中国股市收益率与成交量动态关系的研究——基于工具变量的分位数回归(IVQR)模型

Analysis of Yield and Volume in China's Stock Market——Based on IVQR Models

基于分位数回归(QR)模型分析了不同分位水平下的收益率与成交量的关系,考虑到收益水平对成交量的影响,引入工具变量,构建IVQR模型,更加客观地分析了不同分位水平下成交量对收益率的作用。蒙特卡洛模拟结果表明IVQR估计比QR估计具有更小的偏差和更强的稳健性。实证分析结果表明条件收益率处于较高水平时与成交量正相关,且分位水平越高两者之间的相关性越大;条件收益率处于较低水平时与成交量负相关,且分位水平越低相关性越大;多数分位水平下成交量对收益率的影响并没有较大的差异。同时结果还表明工具变量的分位数回归模型能较好地处理模型中的内生性问题。

Based on quantile regression models,the correlation between volumeand yield on different levels is studied.In order to eliminate the influence of yield's level to the volume,an instrumental variable and instructa IVQR model are introduced,based on which the effect of volume on yield of different levels can be analyzed objectively.Monte Carlo simulation shows that IVQR estimator has less bias and stronger robustness.In empirical analysis,it is found that yield and volume are positively correlated if the yield is on a high level,the higher the level is,the larger the correlation will be;and that yield and volume arenegativelycorrelated if the yield is on a low level,the lower the level is,the larger the correlation will be.We also find that the volume's effects on yields of most levels are similar.The outcome simultaneously shows that IVQR can dispose the endogeneity in models.