"In this paper, we suggest another possibility that causes PEAD rather than market friction; that is, investment horizon of shareholders. Different investors have different investment horizons. For example, some investors like retirees require stable cash flows from their investment for a long time, while other investors like young generation seek relatively high profits in a short-term. We call the former type long-term investors and the latter type short-term investors.  We suppose that long-term investors are inattentive to short-term events such as earnings announcements and slowly respond to the events. On the other hand, short-term investors are likely to pay much attention to short-term events and actively trade stocks to exploit potential profits from the events. Therefore, our hypothesis is that PEAD occurs only among stocks mostly traded by long-term investors."

"Based on the above studies, we adopt the Sturn measure by Kwon and Kim (2014), as a proxy for investment horizon of each stock’s shareholders. Specifically, we use U.S. mutual fund data and calculate share-weighted portfolio turnover levels of mutual fund shareholders for each stock. High Sturn measure implies that shareholders of the stock have short-term investment horizon. We also estimate a standardized unexpected earnings (SUE) measure using analyst forecast data. Stocks are divided into three groups based on Sturn, and PEAD is estimated for each Sturn group. As a result, PEAD is strong (1.55% with t-value 7.08) for the low-Sturn group, while it is weak (0.48 with t-value 1.44) for the high-Sturn group. The difference is -1.09% with t-value 3.39, which is economically significant as well as statistically significant since its annualized value is about 14%. We conclude that PEAD is strong for stocks mostly held by long-term investors."

"Interestingly, the results contradict the prediction of the previous studies that PEAD should concentrated on stocks with high transaction costs and low investor recognition. Stocks in the low-Sturn group have smaller transaction costs and higher investor recognition than stocks in the high-Sturn group; however, PEAD is stronger for stocks in the low-Sturn group. It strongly supports the idea that PEAD is caused by underreaction of long-term investors. Cross-sectional regression results reconfirm that investment horizon of shareholders remains a strong determinant of PEAD even after transaction costs and investor recognition are controlled for."