The assumption that daily stock returns are normally distributed has long been disputed by the data. In this article the normality assumption is tested (and clearly rejected) using time series of daily stock returns for 13 European securities markets. More importantly, four alternative specifications are fitted to the data, overall support is found for the scaled- t distribution (and partial support for a mixture of two Normal distributions), and the magnitude of the error that stems from predicting returns by using the Normal distribution is quantified. Data also show that normality may be a plausible assumption for monthly (but not for daily) stock returns.
展开▼
