论文部分内容阅读
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。
In this paper, we propose a multi-time scale CAPM model based on wavelet multi-resolution analysis, and obtain multiple time scale β coefficients according to the different asset-holding periods of investors. As investors pay more attention to the downside risk while investing, this paper further proposes a method of calculating the β coefficient of downward β based on the martingale semi-variance and the weighted semi-variance of the martingale. Empirical tests prove that using the semi-variance of the martingale and the weighted martingale semi-variance β Coefficient to measure system risk is more reasonable and superior than other methods. Finally, this paper combines wavelet multi-resolution analysis with the martingale semi-variance β coefficient and the weighted martingale semi-variance β coefficient to obtain the martingale semi-variance β coefficient and weighted martingale semi-variance β coefficient calculation method under multiple time scales.