Gold and the "weekend effect"

• Published in: The journal of futures markets Vol. 2; no. 2; pp. 175 - 182
• Main Authors: Ball, Clifford A., Torous, Walter N., Tschoegl, Adrian E.
• Format: Journal Article
• Language: English
• Published: New York Wiley Subscription Services, Inc., A Wiley Company 1982; Published by J. Wiley in affiliation with the Center for the Study of Futures Markets, Columbia University; Wiley Periodicals Inc
• Subjects: Clay; Foreign exchange rates; Gold; Gold markets; Hypotheses; Prices; Regression analysis; Return on investment; Statistical analysis; Statistics; Studies; Variables
• ISSN: 0270-7314; 1096-9934
• DOI: 10.1002/fut.3990020209

Recent studies on US common stocks have found that average daily return was significantly negative for Monday while being positive for the other 4 days of the week. This is inconsistent with both the trading time hypothesis, which implies an identical return distribution across all trading days of the week, and with the calendar time hypothesis, which implies a significantly higher mean and variance of return following weekends. In examining the return to gold, data on morning and afternoon fixing prices in London between January 2, 1975, and June 30, 1979, were used. The results indicated that overnight changes were less variable than changes occurring during the day. Daily variances were not equal, and variances for weekends were not significantly different than those occurring within the trading day. Mean returns to gold varied greatly according to the day of the week and whether changes occurred overnight or within the day. This implies that such factors as settlement effects should be considered. However, international markets tended to behave differently from US markets with respect to weekends.