Empirical Research On Asset Allocation In China’s Stock Market Based On Emergencies

Since 2020,China has become the country with the largest number of shareholders in the world,and the stock market is growing day by day.In the context of global economic integration,the financial markets of various countries are becoming increasingly interconnected.On the one hand,it promotes the circulation of funds in the financial market and strengthens economic exchanges among countries;On the other hand,it can also easily lead to financial market panic,exacerbating market risks in various countries.More and more scholars are paying attention to the optimization of asset allocation in the stock market,through model innovation and method innovation in order to obtain stable returns and reduce investment losses.On the basis of existing research,this article focuses on the investment portfolio issues before and after the occurrence of unexpected events.By constructing a dual risk model and improving it,an empirical analysis is conducted on asset allocation during different periods of unexpected events,and the superiority of the model is explored.The specific content is as follows:Firstly,empirical analysis was conducted using 50 stocks on the Shanghai Stock Exchange,and a mean variance CVaR dual risk model was constructed to compare and analyze with traditional single risk models.Secondly,considering the return of investment guided by the single risk model and the double risk model under the abnormal fluctuation of the stock market,according to the fluctuation of the stock price and the background of the historical economic era,the research interval is divided into key points.Based on the financial crisis in 2008,the European debt crisis in 2010,the stock disaster in 2015 and the global COVID-19 in 2020 Explain and empirically analyze the market situation in the mid-term and later stages,and use iterative methods to determine the optimal risk allocation coefficient for the investment portfolio dual risk model constructed in each period.Finally,in order to further improve the performance of the model,a stock price prediction model is embedded in the established dual risk model.The stock price is predicted using machine learning models,deep learning models,and ARIMA models,and the prediction results are then introduced into the investment portfolio model.The empirical results show that:(1)the investment portfolio guided by the dual risk model before and after emergencies has good robustness;(2)Embedding stock price prediction models in dual risk models has more stable returns than models that do not include predicted values;(3)There are differences in the accuracy of stock price prediction between different models.When predicting stock prices through the sliding window,random forest has the best prediction effect,LSTM neural network has the largest deviation from the real value,and GBR model has a predictive ability between the two.