Financial statement information and abnormal stock returns : a test of increased market efficiency over time in the Swedish stock market

Abstract: This study revisits the question of whether publicly available financial statement information can be used to generate abnormal returns. The study tests the hypothesis that the Swedish stock market has become increasingly efficient over time with respect to publicly available financial statement information, suggested by Skogsvik and Skogsvik (2010), by applying their investment strategy, combining the estimated probability of an increase in mid-term ROE with the implied market expectations for future mid-term ROE estimated from a RIV-model. The sample consists of listed Swedish manufacturing companies between 2008-2019. No evidence of the investment strategy yielding any abnormal return during the sample period is found, supporting the notion of increasing market efficiency over time. These results are robust and consistent even after (i) introducing the risk of bankruptcy to the RIV-valuation framework, (ii) considering an alternative statistical method for estimating the ROE prediction model, and (iii) evaluating the abnormal returns using the Fama-French-Carhart model. By applying the Fama-French-Carhart model, this study also addresses the concerns raised by previous scholars regarding that the predicted probability of an increase in some key value driver is a proxy for systematic risk factors, where we find support to this notion, concluding that the estimated probability of an increase in ROE has significant covariance with the size factor.