Do Security Analysts Speak in Two Tongues?
University of California, Berkeley - Department of Economics; University of California, Berkeley - Haas School of Business; National Bureau of Economic Research (NBER); Centre for Economic Policy Research (CEPR); Institute for the Study of Labor (IZA)
Devin M. Shanthikumar
University of California, Irvine - Paul Merage School of Business
AAA 2006 Financial Accounting and Reporting Section (FARS) Meeting Paper
Why do analysts display overoptimism about the stocks they cover? According to the selection hypothesis, analysts are truly too optimistic about the stocks they choose to cover. According to the conflict-of-interest hypothesis, analysts choose to distort their view to maximize profits via commissions and underwriting business, in particular if affiliated with an underwriting bank. We analyze the concurrent issuance of recommendations and earnings forecasts to differentiate between these two hypotheses. The selection hypothesis implies a positive correlation between overoptimism in recommendations and in forecasts. Under the conflict-of-interest hypothesis, analysts may choose to distort recommendations but to prove their analyst quality in their forecasts, which are directed towards more sophisticated, institutional investors. We find that, while affiliated analysts' recommendations are more optimistic than unaffiliated recommendations, affiliated analysts' earnings forecasts are more pessimistic than unaffiliated forecasts. Similar discrepancies between the timing and persistence of recommendations and forecasts confirm this interpretation. Additional results on trading reactions indicate that small traders react indeed more strongly to recommendations, while large traders discount recommendations and react more strongly to analyst earnings forecasts.
Number of Pages in PDF File: 43
Date posted: October 3, 2005