How long should CEOs stay?

Professor Paul Kerin explains why boards should regularly and carefully review the performance of a chief executive officer.

It is often said that a board’s most important job is to appoint a chief executive officer (CEO). However, an equally important, and more difficult, decision is when to end a CEO’s tenure.

Boards make that decision more often than is commonly thought. Reported reasons for CEO exits suggest that around 20 per cent are “forced” by boards. However, econometric estimates indicate that about 40 per cent of exits are performance-related. Even the right CEO may not be “right” forever. On average, shareholder value creation is hill-shaped with respect to CEO tenure. Extra tenure generates diminishing incremental benefits but rising incremental costs. Incremental benefits are greatest in the early years. Incremental costs rise with tenure. Studies find that optimum tenure is typically eight to 12 years.

Economists sometimes do macabre things to prove a point, like studying share price reactions to unexpected CEO deaths. On average, share prices dive by 2.3 per cent if a CEO’s tenure is “short” (under 9.5 years), but rise by 1.4 per cent if it is long.

Boards and CEOs generally get it about right. The tenures of the most recent CEOs of Australia’s top 10 companies at exit ranged from five to 23 years; the median was 6.6 years. Across the market, the median is about eight years; 90 per cent of CEOs stay less than 12 years.Of course, boards don’t always select the right CEO. Even with excellent due diligence, boards and markets will always be uncertain about how well a new CEO will actually perform. Key uncertainties include the CEO’s abilities and fit with the firm, its current environment and the opportunities/challenges it faces.

Boards, markets and CEOs all attempt to manage these uncertainties. Boards monitor CEO and firm performance more when uncertainty is high. On average, boards hold more meetings in a CEO’s first three years. As boards gain more certainty, they act. Forced exits are most likely to occur within the first four years. Boards generally get retention decisions right. The average annual total shareholder return (TSR) over a CEO’s tenure is 9.6 per cent higher for CEOs lasting eight years than those lasting four years.

If, after gaining more certainty, boards retain their CEOs, they rationally make different choices. With increased confidence, boards grant CEOs more autonomy, reduce monitoring and react less to performance blips. Short-term profits have less impact on the probability of exit for higher-tenure CEOs, although it still has some impact. While board monitoring falls with tenure, it rises when TSR falls. In years with forced CEO exits, the frequency of board meetings is 17 per cent higher.

Market participants also gain more certainty about new CEOs by scrutinising firm and CEO performance. Share price reactions to news are stronger for early-tenure CEOs, as the market is unsure whether news reflects CEO ability/fit or other factors. As a CEO becomes more of a “known quantity”, share price volatility falls. On average, uncertainty about CEO ability/fit accounts for 25 per cent of total TSR volatility when a CEO is appointed. However, TSR volatility falls by 10 per cent over the first three years.

Like boards, market participants adjust their behaviours once they gain more certainty. If they decide a CEO is high quality, they mark the shares up and react less strongly to new information; if not, they mark the shares down and may agitate for change. New CEOs have strong incentives to convince their boards and the market that they are high quality. Doing so means they keep their jobs, improve their career outlooks and receive greater compensation and autonomy. These incentives affect their behaviours. Early-tenure CEOs are more likely to work harder and manage discretionary accruals and expenses to boost short-term earnings.

However, every situation is different. Some CEOs are incredible leaders and can reinvent themselves as necessary. CEO fit may change at different rates. Optimum CEO tenure in slow-moving industries is almost double that in fast-moving industries. Turnaround specialists have shorter optimum tenures than growth specialists. If they haven’t achieved a turnaround within several years, they’ve failed and are fired. Conversely, a successful turnaround does a turnaround specialist out of a job; a successful growth strategy doesn’t. Boards should regularly and carefully review whether or not to retain even high-performing CEOs.