CLS Blue Sky Blog

What Does CEO Turnover Data Tell Us About Succession Planning?

Over the past several decades, researchers have taken a serious look at the quality of CEO succession planning at publicly traded corporations. The results have not been encouraging. The evidence suggests that many companies are slow to terminate an under-performing CEO, are caught flat-footed in the event of a sudden CEO departure, and are often unprepared to identify a viable or permanent successor.

The research, however, is not without its shortcomings. A central challenge facing researchers is that it is very difficult for outside observers to determine whether the board in fact terminated the CEO. Rarely does a board explicitly state that the CEO was fired. Instead, shareholders are much more likely to hear that a CEO retired or left to pursue other opportunities, to spend more time with family, or for personal reasons. From statements such as these, outsiders must read between the lines before categorizing terminations as voluntary or involuntary, and for researchers dealing with hundreds or thousands of turnover events, clear-cut determinations are difficult.

In a recent paper, we examine CEO turnover and succession using a unique and highly comprehensive data set from the company known as exechange. Our sample includes 1,399 CEO turnover events at Russell 3000 companies over the five-year period from January 1, 2017 to December 31, 2021. Exechange applies a novel methodology that evaluates the circumstances surrounding a CEO departure to determine the degree to which it might be voluntary or involuntary (called the Push-Out Score). With this data, we are able to provide new statistics on CEO departures and fresh analysis of CEO turnover to arrive at a more current assessment of the quality of succession planning among publicly traded U.S. companies.

CEO Departures

Description of Push-Out Scores

The premise behind the Push-Out Score is that CEO departures are not necessarily black-and-white decisions. Instead, the Push-Out Score comprises a scale of 0 to 10 to reflect that departure decisions often fall on a spectrum. A score of 0 indicates that it is “not at all likely” the CEO was terminated, and a score of 10 indicates termination is “evident.” Numbers between these reflect the extent to which circumstantial evidence indicates the CEO was pressured to leave.[1]

Distribution of Push-Out Scores

First, we examine the distribution of Push-Out Scores. We see they are somewhat evenly distributed across the scale. This demonstrates the inherent difficulty of categorizing CEO departure decisions as voluntary or involuntary and shows how many turnover events appear to fall somewhere in the middle. Of note, high scores are slightly more prevalent than low scores (see Figure 1).

Figure 1: Distribution of Push-Out Scores (2017-2021)

Source: Data from exechange.

This is unexpected. Most traditional research finds that CEOs terminations are less frequent than voluntary resignations, whereas Push-Out Scores suggest that terminations are more frequent. Boards might be more willing to terminate a CEO than researchers have historically recognized.

Stated Reason for CEO Departures

Next, we look at the official reason companies provide in press releases to explain CEO departures and the relation between these and Push-Out Scores. The most frequent explanation companies provide is that the CEO “retired,” followed by “resigned,” and “stepped down.” Other descriptions are “left, departed, or no longer serve at the company;” “terminated, removed, or separated from the company;” and “transitioned” leadership (see Figure 2).

Figure 2: Stated Reason for CEO Departures

Source: Data from exechange. Analysis by the authors.

All of these phrases – with the exception of termination language – are associated with a range of Push-Out Scores. Only the phrases “terminated, removed, or separated from the company” are clearly indicative of involuntary departure (see Figure 3).

CEO “Retired”

CEO “Stepped Down”

CEO “Resigned”

CEO “Left, Departed, or No Longer Serves”

CEO “Terminated, Removed, or Separated from the Company”

Figure 3: Stated Reason for Departure and Push-Out Scores

Source: Data from exechange. Analysis by the authors.

These patterns underscore the ambiguity of language used to describe CEO departures and demonstrate why outsiders have difficulty relying on company-provided explanations to determine whether a CEO was fired or resigned.

Sensitivity of Turnover to Performance

Here we evaluate the relation between Push-Out Scores and the short- and long-term stock-price performance of companies leading up to the CEO departure data to gain insight into the sensitivity of CEO turnover to performance. To adjust for the fact that the Push-Out Score model includes recent stock-price performance as an input, we compute a modified score on a scale of 0-9 that strips out this variable.

We find a strong association between a lower stock price and a greater likelihood that a CEO is pressured to leave. This indicates that Push-Out Scores provide informative assessments of whether a termination event is involuntary. It also indicates that boards might be more likely to hold CEOs accountable for performance than prior studies suggest (see Figure 4).

