论文部分内容阅读
风格投资已逐渐成为基金构建投资组合的一种主流量化投资方法.本研究在分析风格资产收益呈分形特征的基础上,通过引入滑动窗口技术对传统的多重分形消除趋势波动分析法(MFDFA)加以改进,并对中信标普公司推出的6种股票纯风格资产指数日收益率序列的波动特征进行研究,实证结果发现:滑动窗口技术能有效减少因分割连接点处的不连续性而产生的伪波动误差;风格资产指数日收益率序列均具有相关多重分形特征,即原始序列具有持久性,位置重构序列均具有反持久性,且位置重构序列的多重分形特征显著弱于原始序列的多重分形特征,表明风格资产指数收益序列的持久相关性是形成多重分形特征的主要原因;价值、成长型比规模型风格资产具有更规律的多重分形特征,表明价值、成长型比规模型风格资产的分形规律更明显.本研究对基金公司、基金经理及时准确地把握股市风格动向以便构建适度风格漂移策略具有重要的理论价值与现实意义.
Style investment has gradually become a mainstream quantitative investment method of fund construction investment portfolio.Based on the analysis of the fractal characteristics of the return on style assets, the paper introduces sliding window technology to the traditional multifractal trend elimination analysis (MFDFA) The paper also studies the fluctuation characteristics of the daily yield series of six kinds of purely asset-backed index stocks released by CITIC-S & P. The empirical results show that the sliding window technology can effectively reduce the false generation caused by the discontinuity at the splitting connection point Volatility error and daily return of style asset index all have the related multi-fractal characteristics, namely, the original sequence is persistent and the position reconstruction sequence has anti-persistence, and the multifractal characteristics of the position reconstruction sequence are significantly weaker than the original multiple sequences The fractal characteristics indicate that the long-term correlation of the returns of style asset indices is the main reason for the multifractal character. Value and growth have more regular multifractal characteristics than those of scale-style assets, indicating that value, growth, The fractal rules are more obvious.This research is to fund companies and fund managers in time Correctly grasp the trend of the stock market-style policies in order to build a modest style drift has important theoretical and practical significance.