Dynamic Alpha Quantitative Investment Strategy Based On Spectral Analysis

The performance evaluation standard of traditional quantitative investment strategy is measured by alpha,beta,standard deviation,tracking error,sharp ratio and information ratio.However,these measures do not describe the dynamic components of investors' activities,nor do they consider the time range of weight changes.In order to solve the measurement problem of dynamic returns generated in the process of active investment,this paper adopts a new dynamic Alpha method.It can capture the part of income generated by active investment in investment decisions,namely dynamic Alpha.This dynamic Alpha technique uses spectral analysis to decompose the return rate and its weight of individual stocks in the portfolio into frequencies,and the result consists of dynamic and static parts.When the return frequency and the weight frequency are in phase,a positive dynamic Alpha will be generated,thus bringing positive returns to the investment portfolio.On the contrary,it will reduce the income of the investment portfolio.Firstly,this paper introduces active/passive decomposition,which is the theoretical basis of dynamic Alpha.the expected return rate of the portfolio is divided into two parts to evaluate the investment performance,one part is the covariance of the portfolio weight and return,the other part is the product of the expected weight and return rate,and there are some numerical analysis in this section as well.Then,this paper introduces the spectral analysis principle used in dynamic Alpha,including discrete Fourier transform(DFT),power spectrum and general moment characteristics of power spectrum.Then,this paper introduces the dynamic Alpha theory.Through spectral analysis,the individual stock return rate and its weight of the portfolio are divided into dynamic and static parts to measure the performance of the portfolio and make numerical analysis.Finally,in the fourth chapter,five stocks listed in China before 2015 are selected to form a 5-day mean regression dynamic Alpha strategy,and are compared with the Shanghai and Shenzhen 300 index returns.The results show that 106.2% dynamic Alpha can be generated by actively adjusting the weight of the portfolio,thus increasing the return of the portfolio.