ANALYST CHARACTERISTICS AND EARNINGS FORECAST ERRORS

In this article, we investigate the association between analysts’ earnings forecast errors and analyst characteristics. These characteristics include working experience, company experience, brokerage house size, and boldness in making recommendations. We use t-tests and time-series regressions to examine the relationship. Our results reveal that analysts’ years of experience is the most important factor. The more experience analysts have, the more accurate their forecasts tend to be, which is consistent with the learning-by-doing theory. For the brokerage house size, analysts from larger brokerage houses are more likely to provide accurate forecasts. For boldness, when an analyst’s recommendation is far away from mean recommendation, the earnings forecast tends to be less accurate. Company-specific experience has no robust relation with forecast errors. We conclude that experience of analysts, brokerage size, and analysts’ recommendation dispersion are correlated with forecast accuracy.