Figure 4: Cumulative Abnormal Returns (CAR) Prior to CEO Departure (Long Windows)

Source: Data from exechange and Center for Research in Securities Pricing (CRSP). Analysis by the authors. Push-Out Scores modified by the authors to remove recent stock-price performance as an input variable.

Push-Out Scores are also related to stock-price performance in short trading windows around the departure announcement date. Low scores are associated with very modest abnormal returns in the five-day and one-day periods surrounding announcement, while high scores are associated with large abnormal losses over these periods (see Figure 5).

Figure 5: Cumulative Abnormal Returns (CAR) Around the CEO Departure (Short Windows)

Source: Data from exechange and Center for Research in Securities Pricing (CRSP). Analysis by the authors.

Preparedness to Name a Successor

Here, we examine the degree to which companies are prepared to name a permanent successor and whether forced terminations (high Push-Out Scores) are associated with lower levels of preparedness.

While most researchers identify two types of succession (permanent and interim successors), in fact three categories exist:

67 percent of succession events involve the naming of a permanent successor, 10 percent a delayed-permanent successor, and 22 percent an interim successor.

The choice of an interim successor is associated with significantly lower  stock price in comparison with the other two choices over all long-term measurement windows leading up to the departure announcement. Delayed-permanent successors are also associated with a lower stock price than are permanent successors over three-year and one-year measurement periods. This confirms earlier research studies that find that companies in worse financial shape are less likely to name a permanent successor.

Succession choices vary by Push-Out Score. Low Push-Out Scores almost always involve naming a permanent CEO. Middle scores primarily involve permanent successors but also a mix of delayed permanent and interim. High scores are fairly evenly split between interim successors and permanent successors but rarely involve a delayed permanent successor (see Figure 6).

Figure 6: Successor Type by Push-Out Score

Source: Data from exechange and research by the authors. Analysis by the authors.

This suggests that most companies that name an interim successor do so because the CEO was terminated and not because the CEO resigned unexpectedly. It also suggests that boards in these situations are unprepared with a long-term succession plan, despite the fact that the board precipitated the turnover event.

Internal versus External Successors

Finally, we analyze decisions to appoint internal or external executives as permanent successors. Over our measurement periods, 59 percent of eventual permanent successors come from within the company, 33 percent are sourced from outside, and 8 percent are a nonexecutive director serving on the board.

The choice of an internal successor varies by Push-Out Score. When the CEO leaves voluntarily, his or her successor is overwhelmingly likely to be an internal executive (89 percent). With high Push-Out Scores, internal replacements are chosen only 41 percent of the time. This is consistent with the view that internal successors are more likely to be named when the CEO departs voluntarily, and external successors are more likely to be chosen following a forced termination or when pressure to resign was greater.

Lastly, we study the stock-price performance of successors by Push-Out Score and by type of successor. We find that, when the successor is chosen from the board, the announcement generally prompts a  lower stock price than does an announcement of internal and external successors and that board successors exhibit worse three-month and six-month returns following their appointment.

Why This Matters

  1. One of the fundamental responsibilities of the board is to hire and fire the CEO. We find in a third of cases that boards are unable to name a permanent successor when the departing CEO steps down. Does this indicate these boards have not taken succession planning seriously?
  2. We find that most companies name an interim CEO when the CEO was terminated rather than resigned voluntarily. What does this say about succession planning?
  3. In a non-trivial percent of cases, companies turn to a current board member as long-term successor to the outgoing CEO. Is this indicative of good succession planning because of the board member’s knowledge of the company or poor planning because a more suitable internal candidate could not be found?
  4. We also find that, when a current board member is chosen, the company tends to perform worse than when an internal or external candidate is selected. Is this due to the selection of the board member as CEO or because of serious strategic or operating issues at the company?
  5. The stock market clearly pays attention to succession events. How might shareholders use Push-Out Scores to inform investment decisions around CEO changes?
  6. Our data support the concept that CEO resignations are not simple binary events in which the CEO either was or was not terminated. Instead, CEO resignations appear to vary based on the pressure the board exerted on the executive to resign. How would the concept of board pressure, rather than strict termination, change how researchers evaluate the sensitivity of CEO turnover to performance, board preparedness to name a successor, and the source of that successor?


[1] For a more detailed discussion of the Push-Out Score methodology, see

This post comes to us from David F. Larcker and Brian Tayan at Stanford University Graduate School of Business and Edward M. Watts at Yale School of Management. It is based on their recent article, “Firing and Hiring the CEO: What Does CEO Turnover Data Tell Us About Succession Planning?” available here.

Exit mobile version