中国股市β系数的多尺度特性分析

摘要

This paper presents β coefficient estimation method based on wavelet variance and wavelet covariance, and estimates the risk coefficient on different scales through multi-scale decomposition of wavelet variance and wavelet covariance, then uses this method to carry on multi-scale analysis for China securities A-share market branch industry and the investment portfolios coefficient. The empirical results show that China's stock markethas the characteristics of complex multi-scale fluctuations, stock market riskson different time scales exhibits different risk. The risk of short-term investments mainlyin the high-frequency fluctuations, investors should consider a low scalesβcoefficient, while long-term investment risk mainly forlow-frequency fluctuations, investors should consider a large scalesβcoefficient.%本文提出一种基于小波方差和小波协方差的β系数估计方法，并通过小波方差和小波协方差的多尺度分解估计出不同尺度上的风险系数，用该方法对中国证券A股市场分行业及投资组合的β系数进行了多尺度估计分析。实证结果表明，我国股市具有复杂的多尺度波动的特征，不同时间尺度上证券市场所表现出的风险不一样，短期投资的风险主要表现在高频波动，投资者应当考虑低尺度下的β系数，而长期投资风险主要表现为低频波动，应当考虑大尺度下的β系数。