Mutual Fund Persistence Using Kalman Filter Models

Using a dynamic state-space framework, we examine the persistence of US equity mutual fund performance. Performance is estimated from standard factor models via a Kalman filter following that of Mamaysky et al. (2008), which has not been used in the persistence literature until now. Persistence is tested across monthly, quarterly, biannual and annual horizons from 1995 to 2015. In opposition with current literature, we present strong evidence that funds exhibit persistence in risk-adjusted performance across all measurement horizons, and is not driven by fund fees. However, persistence in risk-adjusted performance is unable to translate into immediately higher net returns for investors. These findings lend support to the notion that some fund managers do possess skill in their ability to select stocks