We show there exists a monthly cycle of information production that has a pronounced effect on stock prices. Specifically, for stocks with high returns (likely with good information) in a return horizon ranging from one month to three years, their returns spike during early month relative to the rest of the month for each month within the return horizon. In addition, stocks with high returns exhibit more price jumps that are positive, higher market beta, and higher volatility in early than the rest of the month. These are consistent with the hypothesis that (i) some information is produced in a lumpy way around the end of a calendar month; and (ii) such information, if good, is speedily revealed to the market early next month. For stocks with low returns (likely with bad information), we find evidence of slow adjustment in price to bad information---such stocks exhibit less pronounced intra-month variation in average return; and they have more price jumps that are negative, higher market beta, and higher volatility in late month relative to earlier in the month. Consistent with theories on strategic disclosure, the difference between individual-stock return skewness and market return skewness is bigger in late month than earlier in the month. The asymmetry between the dissemination of good and bad information implies that the aggregate stock market is also affected by the monthly information cycle, which we confirm.
展开▼
