“Passing the Baton”: The effects of CEO succession planning on firm performance and volatility

• Date: 01 January 2019
• DOI: http://doi.org/10.1111/corg.12251
• Author: Hong Zhao,Ran Tao
• Published date: 01 January 2019

ORIGINAL MANUSCRIPT

“Passing the Baton”: The effects of CEO succession planning on

firm performance and volatility

Ran Tao |Hong Zhao

NEOMA Business School, France

Correspondence

Hong Zhao, Department of Finance, NEOMA

Business School, 59 Rue Pierre Taittinger,

Reims 51100, France.

Email: hong.zhao@neoma‐bs.fr

Abstract

Manuscript Type: Empirical

Research Question/Issue: This paper seeks to understand the way in which CEO

succession planning affects firm performance and volatility during CEO turnover.

Specifically, the paper examines a succession type referred to as relay succession, in

which the incoming CEO has been groomed as an heir apparent before the turnover.

Research Findings/Insights: Heirs apparent are identified by comparing all of a

firm's non‐CEO top executives' promotion likelihood estimated based on a set of

characteristics. Applying this heir apparent measure to a large sample of CEO

turnovers from ExecuComp, the paper delivers robust evidence that firms with relay

successions achieve higher post‐turnover accounting performance, higher long‐term

stock returns, and lower volatility. Further, the positive effect of relay succession on

performance is stronger for firms with higher human capital requirements.

Theoretical/Academic Implications: The paper provides comprehensive evidence

of the important role of relay succession in smoothing CEO transitions. It also sheds

new light on previous research on outside successions by showing that the difference

between inside and outside successions is explained largely by the length of relay.

Practitioner/Policy Implications: The paper suggests that firms should consider

using relay succession to improve performance and reduce volatility during the CEO

transition period. The paper's findings also support the US Securities and Exchange

Commission's recent appeal for more firm disclosure on its succession planning.

KEYWORDS

Corporate Governance, Heir Apparent, Relay Succession, Performance, Volatility

1|INTRODUCTION

The succession of a chief executive officer (CEO) is a significant event

for a firm, as it can lead to considerable changes in firm structure, per-

formance, and volatility (e.g., Clayton, Hartzell, & Rosenberg, 2005;

Denis & Denis, 1995; Huson, Malatesta, & Parrino, 2004; Weisbach,

1995). To achieve a smooth transition, an heir apparent is often desig-

nated several years before the expected turnover and works with the

incumbent CEO until she steps down. This type of succession planning

is referred to as relay succession, or sometimes as “passing‐the‐baton”

(Vancil, 1987). The CEO transition at Goldman Sachs in 2006

represents one such case. Lloyd Blankfein had been groomed as

Goldman's next CEO years before he took the helm, so when Henry

Paulson (the incumbent CEO) wastapped as Secretary of the Treasury,

Wall Street viewed Blankfein's elevation as a “foregone conclusion”and

a“seamless transition”(Enrich, 2006). A contrasting example is the CEO

turnover drama at Coca Cola in 2004. When Douglas Daft announced

his plan to retirement, the board split between promoting internal

candidates or searching for external candidates. The months‐long

turmoil at the top of Coca Cola ended with the loss of several of its

top executives and greatly increased investors' concerns about the

company (Terhune & Lublin, 2004; Terhune, McKay, & Lublin, 2004).

Received: 17 October 2017 Revised: 13 June 2018 Accepted: 15 June 2018

DOI: 10.1111/corg.12251

Corp Govern Int Rev. 2019;27:61–78. © 2018 John Wiley & Sons Ltdwileyonlinelibrary.com/journal/corg 61

The goal of this paper is to examine whether the benefits of relay,

as suggested by the anecdotes above, hold systematically across a

broad sample of firms. To begin with, ample theoretical evidence of

the benefits of relay exists in the extant literature. Large changes in

the top management team could lead to organizational disruption

(Hannan & Freeman, 1984; Helmich & Brown, 1972), and relay could

minimize the risk of such disruption during a CEO turnover by improv-

ing complementarity between the incoming CEO and other senior

managers (Hayes, Oyer, & Schaefer, 2006). In addition to disruption

minimization, relay also provides the incoming CEO with opportunities

to accumulate human capital (Zhang & Rajagopalan, 2004), and the

firm with time to test its heir apparent (Shen & Cannella, 2002).

Hence, we hypothesize that firms with relay should have higher

post‐turnover performance. Besides performance, our paper also

investigates the effect of relay on firm post‐turnover volatility, as high

volatility can depress firm investments, intensify agency problems, and

exacerbate conflicts between stockholders and bondholders. We

hypothesize that firms with relay have lower post‐turnover volatility,

as these firms face less uncertainty about their upcoming strategic

planning, their new CEO's quality, as well as changes in other senior

managers (Clayton et al., 2005).

