BrianWhiteby Brian White

Earnings are central to financial reporting. They are a major focus for managers, analysts and investors, and earnings announcements generate headlines and often move stock prices. But how should earnings be measured and presented to financial statement users? For example, how should we deal with transitory items like many of the items that are classified as other comprehensive income (OCI)? Transitory items are those, like foreign currency translations, that are unpredictable from one year to the next. They are contrasted with persistent earnings elements, which are more predictable and easier to forecast.

There are multiple possibilities for dealing with transitory elements in earnings. They could be combined with persistent elements into a single earnings metric (think comprehensive income), they could be disaggregated into a separate metric (think OCI), or they could be excluded from earnings altogether (think net income, at least compared to comprehensive income, because net income usually includes some transitory items too). This issue is particularly timely, as the Financial Accounting Standards Board (FASB) now requires firms to display comprehensive income more prominently in financial statements.

In a recent paper that I coauthored with colleagues from the University of Illinois, we tested the effects of several different earnings metrics, each of which treated transitory elements in a different way, on investor behavior and market prices in laboratory markets. Our results provide several new insights. When we included transitory elements in a prominently disclosed earnings metric, investors searched unnecessarily for further information about transitory items, and they overestimated the effect of transitory items on firm value. In short, including transitory elements in earnings metrics cause investors to overestimate their importance. This was true whether transitory items were aggregated with persistent earnings or disaggregated in a separate metric. These effects could also be seen in prices. In fact, prices were most efficient—in other words, prices deviated least from fundamental value—when we excluded transitory elements from earnings metrics altogether.

We hope that our study will provide timely evidence on a matter of interest to standard setters and preparers. While increasing the prominence of transitory items in earnings is perhaps understandable on the basis of transparency, our study suggests that any transparency benefits should be weighed against the negative effects that we document for investor behavior and price efficiency. Although these negative effects can be mitigated to some extent by disaggregating transitory and persistent earnings elements, we find that excluding transitory elements from earnings metrics yields the greatest benefits.

 

W. Elliott, J. Hobson & B. White. 2015. Earnings metrics, information processing, and market price efficiency. Forthcoming in Journal of Accounting Research.