Do Firms Use Interim CEO Position as a Testing Ground for CEO Candidates?

Transcription

1 Do Firms Use Interim CEO Position as a Testing Ground for CEO Candidates? Xiaoxiao He Southwestern University of Finance and Economics Margaret Rui Zhu City University of Hong Kong Dec 18, 2017

2 Conventional Managerial Role of Interim CEO Seat-warmer: A person appointed by a company s board to assume the role of chief executive officer temporarily during a time of transition or as the result of the sudden departure of the company s previous CEO Reaction to sudden CEO departure: maintain daily operation and make everything ready for new formal CEO (Farquhar, 1995; Mooney et al., 2012) Internal discipline: enable firms to replace poor performed CEOs even without suitable successor (Ballinger and Marcel, 2010)

3 Conventional Managerial Role of Interim CEO Interim CEO might destroy firm value: Leadership Vacuum: compared with firms without interim period, firms with interim period are underperformed in the long-run. (Ballinger and Marcel, 2010; Liang et al., 2012) Uncertainty of Future Leadership: hold the position until a formal CEO successor is identified Limited CEO Authorities: maintain daily operation

4 Recent Trend of Using Interim CEO A large fraction (1/3) of interim CEOs are appointed as formal CEO after interim period, i.e. Steve Jobs Surprising higher performance during interim period (Business Wire, 2006, titled Interim CEOs on the rise and Underestimated )

5 Recent Trend of Using Interim CEO

6 Recent Trend of Using Interim CEO

7 Try-out Theory Firm use interim CEO position as a test-ground for CEO candidate: use interim CEO position to test potential CEO contenders who have uncertain leadership credential (Liang et al, 2012; Mooney et al., 2012). Motivation: firms board do not have fully confidence of the candidates leadership qualification Incentive: fail the test will dramatically reduce future promotion opportunity Criteria: to convince the board, interim CEO need to show their managerial ability during the interim period Outcome: the likelihood of being named as official CEO increases as the increase of interim period firm performance

8 Empirical Prediction Try-out Theory Prediction and Hypothesis tested: H 1 : Candidate with higher uncertainty of leadership. qualification are more likely to be named as interim CEO H 2 : Interim CEOs take non-passive activities during the interim period. H 3 : Interim CEOs probability of being promoted will be positively related to interim period firm performance. H 4 : Failing the try-out process will decrease the likelihood of future promotion so that managers are more likely to leave the firm.

9 Summary of Baseline Result Supportive evidence for the try-out theory H 1 : Candidates with uncertain managerial qualification (Not Heir Apparent) are more likely to be named as interim CEO. H 2 : Both promoted and not-promoted interim CEO take non-passive activities during the interim period. H 3 : Interim CEOs are more likely to be promoted to formal CEOs if the firm performs better during interim period.

10 Summary of Baseline Result Supportive evidence for the try-out theory H 4 : Within the group of failed interim CEO, interim CEOs who previously served as firms top managers and who have less managerial power are more likely to leave the company within 12 months after the interim period. And the relationship concentrates on interim CEOs whose ages are not close to retirement age.

11 Alternative Explanation Difficulties in attracting optimal CEO successors, e.g. firms promote interim CEO when they cannot find optimal CEO successors in the labor market. Promoted interim CEOs is sub-optimal and will be fired as soon as the firm finds an optimal one: Promoted interim CEOs do not suffer abnormally higher forced turnover rate during their official CEO tenure than their peers.

12 Robustness: Endogeneity Unobserved firm characteristics affect promotion decision. Two-stage OLS model use industrial performance as IV, as industrial performance are less likely be affect by its affiliations The effect of interim period firm performance maintains.

13 Contribution to Literature Find supportive evidence for the try-out theory on the motivation of using interim CEO. Provide additional evidence that performance is used to evaluate (Interim) CEOs (Fee and Hadlock, 2003). Contributes to the general CEO turnover literature: Previous studies in CEO turnover mainly focus on the outgoing CEOs and the incoming CEOs.

14 Empirical Predications of Try-out Process Prediction Model Dep Var. Indep Var. Contender H1 Probit D(Interim) 1 Heir Apparent 2 - H 2 Probit D(Promoted) 3 Interim period performance Dummy var. with one if the candidate are named as interim CEO first rather than official CEO and zero otherwise 2 Dummy var. with one if the candidate previously served as firms heir apparent and zero otherwise 3 Dummy var with one if the interim CEO are named as official CEO 4 Stock performance during interim period

15 Sample Treatment Sample: CEO succession with interim period hand collected data 1. Identify interim CEO position in all non-financial US listed firms from BoardEx. The sample period is 1992 to Interim CEO tenure in all observations must be completed and formal CEO successor is identified by Further divide treatment sample into Group Promoted: firms that promote their interim CEO into formal CEO position Group Not Promoted: firms whose interim CEOs fail the test.

16 Sample Control Sample: CEO succession without interim period (Group Control) 1. Identify CEO succession from Execucomp based on the change of CEO identification from fiscal year 1993 to Merge with turnover type dataset provided by Prof. Drik Jenter, Prof. Florian Peters, and Prof. Alexander Wagner. 3. Merge with BoardEx dataset and manually exclude succession with interim period.

