CEO turnover, leadership vacuum, and stock market reactions

ABSTRACT

CEO departures with a delay in successor appointment create a leadership vacuum inducing operational disruption and strategic uncertainty. Such departures also produce turnaround benefits from cutting ties with a poorly performing CEO and by allowing additional time to search for a qualified successor. Prior studies fail to disentangle these perceived costs and benefits associated with CEO dismissal. After filtering out the turnaround benefits, we find that the market reacts incrementally negatively to CEO departure announcements with a delay in successor appointment than those without such delay, capturing incremental switching costs caused by a leadership vacuum. We also find that the leadership vacuum cost is larger in a more volatile environment or with abandonment of a relay succession plan. Our findings contribute to CEO turnover literature by suggesting that a temporary leadership vacuum is an indicator of abandonment of a succession plan that has been influenced by a poorly performing CEO and that such abandonment creates an opportunity to achieve performance turnaround through a better successor.

KEYWORDS:

• CEO turnover
• leadership vacuum
• turnaround benefits
• switching cost
• stock market reactions

JEL CLASSIFICATION:

• G14
• G30
• G34
• M12

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ The expected costs and benefits of a leadership vacuum could also lead to a board’s decision to retain, rather than replace, an incumbent CEO. However, we cannot test market reactions to the board’s CEO retention decision because the decision date is generally unknown to researchers. Therefore, our study focuses on the board’s selection between a DO and DA announcement and stock market reactions to these announcements.

² We find that 80.3% and 15.4% of heirs apparent in DO- and DA-announcing firms are not appointed as a successor, corroborating our argument that DO announcements signal the abandonment of a relay succession plan.

³ Similarly, as the unobservable portion of the contract termination value decreases, the probability of selecting a DO over a DA announcement decreases, implying a decreasing $u_i$ and ${\epsilon }_{0i}$ (and thus ${\rho }_{u\epsilon 0}>0$).

⁴ Similarly, a positive ${\rho }_{u\epsilon 0}$ is moderated by the unobservable innate candidate supply abundance and efficient succession planning that decrease the probability of a DO over a DA announcement (i.e. a lower $u_i$), reduce search costs (i.e. a lower ${\upsilon }_{sc,i}$), and increase market reaction to the DA announcement (i.e. a higher ${\epsilon }_{0i}$)$.$

⁵ In a more volatile operating environment, a new permanent CEO, elected on or after the departure of an incumbent CEO, may produce more uncertain performance during her early tenure during which he or she must adapt to a new environment. However, such performance uncertainty does not stem from her short horizon and limited authority to make long-term decisions. In a competitive labour market, such performance uncertainty may be incorporated into her initial compensation contract.

⁶ The departure of Douglas Daft, the former Chairman and CEO of Coca-Cola Co., is a case in point (Terhune and Lublin 2004). On 19 February 2004, the company announced the CEO’s plan to retire and began an outside search for a successor. President and COO Steven J. Heyer, the heir apparent, was not selected as the successor. The three-day announcement period size-adjusted returns for Coca-Cola were -3.05%.

⁷ To avoid repetition, we do not tabulate the results of the first-stage regression using the 2SLS method, which are similar to those using the FIML method.

⁸ When we also add AGE and AGEDUM to market reaction regression, these variables do not come in significant and our primary results remain unchanged.

⁹ The difference in intercepts becomes larger and regains significance when we exclude RTNSTD and RD_INTENSITTY from the fully augmented market response equations in Panel B of (Table 5). This (untabulated) result indicates that these two variables capture key characteristics of the cross-sectional variations in leadership vacuum-induced switching costs.

¹⁰ The U.S. firms designate a non-CEO COO as an heir apparent, suggesting that they consider hierarchy order disturbance. Studies regarding CEO successions with hierarchical jumps within top management team in the Chinese setting report that CEO successions with hierarchical order disturbance reduces agency costs (Sarfraz et al. 2020; Shah et al. 2019) and interact with board independence to increase CSR activities (Shah, Sarfraz, and Ivascu 2021).