Which Analysts Make the Best Stock Recommendations?

Abstract: This paper uses a set of 4566 stock recommendations for companies traded on the Stockholm stock exchange during the five year period 2010-2014 to investigate what factors affect analysts' stock-picking ability. We begin our analysis by using a buy-and-hold strategy to establish the presence of stock recommendation announcement effects and long term over performance. We find average three-day event window abnormal returns of 1.58%, as well as average six-month holding period abnormal returns of 2.54%. Subsequently, we attempt to deepen our understanding of analysts and their working conditions, by posing four hypotheses regarding analyst-level factors that could affect the abnormal returns of their recommendations. Number of individual stocks covered has a strong statistically significant impact, with incremental abnormal returns of 0.21% per stock covered, for a six-moth holding period. Nationality of employer has no impact on long-term abnormal returns, but there is a significant difference in announcement effects, with 0.8% higher average three-day event-window returns for local firms. However, both of these factors have ambiguous economic interpretations. Recommendations made by analysts who work in teams are associated with 2.55% higher abnormal returns for a six-month holding period, with high statistical significance. Years of experience has a slightly positive relation to stock-picking ability, but statistical significance is low. These two factors on the other hand, both align with our hypotheses. As a last step, a transition matrix of analyst rankings is created and shows that analyst performance persistence is very low, suggesting that there might not be strong determinants of performance on the analyst level.