One difficulty in studying relay succession is measuring its duration

(relay length), which is defined as the number of years an incoming CEO

has been groomed as an heir apparent before assuming the CEO posi-

tion. The conventional way in which the literature identifies an heir

apparent is based on whether the firm has a non‐CEO executive hold-

ing the president and/or COO titles (e.g., Naveen, 2006; Zhang &

Rajagopalan, 2004). However, ExecuComp data show that on average

only 30% of executives who ever hold the president/COO title eventu-

ally become a CEO, indicating that a president/COO is not necessarily

the chosen successor. Moreover, when the president and the COO

positions are held separately by two non‐CEO executives (in 25% of

firm‐years in ExecuComp), it is unclear who should be considered the

heir. To address these issues, for each non‐CEO top executive in a firm,

we estimate her probability of promotion to the CEO position based on

a set of characteristics (including titles, directorship, compensation,and

age), and define her as the heir apparent if she is the only executive in

that firm‐year with a sufficiently high probability of promotion. Com-

pared to the conventional president/COO measure, heirs apparent

defined by this new measure have a doubled realized probability of

becoming the next CEO. Further, approximately 40% of incoming CEOs

in the sample have at least 1 year of relay experience, which is consis-

tent with the prevalence of relay succession observed in the field

(Brickley, Coles, & Jarrell, 1997; Vancil, 1987).

Based on this heir apparent measure, we investigate the way in

which a firm's relay length affects its post‐turnover performance and

volatility using a sample of 2,542 CEO turnovers from ExecuComp.

Consistent with the hypothesis, we find that firms with relay succes-

sion outperform those without relay. Regression results imply that a

1‐year increase in relay length significantly increases a firm's pre‐turn-

over to turnover year change in operating return on assets by 0.008

(6% of sample mean). The positive effect of relay decreases, but

remains significant in years 1 and 2 after the turnover, and dissipates

in year 3. Similar results are obtained from a market‐based perfor-

mance analysis with calendar time portfolios. Portfolios consisting of

non‐relay firms do not have abnormal returns either before or after

CEO turnovers. On the other hand, portfolios consisting of relay firms

earn significantly positive abnormal returns after turnovers and not

before. We also find that the positive effect of relay on performance

is significantly stronger for larger, more diversified, more R&D intense,

and more M&A intense firms. This is consistent with the view that

relay successions help incoming CEOs accumulate human capital and

are, therefore, more valuable for firms with higher human capital

requirements.

We next examine relay's effect on volatility. The empirical results

again support our hypothesis. There is a significantly negative relation

between relay length and pre‐to post‐turnover volatility change. In

the turnover year, firms with a 1‐year relay have on average a 0.016

lower volatility (4% of sample mean) than do those without relay. Sim-

ilar to the results on performance, the effect of relay is strongest at

the time of turnover and decreases gradually thereafter.

We use several methods to carefully address the issue of

endogeneity. First, in all regressions, we use pre‐to post‐turnover

changes rather than post‐turnover levels as dependent variables. This

controls for firm pre‐turnover trends in performance/volatility and

helps address reverse causality. We also control industry‐wide trends

and mean reversions of firm‐specific factors using industry and

matching group benchmarks. Second, we rerun the main regressions

with the inclusion of firm fixed effects, which absorbs unobserved

but time‐invariant firm characteristics. The results from firm fixed

effects are similar to the main findings, although they are slightly

weaker statistically. Endogeneity still could arise, however, if omitted

variables vary over time. Therefore, our third method is to exploit dif-

ferences in relay length caused by exogenous CEO turnovers. Using

turnovers attributable to former CEOs' health issues or acceptance

of another position, we continue to find results on the effects of relay

consistent with those described above.

Although the paper focuses on the effects of relay succession, we

also investigate whether forced turnovers and outside successions

have any predictive power on performance and volatility above and

beyond relay. Compared to firms with voluntary turnovers, those with

forced turnovers are associated with lower post‐turnover perfor-

mance and higher volatility relative to pre‐turnover benchmarks. On

the other hand, we find that the predictive power of outside succes-

sions on performance and volatility, as shown in Clayton et al.

(2005) and Huson et al. (2004), disappears after relay length is

controlled. This implies that it is the relay that matters, not whether

the CEO is internal (without relay) or external.

The findings in this paper contribute to the literature in several

important ways. First, to the best of our knowledge, previous research

in the literature always measures heir apparent using president/COO

title (Behn, Riley, & Yang, 2005; Cannella & Shen, 2001; Fee & Hadlock,

2003; Naveen, 2006; Shen & Cannella, 2003; Zhang & Rajagopalan,

2004), which actually has a low predictive power of realized promotion

to CEO. We propose an alternative measure using readily available data

and estimation models, and show that our heir apparent measure

improves on the conventional president/COO measure in a significant

way. Second, our paper contributes to the study of relay succession.

There is in general a consensus among researchers that relay helps

firms achieve seamless succession and is therefore value enhancing

62 TAO AND ZHAO