A New Method to Estimate PIN (Probability of Informed Trading)

Abstract

In spite of the many successful applications of the concept of probability of information-based trading (PIN), the method used for its estimation (Easley et al., 1996) suffers several serious shortcomings. 1) It fails to deliver the PIN to a huge number of stocks. Using the most widely cited PIN estimates posted by Havidkjaer as an example, the joint market share of stocks with valid PIN estimates is only 52% of the entire market in 2001. The current trend is even worse because the steady increase of trading activities increases the failure rate. 2) It assumes that trading days or event days are independent, which is refutable. In this paper, a new method is developed to correct these problems. For example, the assumption of independence is discarded. The new method is designed to deliver PINs to all stocks. In contrast to the current method, the performance of the new method is an increasing function of the trading activities. After comparing the proposed new PINs with those in Easley et al. (1996) and Havidkjaer, many interesting findings are derived, which are presented here.

Keywords: PIN, informed trading, systematic bias, information risk

JEL Classification: G12, G14