This article investigates the modelling of style returns in the United States and the returns to style 'tilts' based on forecasts of enhanced future style returns. We use hidden Markov model to build our forecasts for data from 1975 to 1998. We do not include more recent observations as the subsequent trend and volatility sways the analysis. Our finding that style returns are less forecastible in the late 1990s is consistent with the hypothesis that style returns are the result of anomalies rather than risk premia. The erosion of anomalous returns as public awareness of their presence is translated into strategies that arbitrage away the excess returns seems to be a hypothesis consistent with our modelling results.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados