基于卡尔曼滤波的中国证券市场投资者情绪测度研究

Research about the Measure of Chinese Securities Market Investor Sentiment Based on Kalman Filter

投资者情绪准确测度较为困难,因为其固有的动态复杂性和变化性。本研究通过对情绪测度理论方法进行系统分析,结合中国证券市场具体情境,基于卡尔曼滤波方法过滤掉市场噪声,选取封闭式基金折价率、新股上市首日收益率、新股发行数量、新增开户数四个指标,并通过滚动合成投资者情绪指数进而验证其有效性。

Accurate measure of investor sentiment is particularly difficult because of its inherent complexity and dynamic variability.This paper has carried on the systematic analysis of sentiment measure theory combined with specific situation of Chinese stock market. This paper filters out noise based on the kalman filter method and built four indicators（Closed-end fund discount rate; IPO first-day returns; Number of IPOs; New accounts） into investor sentiment index. Finally,this thesis verifies the validity of sentiment index.