Seasonality in returns and volatility: evidence from the European Real Estate sector.

Abstract

In this study, we investigate the day of the week effect on thirteen European real estate index returns and conditional variances (volatility). The empirical research was conducted using non-linear models from the GARCH family and daily returns from several European securitized real estate indices between 1990 and 2010. We permit two specifications for the error distribution – Student’s t and generalized error distribution. Furthermore we are trying to explain the observed anomalies by reference to market risk in a CAPM-type framework, using as risk proxy two market indices (EPRA/NAREIT Global index and EPRA Europe index). In order to investigate if the significant seasonality remains significant for the whole sample period, we estimate the mean coefficients of GARCH models in a rolling framework.