Information Risk and Long-Run Performance of Initial Public Offerings

Abstract

While the asset pricing literature has provided empirical evidence for broad samples of firms that information risk is priced by investors, the pricing of information risk in a newly public firm has not been explored. I argue that it is hard for investors to assess the information risk of newly public firms due to the lack of historical information. Using the industry average as expected information risk parameter for an IPO firm, I show that the cross-section of post-IPO long-run abnormal returns, positive as well as negative, can be explained by the difference between realized and expected information risk. In particular, initial underestimation (overestimation) of information risk results in a negative (positive) abnormal return over the period investors update their beliefs. A second set of tests confirms that positive abnormal returns on better-information firms are less persistent than the negative abnormal returns on worse-information firms. The findings are consistent with rational Bayesian investors being able to gradually assess the information risk of newly listed firms; the better the information quality, the quicker is the adjustment.