The day-of-the-week effect is weak: Evidence from the European real estate sector

The day-of-the-week effect for the securitized real estate indices is investigated by employing daily data at the global, European and country level for the period 1990 to 2010. We test for daily seasonality in 12 countries using both full sample and rolling-regression techniques. While the evidence for the former is in line with the literature, the results for the latter cast severe doubts concerning the existence of any persistent day-of-the-week effects. Once we allow our sample to vary over time, the average proportion of significant coefficients per day ranges between 15 % and 24 %. We show that higher average Friday returns evident in previous literature, remain significant in 21 % of the rolling samples. We conclude that daily seasonality in the European Real Estate sector is subject to the data mining and sample selection bias criticism.