超额收益的可预测性与资产配置——基于中国股票市场数据的研究

发布时间：2018-05-19 01:06

本文选题：超额收益 + 可预测性　；参考：《复旦学报(社会科学版)》2013年06期

【摘要】：本文首先通过在三个限制变量(M1月环比增长率、CPI月环比增长率和银行间市场回购利率)之中选取变量来构建预测模型,检验中国股票市场的可预测性在统计上是否显著;之后将交易成本和各种交易限制纳入考虑,通过最大化投资者效用,基于不同的预测模型进行实时环境下的资产组合配置,并考察统计显著性和投资绩效之间的关系。研究表明,中国股市超额收益与M1月环比增长率之间具有稳定而显著的线性关系。在短期内,统计显著性与优化的投资绩效之间并非完全相关,而长期内,依据M1月环比增长率构建预测模型的投资策略可能带来良好的绩效。
[Abstract]:In this paper, we first choose the variables among the three limiting variables: M1 monthly ratio, CPI monthly ratio and interbank repo rate, to test whether the predictability of Chinese stock market is statistically significant. Then the transaction cost and various transaction restrictions are taken into account. By maximizing investor utility portfolio allocation in real time environment is carried out based on different forecasting models and the relationship between statistical significance and investment performance is investigated. The study shows that there is a stable and significant linear relationship between the excess return and the growth rate of M1 month ratio. In the short term, the statistical significance is not completely related to the optimal investment performance, but in the long run, the investment strategy based on M1 monthly ratio growth rate may bring good performance.
【作者单位】：复旦大学国际金融系;
【基金】：国家自然科学基金项目“公开信息冲击下的投资者交易策略高阶期望及其实证分析”(项目批准号:70671027)的资助
【分类号】：F832.51;F224

【参考文献】