The most striking phenomenon among the stock market anomalies is the interaction of autocorrelations and the day-of-the-week effects. International evidences suggest that return autocorrelations are the highest on weekends, i.e., between returns of Friday and Monday. In this paper, differences in autocorrelations of S&P CNX Nifty Index-returns across the different weekdays have been investigated. The behavior of autocorrelations was not only found varying according to the day-of-the-week, but the positive and the negative returns were also showing different patterns of autocorrelations. In conformity to the international trend, the highest positive first-order autocorrelation was observed at the weekend for Friday returns, but this weekend effect was found significant for negative returns only. The authors also observed significantly high negative second-order autocorrelation for negative Monday returns. In both the cases, the autocorrelations were found about seven times higher than the average unconditional first-order and second-order autocorrelations, respectively.

When stock returns show certain empirical regulations, which are difficult to explain from asset-pricing theories, they are called stock market anomalies. Among different stock market anomalies, two are very widely documented—the day-of-the-week effect and the positive first-order autocorrelation. In an efficient market the stock prices must be independent of the day, week, month and other calender dummies. But, studies have found, inter alia, that average stock returns significantly vary across the weekdays. Abnormally high positive average return is observed for last trading day of the week (which is usually a Friday); and the first day of the week (i.e., Monday) shows a significantly low or a negative average return. Similarly, the Efficient Market Hypothesis (EMH) postulates that the stock prices show a random walk behavior and past prices cannot be used to predict future prices. In contrary to this postulation some studies have found that daily stock return show a significant positive first-order autocorrelation and thus, tomorrows expected returns are not independent of today’s realized returns. However, the more striking observation is that the autocorrelation effect and the day-ofthe- week effect interact with each other. The highest and statistically significant positive first-order autocorrelation is observed between Friday returns and Monday returns i.e., positive Friday returns tend to be followed by positive Monday returns and negative Friday returns tend to be followed by negative Monday returns.

Applied Finance Journal, Stock Market Index, Stock Market Anomalies, Efficient Market Hypothesis, Asset-Pricing Theories, Indian Stock Market, Emerging Markets, Financial Markets, Derivative Products, Regression Models, Autoregression Model, Indian Equity Market.