基于高频数据的股票市场价格集聚效应特征研究

High-frequency Data Based Empirical Study on Price Clustering in China Stock Market

股票价格常常呈现出集聚效应,并对市场质量产生重要影响,因此需要利用高频数据深入研究中国股票市场价格集聚效应的存在性及其表现特征,同时解释其原因。通过对沪市高频交易数据的统计检验和定量实证研究表明:价格尾数在"0"、"5"和"8"上存在显著的集聚效应,并且可以用"价格决定"假说加以解释,但是不能用"价格吸引"假说解释。投资者通过在这些点位上变动一个最小报价单位进行报价,将可以最小成本获得交易的优先权。

Stock prices often display clustering, which will have an important influence on market quality. So we should do a deep research on the existence, characteristics and reasons of price clustering on our stock market by high frequency data analysis. Through statistical test and empirical study on the high frequency data of Shanghai Stock Exchange market, we conclude that prices show significant clustering on ＇0＇, ＇5＇ and ＇8＇ This phenomenon can be explained by ＂price resolution＂ hypothesis, but can not be explained by ＂price attraction＂ hypothesis. Investors can get the transaction priority with the smallest cost by only change a tick to bid or ask.