CEO Investment Cycles

Assets and employment grow more slowly early in the CEO's tenure than in later years.

The chief executive officer (CEO) and the top management team are widely viewed as critical to a company's success or failure, yet it is often difficult to identify specific actions by corporate leaders that affect firm fortunes. One way to address this question is to study how a firm's activities vary over the CEO's term in office. A CEO's incentives and power inside the firm are likely to vary over the course of his or her tenure, and this may lead to systematic differences in firm behavior. In CEO Investment Cycles (NBER Working Paper No. 19330), authors Yihui Pan, Tracy Yue Wang, and Michael Weisbach document patterns in corporate investment and disinvestment activities over the "CEO cycle" in a large sample of publicly traded U.S. firms.

The authors find that disinvestments are fairly common in the early years of a CEO's tenure, and that these disinvestments decrease with tenure. Investments, on the other hand, are relatively low in the early years of a CEO's tenure and increase over time. As a result, the firm's assets and employment grow more slowly early in the CEO's tenure than in later years. These cyclical changes in investment are observed using a variety of distinct measures of investment and disinvestment, and regardless of the reason for the CEO turnover, the CEO's background, or the industry conditions at the time of the turnover.

The authors find that the annual investment rate, measured as the investment-to-capital-stock ratio, is about 6 percentage points lower, and the asset growth rate is about 3 percentage points lower, in the first three years of a CEO's tenure than in later years. The median investment rate in their sample is 24 percent and the median asset growth rate is 7.6 percent. The magnitude of the "CEO cycle effect" is substantial, and is comparable to that of other factors that are known to influence investment, such as the business cycle, political uncertainty, and financial constraints.

The authors suggest that early in the CEO's tenure, the CEO disinvests poorly performing assets that his predecessor was unwilling to abandon. Subsequently, when the CEO gains more control of the corporate board, he may over-invest. Empirically, both the increase in the quantity and the decrease in the quality of investment appear to be a function of the CEO's growing control over the board during his tenure. There is no evidence that the investment cycles occur because of shifting CEO skill or productivity shocks.

--Les Picker