多重时间标度鞅半方差与加权鞅半方差β系数研究

Research on Multi-time-scale Martingale-based Semi-Variance and Weighted Martingale-based Semi-Variance Betas

本文根据投资者持有资产周期的不同,提出了基于小波多分辨率分析的多重时间标度CAPM模型,得到多重时间标度β系数。由于投资者在进行投资时更关注下行风险,本文进一步提出了基于鞅半方差与加权鞅半方差的下行β系数计算方法,并通过实证检验证明了用鞅半方差β系数与加权鞅半方差β系数来衡量系统风险较其他方法更具合理性及优越性。最后,本文将小波多分辨率分析与鞅半方差β系数及加权鞅半方差β系数结合,得到多重时间标度下的鞅半方差β系数及加权鞅半方差β系数计算方法。

Based on different investment horizons of investors, we introduce multi-time-scale CAPM and multi-time-scale beta coefficients are estimated. Since investors concern more about downside risk, we propose new methods to calculate downside betas based on martingale semi-variance and weighted mar- tingale semi-variance and the reasonableness and advantages of martingale semi-variance and weighted martingale semi-variance betas were examined. By combining multi-resolution analysis of wavelet and martingale semi-variance and weighted martingale semi-variance betas, we obtain multi-time-scale mar- tingale semi-variance and weighted martingale semi-variance betas.