An improved estimation method and empirical properties of the probability of informed trading

► We develop an algorithm to overcome the estimation bias caused by boundary solutions. ► We estimate PIN for nearly 80,000 stock-quarters between 1993 and 2004. ► We study the methods in Easley et al. (2010) and Lin and Ke (LK, 2011). ► Boundary solutions appear with a very high frequency when the LK method is used. ► We document several interesting empirical properties of PIN.