Does Analyst Behavior Explain Market Mispricing of Foreign Earnings for U.S. Multinational Firms?

Abstract

The general theme of examining components of earnings is of interest to both practitioners and academics, given its potential for allowing investors and analysts to more precisely forecast earnings and estimate firm value. For U.S. multinational firms, prior research indicates that while the foreign and domestic earnings components are value relevant, the market fails to fully incorporate the higher persistence of foreign earnings. However, findings of potential market mispricing can be attributed to alternative explanations such as unidentified risk factors or flaws in research design.

In this study, we seek to circumvent these concerns by examining whether financial analysts (who are deemed to be sophisticated users of financial statements and act as intermediaries between companies and investors) fully incorporate the time-series properties of foreign and domestic earnings in formulating their earnings forecasts. Similar to market-based studies, we find that analysts do distinguish between foreign and domestic earnings. However, we find that analysts also fail to fully incorporate the higher persistence of foreign earnings. Finally, we find that analysts' underestimation of foreign earnings persistence appears to parallel the similar finding at the market level.

We discuss how our findings relate to the prior literature, and what they imply for accounting research and practice. Specifically, our study complements previous research which appears to undermine the notion of an efficient market. Also, we find that analysts' bias is in the direction of lowering earnings expectations by underestimating the persistence of foreign earnings, which suggests that collusion with management to inflate earnings expectations is not a plausible explanation for the bias. Collectively, our findings suggest analysts' lack of effectiveness as information intermediaries.