The Propensity to Split and CEO Compensation

• Published date: 01 March 2018
• Author: Richard S. Warr,Erik Devos,William B. Elliott
• Date: 01 March 2018
• DOI: http://doi.org/10.1111/fima.12179

The Propensity to Split and CEO

Compensation

Erik Devos, William B. Elliott, and Richard S. Warr∗

We analyze the relation between the delta and vega of a chief executive officer’s (CEO) com-

pensation and the propensity of the firm to engage in a split. Controlling for other well-known

factors, we find that CEOs with compensation that has higher levels of delta are more likely to

split their shares. Furthermore, the choice of split factor is inversely related to delta. Our results

are economically significant: for the average (median) firm in our sample, a stock split results in

a CEO wealth gain of $4.9 million ($84,000).

Early studies of stock splits document a strong and positive abnormal return upon their an-

nouncement (see, e.g., Ikenberry, Rankine, and Stice, 1996). Subsequent authors have attempted

to identify factors that either trigger stock splits or that are related to the abnormal announcement

returns resulting from the split announcement. Researchers have posited a variety of explanations

for stock splits, including signaling to reduce information asymmetry (Brennan and Copeland,

1988; Dharan and Ikenberry, 1995; Ikenberry et al., 1996), adjusting the stock price to an optimal

trading range (Copeland, 1979; Fernando, Krishnamurthy, and Spindt, 2004; Dyl and Elliott,

2006) or to an optimal tick size (Angel, 1997; Harris, 1997), and to increase the tax option

value for investors (Lamoureux and Poon, 1987). In addition to the announcement return effect,

beginning with Ohlson and Penman (1985), research has documented a significant increase in

post-split stock volatility.

In this article, we examine whether the composition of a chief executive officer’s (CEO)

compensation portfolio increases the likelihood that the firm will announce a stock split.1Because

the announcement of a stock split frequently results in a substantial increase in both the price level

and the return volatility, wehypothesize that a CEO whose compensation portfolio is sensitive to

these effects would be more likely to favor a stock split.2We base our theses on the broad idea

that CEOs with option and stock compensation portfolios benefit from stock splits in two broad

ways. First,an increase in the stock price associated with the split announcement will increase the

Wethank Edward A. Dyl, VincentIntintoli, Murali Jagannathan, Srini Krishnamurthy,Raghu Rau (Editor), Ousmane Seck,

an anonymous referee, seminar participants at North Carolina State University and University of Texas–El Paso, and

session participants at the 2010 meetings of the Financial Management Association for helpful comments. The authors

gratefully acknowledgethe contribution of Thomson Financial for providing analyst data available throughInstitutional

Brokers’ Estimate System (IBES). These data wereprovided as part of a broad academic programto encourage earnings

expectations research. We retain responsibilityfor any remaining errors.

∗Erik Devos is a Professor of Financeand holds the J.P.Morgan Chase Professorship in Business Administration in the

College of Business Administration at the University of Texas–El Paso in El Paso, TX. William B. Elliott is a Professorof

Finance and holds the EdwardJ. and Louise E. Mellen Chair in Finance in the Boler School of Business at John Carroll

University in University Heights, OH. Richard S. Warr is Professor of Finance in the Poole College of Management at

North Carolina State University in Raleigh, NC.

1In an unpublished working paper, Baghai-Wadji and Gabarro (2009) suggest a link between CEO stock ownership and

stock splits.

2Obviously,a CEO may not unilaterally cause the firm to split its shares. At a minimum this decision requires a vote of

the board of directors. However,there is little doubt that the CEO has significant influence on the decision.

Financial Management •Spring 2018 •pages 105 – 129

106 Financial Management rSpring 2018

value of both the equity portion of their portfolio and the value of the options in their portfolio.

Second, the increase in price volatilityassociated with a split will increase the value of the options

in the CEO’s portfolio.

Evidence that CEOs know of the economic effects of splits comes from a variety of sources.

First, survey evidence suggests that CEOs are aware of the positive stock price effects of an-

nouncing a split. For example, in Baker and Powell’s (1993) survey of the management of 251

firms, 73.3% of the respondents agreed with the statement: “A stock split has a favorable market

reaction on a firm’s stock price.” Second, evidence exists that suggests managers time splits

to benefit the fir m (e.g., Dyl and Elliott, 2006; Baker, Greenwood, and Wurgler, 2009). Third,

anecdotal evidence suggests that managers are keenly aware of the positive stock price effects

surrounding the split announcement. For example, in 2012, Tim Cook, in response to a question

from a stockholder stated that a stock split results in a “short-term pop.”3Apple announced a

7-for-1 (7:1) split on April 23, 2014, which was greeted by an 8% stock price increase.4Also,

Starbucks’ CEO, Howard Schultz, believes that shareholders get more excited about a stock

split announcement than any other firm announcement.5Finally, Devos, Elliott,and Warr (2015)

provide evidence that CEOs are cognizant of the effects that stock split announcements may

have on their compensation portfolios. They report significant clustering of new option grants

before splits and increased levels of stock sales immediately after splits.6They conclude that

these actions are unlikely to be random and are consistent with the conjecture that option grants

to CEOs are timed such that the CEOs benefit from the largely value-increasing effect of a stock

split.

Toquantify how the price and volatility effects of a split might affect a CEO, we use the metrics

developed by Guay (1999) and Core and Guay (2002), delta (the sensitivity of CEO wealth to a

1% change in stock price, stated in thousands of dollars) and vega (the sensitivity of CEO wealth

to a 1% change in the standard deviation (SD) of stock returns, stated in thousands of dollars),

to study the effect of the CEO’s compensation portfolio on the decision to split. As noted by

Coles, Daniel, and Naveen (2006) and Core and Guay (2002), delta and vegaare superior to other

proxies (e.g., number of options held, value of options held, and number of options granted) of

the sensitivity of CEO compensation to changes in the value and volatility of the firm’s stock.

Our article makes several contributions to the literature. First, after controlling for known

determinants of splits, we show that the decision to split is directly related to the delta of the

CEO’s compensation portfolio. This result is not only statistically significant, but we contend

that it is also economically significant. The average (median) CEO wealth gain (in terms of

her stock and option portfolio) is about $4.9 million ($84,000). To our knowledge, we are the

first to examine the relation between CEO compensation and the propensity to announce a stock

split. A potential criticism of our study is that stock splits tend to occur after significant price

run ups and the financial benef it to the CEO of doing a stock split is probably relatively small

compared to the recent wealth gains that she has experienced because of the stock price run up.

We cannot refute this claim, but we can compare the economic significance of our results with

those of other papers that examine CEO actions relative to their personal gain. For example,

Aboody and Kasznik (2000,) in their study of information disclosures around stock grants find

that CEOs have a median gain of about $18,500. They argue that this amount is an economically

3http://fortune.com/2012/02/23/no-apple-dividend-today/.

4http://www.businessinsider.com/why-tim-cook-decided-to-do-a-7-for-1-stock-split-2014-4.

5http://www.bizjournals.com/seattle/blog/2015/03/starbucks-announces-2-for-1-stock-split-investors.html.

6Specifically, Devos et al. (2015) find that timing grants before a split results in an average gain per CEO of about

$450,000, and selling stocks after the split nets on average about $345,000.