Are Stock Returns Different over Weekends? A Jump Diffusion Analysis of the "Weekend Effect"

Abstract

The distribution of returns on common stocks is, arguably, one of the most widely studied financial market characteristics. The performance of stock prices during breaks in trading has received considerable attention in recent years, especially since the advent of \\"circuit breakers\\" designed to create stability when markets are chaotic. This study examines the distribution of daily returns on five popular stock price indices, with a special emphasis on the difference between returns over weekends and returns over adjacent intraweek trading days. The author revisits the \\"weekend effect\\" in common stock returns, focusing on two characteristics of differential returns over intraweek trading days and over weekends: the \\"drift\\" and the \\"volatility.\\" He finds that the volatility of stock returns over weekends is much smaller than could be predicted from intraweek volatility. This is true of stock returns over weekends both before and after October 1987. He also finds that the difference between intraweek drift and weekend drift is smaller after October 1987 than before. Indeed, it disappears for large companies, suggesting that the poor performance of common stocks over weekends in the 1980s was a financial anomaly that was mitigated over time, as investors incorporated it into the timing of their transactions. The sharp decrease in volatility over weekends is consistent with the view that active trading actually increases volatility, so that a close in trading will be consistent with a reduction in volatility. However, a weekend is a scheduled event, which might simply reduce the rate of new information flow, while a sudden halt in trading might eliminate all information flow from price discovery, creating an environment that elicits the volatility it is designed to mitigate.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 31 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.