An analysis of six stock market calendar and weather anomalies from 1980 to 2003 shows that (1) returns on trading days in which macroeconomic announcements were made generate the anomalies and (2) five of the six anomalies are not present at all on the trading days in which such announcements...