Equity Research Analyst Bias in Stock Recommendations

Abstract: We acknowledge the lack of empirical studies of equity research analysts' ability to generate abnormal returns on the Swedish stock market and the global absence of studies measuring performance throughout complete recommendation periods. A unique sample of 705 equity research analyst recommendations from 2004 to 2013 from leading Swedish investment banks ABG Sundal Collier, Carnegie, Nordea and SEB on the OMXS30 constituent stocks are studied in order to investigate whether analysts have better long-run predictive abilities in their sell relative to their buy recommendations. Analyst issuance of sell recommendations on stocks can incur risks of deteriorating analyst - company relationship and reduced management access. Hence analysts should be significantly more certain in their sell recommendations. Contrary to our hypothesis we find that analyst buy recommendations have positive abnormal returns and sell recommendations generate negative abnormal returns with significant 30-day, 180-day and Recommendation Period Return returns of 1.5%, 3.6%, and 20.9% respectively on buy recommendations adjusted with CAPM and Fama-French factors. Hence, an investor could enjoy positive abnormal returns following buy recommendations from Swedish investment banks given access to such research is available.