錯誤認定交易買賣方引致知訊交易機率估計偏誤之研究

Abstract

In spite of the importance to estimate the probability of informed trading (PIN), a widely-used market measure (Easley et al., 1996, 2002), empiricists usually have no access to the data for distinguishing between buyer- and seller-initiated transactions and thus rely on trade classification algorithms. Nevertheless, prevalent algorithms such as Lee-Ready (Lee and Ready, 1991) may cause misclassifications. Consequently, PIN may be underestimated in extant microstructure studies. Aiming at correcting such bias, this paper proposes a remedial approach useful for both original and extended PIN models. We also demonstrate its effectiveness with simulated and historical data.