Industry Classification and the Efficiency of Intra-Industry Information Transfers

Motivated by recent studies that demonstrate the superiority of the Global Industry Classification System (GICS) relative to the Standard Industry Classification (SIC) system in capital market research, we revisit the stock market anomaly documented by Thomas and Zhang (TZ) ("Overreaction to intra-industry information transfers?" Journal of Accounting Research, Vol. 46, pp. 909–940) and analyze whether the anomaly based on SIC remains evident when intra-industry information transfers are based on the GICS. We first replicate TZ and test whether stock prices of late announcers in response to earnings reported by early announcers in the same SIC industry are significantly related to subsequent price responses of late announcers to their own earnings reports. In the multivariate setting, we then evaluate whether the magnitude and significance of the overreaction anomaly changes under the more comprehensive GICS and across different time periods. We first confirm the over-reaction anomaly based on SIC as documented by TZ. Utilizing a larger sample of firms based on the GICS and extending TZ for a new time period, we then demonstrate that the overreaction anomaly disappears during recent years, a period that is characterized by markedly higher trading activity. The findings provide new insights and contributions to the debate on whether or not market significantly misprices information transfers