Hold 'em? Using Financial Statement Information to Pick Winners and Losers when Consensus Analysts' Recommendations are Neutral

Abstract

Sell-side financial analysts commonly issue recommendations to buy, sell, or hold particular firms' shares. Hold recommendations do not provide investors with directional recommendations to change their investment positions. We develop and test predictions about whether users of analysts' recommendations can exploit financial statement information to identify hold-recommendation firms that will generate abnormal returns in the coming year. Our approach summarizes financial statement ratios into a predictor of the likelihood a firm will generate an earnings increase in the coming year. We find that hold-recommendation firms with the greatest likelihood of future earnings increases generate average abnormal returns that significantly exceed those of hold-recommendation firms with the lowest likelihood of future earnings increases. We find that a trading strategy that utilizes our approach to pick stocks from among the set of hold recommendations generates average annual abnormal returns of 16.4 percent over our 12-year sample period, after controlling for previously identified predictors of expected returns. Our results provide insights for accounting and finance academics, as well as for institutional investors and other users of sell-side analyst recommendations.

Keywords: Analysts' Stock Recommendations, Financial Statements Analysis, Analysts

JEL Classification: M41