An Improved Estimation Method and Empirical Properties of PIN
The probability of information-based trading (PIN) is estimated via numerical maximization of a likelihood function. The maximization solutions frequently fall on the boundary of the parameter space and such boundary solutions cause a systematic bias in PIN estimates. We develop an algorithm to obtain PIN estimates that are free of this bias. We analyze the estimates for over 80,000 stock-quarter pairs and document evidence for temporal and seasonal variations in PIN. These variations are related to the January 2001 decimalization and the year-end tax-loss selling. We also find systematic differences between our estimates and others.
decimalization market microstructure numerical maximization PIN information-based trading tax-loss selling
Yuxing Yan Shaojun Zhang
Wharton Research Data Services, 216 Vance Hall, 3733 Spruce Street, Philadelphia, PA, 19104 Division of Banking and Finance, Nanyang Business School, Nanyang Technological University, S3-B1A-0
国际会议
昆明
英文
1-49
2005-07-05（万方平台首次上网日期，不代表论文的发表时间）