应用门限分位点回归模型估计VPIN条件下CVaR

Evaluating CVaR conditioned on VPIN based on the threshold quantile regression model

知情交易概率(probability of informed trading,PIN)是市场微观结构中度量知情交易的一种重要度量方法.给出了一种基于交易量的PIN估计方法VPIN(volume-synchronized probability of informed trading).并应用门限分位点回归模型分析了VPIN与收益率之间的非线性结构关系,给出了VPIN条件下市场风险(conditional value at risk,CVaR)的度量方法.最后对上证综指数据进行了实证分析,实证结果表明VPIN和日收益率之间存在着较为显著的关系,VPIN越大,相应的市场风险CVaR越小.

Probability of informed trading （PIN） is an important measure of informed trading in market microstructures. A new measure of PIN based on trading volume was introduced which is called volume- synchronized probability of informed trading （VPIN）. The threshold quantile regression model was used to analyze the nonlinear relation between VPIN and daily return, and the conditional VaR （CVaR） conditioned on VPIN was given. Finally, an empirical analysis of Shanghai composite index was conducted. The result shows that there exists a close relationship between VPIN and daily return： the lager the VPIN, the smaller the CV^R