Key points are not available for this paper at this time.
We decompose PIN, the probability of informed trading, into good-news (PING) and bad-news (PINB) components, which we estimate at a quarterly frequency. We first assess the validity of PIN as a measure of informed trading by calculating its association with measures of the adverse-selection component of the cost of trading. We then provide new evidence that PING and PINB capture informed trading around earnings announcements by showing that they predict positive and negative earnings surprises, respectively. Conjecturing that investors who take long positions will be more concerned about informed selling than about informed buying since the former depresses the sale price whereas the latter raises it, we then investigate asymmetry in the pricing of private information. We find strong evidence of such asymmetry in that the effect of PINB on the cost of equity capital is large and highly significant, whereas the effect of PING is small and statistically insignificant. Data, as supplemental material, are available at http: //dx. doi. org/10. 1287/mnsc. 2015. 2250. This paper was accepted by Neng Wang, finance.
Brennan et al. (Thu,) studied this question.