The impact of CEO quality on firm decisions is a question that has not been much explored from an accounting point of view. Yet there is strong evidence that CEO reputation can matter here. For instance, manager reputation and career concerns have been shown to affect the assessments of users of financial statements on reporting quality, as well as the reporting decisions of CFOs. Anecdotal evidence also suggests highly-reputed managers may engage in earnings management to maintain their reputation.

Given these factors, Allen Huang and Amy Zang of HKUST and their co-authors Jennifer Francis and Shivaram Rajgopal set out to see if there was a way to measure the impact of CEO reputation on earnings quality. They were careful to distinguish between innate earnings - the features attributable to a firm's operating environment and production technology - and discretionary earnings, which can be influenced by actions taken in the short term.

What they found was an interesting interaction between the needs of firms and the abilities of reputed CEOs. Companies with poor earnings quality were more likely to turn to CEOs with a strong reputation, and there was some evidence these hirings could help to improve discretionary earnings quality.

The findings were based on about 2,000 firm-year observations for S&P 500 firms over 10 years, from 1992-2001. Proxy measures were carefully teased out for CEO reputation and earnings quality because neither of these is directly observable. For CEO reputation, the authors relied on coverage in major US and global business newspapers and business newswires. Earnings quality was measured by looking at the standard deviation of firm-specific residuals and a firm's five-year average of absolute value of performance-matched abnormal accruals (large deviations and large accruals indicated poorer total earnings quality).

The authors had hypothesized three possible outcomes for the data. First was that CEO reputation would be associated with better discretionary earnings quality because CEOs had more to lose in terms of credibility and future wages if they systematically portrayed their firm in a more favorable light than warranted. While that concern may exist, the measure of CEO reputation was not linked to better discretionary earnings quality, but instead to poorer earnings quality.

The likely reasons for this were embodied in the authors' other two hypotheses: rent extraction, and matching of firms to CEOs. Both could result in poor discretionary earnings, but with rent extraction, this would be the result of the CEO over-emphasizing his or her personal career enhancement and thereby taking actions that would worsen discretionary earnings quality. With matching, firms that have poor innate earnings to begin with would hire reputed CEOs, which would induce the association between CEO reputation and poor earnings quality.

Numerous tests pointed to matching as the most likely explanation for the link between CEO reputation and poor earnings quality. One test, for example, showed that even when reputed CEOs had more power in a firm, and therefore more opportunity for rent extraction, earnings quality did not deteriorate further.

"The results indicate the reason more reputed CEOs are associated with poorer earnings quality firms is not because the CEOs take discretionary actions to reduce earnings quality; rather, it is because poor earnings quality firms require the talents of more reputed CEOs. The factors that give rise to poor earnings quality, such as volatile operating environments, are the same ones that require the superior skills of more reputed CEOs," the authors said.

"While our tests point to firm-specific factors as providing most of the explanatory power for earnings quality, we also find the human-capital component of the top executive officer is also important. In particular, the CEO's reputation is a significant factor explaining firm's earnings quality."