The evidence from the pooled data is simple and intuitive but is open to criticism because contemporaneous international stock returns are likely to be positively correlated. It is well known that cross-sectional correlation in returns can lead to understated estimates of standard error and inflated t-statistics (., Bernard 1987). However, this concern is unlikely to be overly important in our setting for two reasons. First, Bernard (1987) shows that problems due to cross-sectional dependence in returns are less pronounced for shorter time-series, and are fairly mild for the case of daily returns. Second, Hirshleifer and.
