CEOs occupy the quintessential position of modern corporations and have a vast influence on the firms they lead. At the same time, the CEO job is generously rewarded. In 2019, the average compensation of S&P 500’s CEOs amounted to $14 million – a figure which has grown by about 50% since 2009 (Roe 2019). Such remarkable trends have fuelled an intense debate among investors and the media on the ‘value for money’ of high-paid CEOs (Economist 2020).
Scholars have argued that CEOs often find ways to assign themselves high pay at the detriment of shareholders. A notable example is provided by superstar CEOs (referring to those celebrated in media outlets), who have been shown to enjoy higher compensation while their firms underperform (Malmendier and Tate 2009). In contrast, traditional agency theories suggest that CEO compensation is a tool to incentivise CEOs to increase shareholder wealth. Thus, sizable compensation packages reflect a significant contribution of CEOs to firm performance and the ample availability of outside options in the managerial labour market. Accordingly, the managerial labour market efficiently allocates high compensation to highly demanded CEOs and, at the same time, introduces competition, disciplining CEOs’ behaviour and aligning pay and performance.
Our recent study (Amore and Schwenen 2020) suggests that the labour market may actually provide a vehicle for CEOs to derive personal benefits unrelated to their managerial effort. This is especially true when hiring firms are subject to weak corporate governance. Finally, we show that there are significant costs of hiring CEOs whose pay at their previous firm did not reflect their managerial effort.
To conduct the analysis, we merge data on executive compensation with accounting data for S&P 1500 firms. The data allow us to track the mobility of CEOs within the S&P 1500 sample. We begin by devising the part of CEOs’ pay that is not related to their individual contribution to firm performance. To this end, we follow a common approach (Bertrand and Mullainathan 2001) and identify changes in firms’ market value due to factors which are not related to operational or strategic measures taken by the CEO, such as oil prices or business cycle. Simple principal-agent notions suggest that an optimal compensation aimed at incentivising managerial effort should not reward CEOs for these exogenous ‘lucky’ events. However, extant results show that pay-for-luck is prevalent (Bertrand and Mullainathan 2001, Davis and Hausman 2020).
Our results indicate that luck not only increases CEOs’ pay at their current firm, but also affects CEOs’ prospects in the labour market. First, we show that luck increases the likelihood of CEO transitions: CEOs whose firm value rises due to luck are more likely to leave their company. Conditional on venturing into new positions, luck at the departing firm makes CEOs more likely to be jointly appointed as CEO and board chairman of the hiring firm. Moreover, CEOs’ luck at the departing firm bolsters the compensation that they get at the new firm. Figure 1 depicts the positive association between a firm’s luck (at a CEO’s firm of departure) and the compensation at his/her new firm.
Figure 1 Luck shocks and new CEO compensation
Note: This figure illustrates the linear relationship between a CEO’s luck at his/her departing firm and (the logarithm of) CEO compensation at the new firm.
Importantly, this association is driven by non-cash compensation items (e.g. stocks and options) rather than cash-based items (i.e. salary and bonuses). This finding lends support to the view that lucky CEOs skim hiring firms on the job market by bargaining on pay items that are easier to conceal (Frydman and Saks 2010). In addition, we find that lucky CEOs who move to new firms earn significantly more than their new industry peers.
Extant literature shows that CEOs – especially those in weakly-governed companies – can increase their pay by influencing the pay-setting process in their favour (Bertrand and Mullainathan 2001, Garvey and Milbourn 2006). Hence, we probe into the corporate governance of hiring firms to shed light on the mechanisms driving the matching between hiring firms and lucky CEOs. Our results indicate that lucky CEOs move more frequently towards firms featuring low analyst coverage, which makes them less equipped to evaluate and discipline executives. Moreover, lucky CEOs move more frequently towards firms operating in less-competitive industries, which are typically subject to laxer corporate governance and provide more room for managerial entrenchment.
What do these findings imply for the shareholders of the hiring firms? On the one hand, appointing a lucky CEO may be harmful for shareholders due to a misalignment between his/her pay and firm value. On the other hand, it may be desirable for a firm to attract a talented CEO whose luck may have made retention too costly at his/her previous firm. To separate out these interpretations, we compare the profitability of firms that lucky and unlucky CEOs move to. As shown in Figure 2, our findings indicate that greater CEO’s luck at his/her previous company is detrimental to the hiring firm’s performance.
Figure 2 Performance change around CEO transitions
Note: This figure illustrates the average profitability (ROA, i.e. EBIT/total assets) in each of the four years before and after a CEO appointment separately for incoming low-luck and high-luck CEOs (i.e. CEOs above or below the median luck at their previous firm).
The exposure of a CEO to luck at his/her departing firm is exogenous to the prospects of the hiring firm. However, the formation of new matches between hiring firms and lucky (or unlucky) CEOs may be driven by unobserved factors that, in turn, correlate with future performance. While we control for constant heterogeneity across firms, we reinforce the causal interpretation of our results by showing the absence of pre-turnover diverging trends in the performance of firms that hire low-luck and high-luck CEOs.
Why do firms that hire lucky CEOs subsequently underperform? To answer this question, we explore a number of outcomes related to efficiency and investment policy. Our results indicate that firms hiring lucky CEOs experience an increase in operating costs and a drop in the ability to generate revenues out of the asset base. By contrast, these firms do not exhibit changes in investment policies, as indicated by an insignificant effect of incoming CEOs’ luck on R&D, capital expenditures, and asset growth.
Collectively, our results indicate that lucky CEOs harm firm performance by enjoying the ‘quiet life’ rather than engaging in empire building (Bertrand and Mullainathan 2003). In this regard, another takeaway of our study is that the role of the labour market as a disciplining device for CEOs is imperfect as it also depends on the governance quality of the hiring firms. A ‘quiet life’ is possible whenever proper corporate governance structures are lacking. Our results point to the existence of informational frictions in the hiring process of weakly-governed firms, and underscore the importance of accurately searching for the right CEO profile in order to enable shareholders to secure the benefits of turnover at the helm of the company.