Print
E-MailWilson H.S. Tong is Professor of Practice at the School of Accounting and Finance and Faculty of Business at Hong Kong Polytechnic University. This post is based on an article forthcoming in the Journal of Accounting and Public Policy by Lin Li, Peter Lam, Professor Tong, and Justin Law. Related research from the Program on Corporate Governance includes Lucky CEOs and Lucky Directors (discussed on the Forum here) by Lucian Bebchuk, Yaniv Grinstein, and Urs Peyer; The CEO Pay Slice (discussed on the Forum here) by Lucian Bebchuk, Martijn Cremers, and Urs Peyer; and Golden Parachutes and the Wealth of Shareholders (discussed on the Forum here) by Lucian Bebchuk, Alma Cohen, and Charles C.Y. Wang.
Executive Summary
- Industry and firm-specific returns relate to CEO forced turnovers differently across industrial conditions.
- CEO forced turnovers relate more to idiosyncratic returns during recessions but more to industry returns during booms.
- Stock prices are more reflective of CEOs’ abilities during recessions than in booms.
Agency Problems & Executive Incentive Scheme
To mitigate agency conflicts in a corporation, a proper incentive scheme for the management is considered as essential. The positive incentive is the executive compensation package and the negative incentive is the disciplinary executive forced turnover. Presumably and understandably, such incentives should be linked to firm-specific performance, more precisely, firm performance in excess to the sector/industry performance (so-called relative performance evaluation (RPE)). Indeed, empirical studies in this line of literature typically show that such relative stock performance is positively correlated with executive compensation while negatively correlated with the likelihood of executive turnovers.
Should Executives be Fired due to Poor Industrial Performance?
Yet, an interesting but puzzling phenomenon is that such incentive schemes are found to be related to industry performance. (Please refer to Table 5 of our paper for executive turnover situation.) It is puzzling because optimal incentive contracts should depend only on factors within control of the executive, not on those that she cannot control.
There is no consensus on the explanation and if such decisions are optimal or not. For instance, Jenter and Kanaan (2015) view that firing a CEO under poor industrial performance is not optimal as the CEO has no control on the economic situation of the whole industry. Yet, Gopalan, Milbourn, and Song (2010) argue that industry performance is not entirely out of a CEO’s control as she should have the vision to choose the right strategy for deploying the firm’s assets so as to reduce firm’s exposure to sector movements. Indeed, they find that the asymmetry in pay for sector performance is present in firms with greater strategic flexibility and forced turnover likelihood is sensitive to sector performance only in firms with greater strategic flexibility.
Eisfeldt and Kuhnen (2013) provide a somewhat different explanation. The firm board considers a CEO based on multiple characteristics and makes the hiring decision if the CEO characteristics match the firm’s needs. When industry conditions change, the quality of firm-CEO matches in that industry deteriorates. The manager hired and worked under the old environment does not possess the skills necessary for the new situation and condition of the industry. As a result, firms demand managers with different skills and this leads to managerial turnover, provided the turnover benefit outweighs the cost.
Information Signaling as Our Explanation
Unlike such turnover explanations that are based on the cross-sectional differences in managerial skill sets or talents upon industrial downturn, the novelty of our study is to provide an explanation based on information signaling. We argue that managerial ability will reveal more clearly during industrial downturns than upturns so that the board of directors relies more on RPE to make disciplinary turnover decisions during industrial downturns than in upturns.
Our argument is conceptually consistent with agency models with multiple performance measures. Holmstrom and Milgrom (1991) and Banker and Datar (1989) argue that an optimal contract should place more weight on performance measures that are more precise and sensitive to the agent’s efforts. Empirically, Defond and Hung (2004) provide international CEO turnover evidence that in markets where stock performance is more informative, the turnover likelihood hinges more on the company’s stock performance. On the other hand, in markets where accounting performance is more informative, the turnover likelihood hinges more on the company’s accounting performance.
Yet, for our signaling argument that disciplinary turnovers are driven by the variation in the signaling quality of the performance indicators over the business cycle to be viable, two critical aspects need to hold. First, managerial ability needs to reveal more clearly in industrial downturn than upturn. According to Dessein and Santos’s (2021) cognitive theory on the variation of manager fixed effects on operations across industry conditions, a manager allocates her scarce attention on tasks differently under different environments. Under less complex situation when managerial attention is not scarce, it is optimal for her to pay attention to various tasks as there is little to be gained from focusing on one task. As a result, managers reduce or eliminate any differences in task knowledge and become generalist managers. Yet in complex environments when managerial attention is scarce, it is optimal for her to focus on the task she has expertise on. As such, managers with style will prevail in more complex environments.
To put in our context, to the extent that industrial downturn being a more complex environment than industrial upturn, Dessein and Santos’s (2021) model would suggest managerial style and talents to be more distinctly and easily observed by the board in bust than in boom periods. That is precisely what we have found empirically when using Demerjian, Lev, and McVay’s (2012) managerial quality score as the proxy measure of managerial talent. Specifically, the managerial quality score is found to be more associated with RPE during bust than boom periods, suggesting that managerial quality is indeed reflected more clearly in firm-specific performance during bust periods. (Please refer to Table 3 of our paper.)
The second, critical aspect for our argument to go through is informational efficiency of stock prices would not be weakened during downturns. If more uncertainty and/or more noise exists during industrial downturns, then the firm-specific stock performance would be less informative to the company board even though the firm performance depends more on the executive own style during this period, as suggested by Dessein and Santos (2021). In such case, it is not optimal for the board to fire their CEO based on firm-specific performance.
Yet, we look at two commonly used proxies for stock price informativeness in the finance and accounting literature, i.e. the R-squared measure of price synchronicity and the value relevance of accounting numbers, over the industry cycle. We observe that price synchronicity is approximately 30% smaller during recessions than during booms, while value relevance of accounting numbers is 30% larger. That is to say, stock prices carry more firm-specific information in recessions relative to that in booms. (Please refer to Table 2 C of our paper.) Our findings are consistent with prior evidence on stock price informativeness varying across economic cycles (Bostanci and Ordonez, 2022; Loh and Stulz, 2018).
Concluding Remarks
Based on our large sample of CEO turnovers from 1993 to 2019, we find that the probability of CEO forced turnover is more strongly associated with idiosyncratic stock returns during sector downturns than upturns while it is more strongly associated with sector performance during sector upturns than downturns. We argue with supportive evidence that such asymmetric turnover decisions over the industrial cycle are due to the fact that executive quality and ability are reflected more clearly during industrial downturns than upturns and, simultaneously, firm stock prices are also more informative during industrial downturns than upturns. As such, the board is optimal to make disciplinary turnover decisions based more on RPE during industrial downturns but based more on sector performance during industrial upturns. In fact, we find, as shown below but not inside our paper, that the likelihood of forced turnovers depends more on RPE when the stock prices are more informative and vice versa, which is consistent with Defond and Hung’s (2004) results.