17 Main Variable Motivation of using interim CEO Uncertainty of leadership qualification - Heir apparent: dummy var. with one indicates the candidate is a non-ceo executive held the title of president or COO who is at least 5-year younger than outgoing CEO, following Cannella and Shen (2001). Conventional managerial function - Emergency : dummy var. with one indicates sudden CEO departure due to death or medical issue - Forced Turnover : dummy var.with one indicates forced turnover of previous formal CEO follows Parrino (1997)

18 Main Variable Interim period performance - BHAR - Mkt-adj CAR: market-adjusted cumulative return - Mean Return: average periodical return scaled by previous 12-month return standard deviation

19 Propensity of Using Interim CEO: Univariate Test With Interim Without Interim Dif[(1)-(2)] Mean Mean Dif(Mean) T-Statistics Heir Apparent *** [-14.83] Emergency *** [4.25] Forced Turnover *** [10.59] Homogeneity * [1.81] CEO Experience ** [2.47] Outsider *** [-9.29] Age *** [6.01] Female [0.49] Log(AT) *** [-19.68] Debt Ratio ** [-2.24] ROA *** [-10.76] Return *** [-5.68] Sale Growth [0.82] Volatility *** [9.69] Institutional Ownership *** [-8.87] Board Size [0.99] Indep.Director [0.67] Busy Director *** [-3.73] Nobs

20 Propensity of Using Interim CEO: Probit Regression Dep Var.: D(Interim)=1 if candidate is named as interim CEO and =0 if named as formal CEO Model 1 Model 2 Model 3 Model 4 Heir Apparent -1.17*** -1.08*** [79.72] [61.44] Emergency 1.29*** 1.44*** [25.41] [28.39] Forced Turnover 0.59*** 0.50*** [40.70] [26.29] Controls ear F.E. Industry F.E. Pseudo R Nobs

21 Likelihood of Promotion: Univariate Test Promoted (1) Not Promoted (2) Dif[(1)-(2)] Mean Mean Dif(Mean) T-Statistics BHAR ** [2.02] Mkt-adj CAR [1.46] Mean Return [1.01] Period * [-1.70] No.Position * [-1.77] Forced Turnover [-1.38] Duality *** [-3.19] Top Manager [1.42] D(Outsider) *** [2.92] CEO Experience [-0.28] Age *** [-2.69] Female [-0.32] Log(AT) [-1.05] Debt Ratio [-1.29] ROA [0.35] Return [-1.22] Sale Growth ** [-2.06] Volatility [1.61] Institutional Ownership [0.16] Board Size [0.50] Indep.Director [-0.80] Busy Director [-1.32] Nobs

22 Likelihood of Promotion: Probit Regression Dep Var. =D(Promoted) Model 1 Model 2 Model 3 BHAR 0.48** [6.11] Mkt-adj CAR 0.41* [3.52] Mean Return 0.24** [3.88] Controls ear F.E. Industry F.E. Pseudo R Nobs

23 Distinct Managerial Effort Variable Promoted Not Promoted Mean T-statistics Mean T-statistics Panel A: Activity Equity Issuance 0.48*** [4.06] 0.35*** [3.93] Debt Issuance 0.96*** [3.35] 0.68*** [4.50] Capital Expenditure [-0.11] -0.01*** [-3.15] Inventory Turnover Rate 1.75* [1.78] 0.21 [0.42] R&D [-0.32] [-0.43] Operating Margin 3.13 [1.25] 5.91 [1.42] Debt Ratio [-0.18] 0.02* [1.81] Panel B: Performance Asset Growth 0.00 [0.09] -0.02** [-2.22] Sale Growth 0.04 [0.72] 0.00 [0.05] ROA 0.02 [1.56] -0.01* [-1.86] EPS 0.15* [1.71] -0.13** [-2.32] Nobs

24 Endogeneity Model 1 Model 2 Model 3 OLS Probit OLS Probit OLS Probit Excessive Industry Ret. 5.85*** 7.25*** 8.41*** [3.18] [4.60] [2.99] B^HAR Mkt^-adj CAR 2.87** [4.12] 2.32** [4.12] Mea^n Return 2.00** [4.12] Control ear F.E. Industry F.E. R 2 /Pseudo R Nobs

25 Difficulty in Attracting Optimal CEO Successor Panel A: Control=Directly Named CEO Model 1 Variable Force Voluntary Promoted 0.10 [0.15] Control Performence (t 1,t 2) Adj. Performence (t 1,t 2) ear F.E. Industry F.E. Pseudo R 2 Nobs N *** [6.83] N Model 2 Force Voluntary 0.11 [0.18] N *** [6.88] N

26 Difficulty in Attracting Optimal CEO Successor Panel B: Control=New Formal CEO in Not-Promoted Group Model 1 Model 2 Variable Force Voluntary Force Voluntary Promoted [0.72] Control Performence (t 1,t 2) Adj. Performence (t 1,t 2) N ear F.E. Industry F.E. Pseudo R Nobs [0.65] N [0.82] N [0.60] N

27 Managerial Departure of Not-Promoted InterimCEO Dep Var. = Prob of managerial leave within 12-month after interim period Age<64 Age 64 Variable Model 1 Model 2 Model 3 Model 4 Model 5 Top Manager 0.89*** 0.77*** 0.91*** [11.00] [7.58] [7.70] [0.01] Chairman -0.79** -0.58* -1.16** [6.31] [2.90] [4.68] [0.00] Control ear F.E. Industry F.E. Pseudo R Nobs

28 Conclusion We find supportive empirical evidence for try-out theory. Instead of using interim CEO as seat-warmer, firms use interim CEO position to test the managerial credential of potential CEO candidates. Firms are more likely to appoint candidate with uncertain leadership ability as interim CEO Interim CEOs actively manage the firm during interim period. Interim CEOs likelihood of being promoted to the formal CEO position is positively correlated with interim period firm performance.