A Quantitative Strategy Study Of Multi-factor Stock Selection And Hidden Markov Model Timing Based On ESG Investment Philosophy

In the 21 st century,global warming and internal development problems of human society have become more and more acute,especially under the influence of the COVID-19,the issue of sustainable development has increasingly attracted the attention of countries around the world,and ESG emerged under this background.As a social responsibility evaluation standard that focuses on corporate environment,society and governance,ESG has gradually become a hot topic in the fields of domestic and foreign investment due to its wide applicability and quantifiable characteristics,and has promoted ESG investment philosophy in the international capital market.Up to now,the industry and academia have not yet reached a unified opinion on whether ESG investment can achieve excess returns and the source of potential returns,but the mainstream view is that there is a positive correlation between ESG investment and excess returns,and ESG investment in emerging markets has better excess returns than in developed markets.The research on ESG in China started relatively late and the related research is relatively less.Therefore,this thesis constructs a quantitative stock selection strategy based on ESG rating data to demonstrate whether ESG factors can bring excess returns in the domestic market,and uses quantitative timing strategy based on the hidden markov model(HMM)to optimize the ESG stock selection strategy to tap the potential excess of ESG factors,in order to enrich the strategy research of ESG investment in the field of quantification.Based on the Quant Data ESG rating data from 2017 to 2021,this thesis selects all the dynamic constituent stocks of the CSI 800 Index within the research interval as the stock pool.First of all,according to the traditional multi-factor stock selection process such as data preprocessing,single-factor testing,model construction and backtesting,the ESG stock selection strategy and the Barra stock selection strategy are sequentially constructed.Then,fusing the ESG factors with the Barra factors to generate ESG-Barra stock selection strategy,and then analyze whether there are extra returns in the ESG-Barra stock selection strategy that cannot be explained by the Barra factor,and then judge the effectiveness of the ESG factors.According to this,the daily quotation data of CSI 800 index in the study period is selected as the training sample,and the prediction results of future market movements in different states are generated based on the HMM model with regular rolling training.Finally,the ESG stock selection strategy is optimized for timing based on the forecast results to enhance the backtesting returns of the ESG quantitative strategy.The main findings of this thesis are as follows: First,by comparing the ESG-Barra stock selection strategy with the Barra stock selection strategy,it is found that the backtesting returns of the Barra stock selection strategy is still improved after adding the ESG factors,indicating that there are additional returns of the ESG factors beyond the Barra style factors,i.e.,the ESG factors is effective in the domestic market.Second,according to the backtesting result of ESG stock selection strategy,the return of ESG stock selection strategy in normal period is not much different from the benchmark return,while the return during the COVID-19 far exceeds the benchmark return,indicating that ESG factors have significant risk resistance in the crisis period.Third,the performance of the timing strategy based on the hidden markov model is higher than that of the CSI 800 benchmark index,indicating that the timing strategy based on the hidden markov model has a strong predictive ability for the future stock market state.Fourth,the timing strategy based on HMM model can effectively improve the performance of ESG stock selection strategy,and the return and risk performance of the combined ESG-HMM strategy is significantly better than that of ESG-Barra stock selection strategy,indicating that the combined strategy of "stock selection + timing" is more suitable as a trading strategy than a single stock selection strategy.