.” Availableonline at: http://archive.fortune.com/magazines/fortune/fortune_archive/2001/12/24/315319/index.htm.

Financial Management •Fall 2018 •pages 583 – 616

584 Financial Management rFall 2018

risk incentives embedded in managerial compensation on the readability of the firm’s annual

report.

Several factors lead us to hypothesize a link between managerial risk incentivesand disclosure

obfuscation. Some studies draw a negative link between options compensation and mandatory

disclosure. Options-compensated managers have been shown to strategically report favorable

information and discourage the disclosure of unfavorable information to increase their firms’

stock price and inflate their option value. For instance, Aboody and Kasznik (2000) find that

managers time their voluntary disclosures around stock option awards, and Call, Kedia, and

Rajgopal (2016) find that fir ms award more stock options even to rank-and-file employees when

they are engaged in misreporting, to discourage whistle blowing. Kim et al. (2011) find that firms

whose top executiveshave large options holdings have a greater risk of future stock price crashes

and conclude that options encourage managers to hide bad news to delay stock price declines.2

Armstrong et al. (2013) find that when managers have a higher sensitivity to stock price volatility,

they show a greater propensity to misreport. These studies link option compensation to strategic

manipulation of a firm’s disclosure.

Another stream of research links stock options to managerial actions that reduce firm value.

For instance, Sanders and Hambrick (2007) find that returns are more likely to be extremely

negative than extremely positive in firms with options-compensated executives. They conclude

that options motivate chief executive officers (CEOs) to take big risks or to “swing for the

fences.” Armstrong and Vashishtha (2012) demonstrate that stock options give CEOs incentives

to increase systemic risk even if that does not increase firm value. Cooper, Gulen, and Rau

(2016) find that options compensation is associated with overinvestment and loss of shareholder

wealth. Dong et al. (2010) showthat options-compensated managers overlever their firms, leading

to a suboptimal capital structure. Bloom and Milkovich (1998) show that higher risk firms

that rely on incentive pay exhibit poorer performance than similar firms without incentive pay.

Kolb (2012) writes that a manager’s holding of a substantial number of options provides a

powerful incentive to “try nearly anything” to increase the options’ value. Collectively, these

results make the case that options-compensated managers may make decisions that reduce firm

value.

Finally, some authors suggest that managers obfuscate disclosures to mask unfavorable in-

formation. Li (2008) documents that annual reports of fir ms with lower earnings are harder

to read whereas the positive earnings of firms that issue easy-to-read annual reports are more

persistent. He suggests that managers opportunistically choose annual report readability to hide

adverse information. Laksmana, Tietz, and Yang (2012) examine the proxy statement disclo-

sures of CEO compensation and argue that disclosures have low readability when CEO pay is

not tied to the economic determinants of compensation. The logical conclusion of these three

strands of research is that options-induced risk taking by managers may not be for the benefit

of shareholders, and managers may try to withhold this information by obfuscating mandatory

disclosures.

In contrast, classical economic theory posits that options align managerial and shareholder

interests by mitigating agency costs. For example, DeFusco, Johnson, and Zorn (1990) find

that when a firm approves an executive stock option (ESO) plan, there is a positive stock

price reaction, indicating that shareholders interpret option grants as a positive event. Rajgopal

and Shevlin (2002) explore ESO grants in a high-risk industry—oil and gas exploration—and

document that ESO risk incentives have a positive relation with future exploration risk taking.

If, as theory suggests, options induce managers to increase firm risks that benef it shareholders,

2Hiding bad news via obfuscation aligns with the “incomplete revelation hypothesis” (Bloomfield, 2002).

Chakrabarty et al. rRisk Incentives and Disclosure Readability 585

it is in their interest to communicate their actions through transparent disclosures. Although the

evidence linking options to value-destroying managerial actions predicts disclosure obfuscation,

models that link options to reduced agency costs predict disclosure transparency. It is therefore

an open question whether managers obfuscate firm disclosures when given risk incentives via

options compensation.

Form 10-K obfuscation of the type alluded to in the above arguments is likely achieved by

means of “hidden in plain sight” disclosure practices that bury information in voluminous prose

while providing a cover against potential future litigation (Bloomfield, 2012). A measure of

obfuscation that captures such practices is critical to our study. We use the metric OBSCU, which

is an inverse measure of the overall readability of a Form 10-K (Loughran and McDonald, 2014;

Kubick and Lockhart, 2016). Loughran and McDonald (2014) show that as a measure of financial

disclosure obfuscation, OBSCU is superior to other measures used in the prior literature. OBSCU

is computed as the natural log of the US Securities and Exchange Commission (SEC) Form

10-K submission file size. Larger 10-K file sizes imply higher obfuscation (lower readability).

In robustness tests, we use the FOG index from the computation linguistics literature (Li, 2008)

as an alternative readability measure.3

We measure managerial risk incentive by their options vega (Chava and Purnanandam, 2010;

Croci, Del Giudice, and Jankensgard, 2017), with the options valued using the Black-Scholes

(1973) model.4Testsare based on a sample of 26,579 Form 10-K f ilings (firm-year observations)

covering 1993-2015. To examine whether higher vegas induce managers to obfuscate Form10-K

filings, we regress the readability metric OBSCU on (lagged) vega and a vector of variables that

control for the degree of risk and complexity inherent in the firm’soperations. We find a significant

positive association between the current-period OBSCU of Form 10-K and (lagged) vega. This

result indicates that after controlling for firm complexity, accounting choices, and business risk,

managers with higher vega compensation contracts at time t−1 file less transparent disclosures

at time t. Specifically, we showthat a CEO who is in the top quartile vega f iles a Form 10-K that

is, on average, 15.4% more voluminous than the Form 10-K filed by a CEO who is in the bottom

quartile vega.

It may be argued that options induce managers to undertake complex projects that require

lengthy disclosures. If this alternative explanation is true, our finding of a positive association

between vega and Form 10-K obfuscation may represent an operational necessity rather than

managerial choice. To rule out this alternative, weperfor m the following tests. First, we examine

the link between vega and disclosure obfuscation in a sample of firms that change CEOs. When a

firm has a new CEO, the CEO vega changes because the incoming CEO has a new compensation

package, yet it is unlikely that the nature of the firm’s operations has immediately changed.

In this CEO change sample, we find that disclosure complexity increases (decreases) when

the new CEO has a higher (lower) vega. Additionally, we find the options-induced obfuscation

effect is prominent for internally appointed (rather than externally hired) CEOs, consistent with

entrenched CEOs taking more value-destroying actions (Harford,Humpher y-Jenner, and Powell,

2012), which they likely obfuscate in disclosure filings.

3Loughran and McDonald (2014) compare the FOG Index, word counts, and several other readability measures with

10-K file size. They find that f ile size is a superior measure in quality and simplicity and less misspecified in the context

of financial disclosures.

4We note that the Black-Scholes (1973) valuation model to calculate the Delta and Vega of options may not represent

perfectly the degree of managerial risk aversion. Black-Scholes (1973) valuation assumes risk neutrality of managers

although they may actually be risk averse. See Chava and Purnanandam (2010), who also use Black-Scholes (1973)

valuation, for a detailed discussion.