Research On The Price Prediction And Trading Strategy Of Stock Index Futures Based On Markov Revision

With the development of the Chinese financial market,stock index futures have become important financial instruments for hedging,price discovery,risk management,and speculative arbitrage.Therefore,forecasting stock index futures prices can better guide investors in decision-making and has practical significance.However,the characteristics of stock index volatility,large amplitude,and long duration make forecasting difficult.It is difficult to apply the nonlinear data characteristics of stock index futures using traditional time series forecasting methods,and the use of machine learning methods for forecasting has a small amount of data and is easy to pass Risk of fitting.In this context,based on the study of general forecasting models,this paper uses Markov's method to modify the forecasting results,and conducts an empirical analysis of futures strategies based on the forecast result.In this paper,the grey prediction model and the Long Short Term Memory neural network model are used to predict the closing price of the main contract of the 500-finger futures.After the prediction results are obtained,the state space is divided according to the relative error distribution between the predicted value and the real value.Then,the Markov state transfer model is used to correct the original prediction results.The results show that the Markov modified model improves the lag effect of grey prediction,and weakens the problem that the prediction results of Long Short Term Memory neural network are generally smaller than the real value,the relative error of prediction becomes smaller,and the judgment of the direction of rise and fall is more accurate.After that,based on the prediction results of grey prediction model,Long Short Term Memory neural network and Markov modified model,the improved average strategy is constructed respectively,and the contrast strategy of random buying or selling is added to verify the effectiveness of the improved prediction model.It is found that the strategy based on grey prediction model and short-term and short-term memory neural network prediction results is generally effective,and basically cannot obtain higher returns than the benchmark and excess risk returns.Markov modified model performed better,with higher excess returns,Sharp ratio,Alpha value and smaller maximum withdrawal rate,volatility.Finally,this paper summarizes and gives the risk tips and suggestions according to the actual situation,and puts forward the shortcomings of the article.All in all,this paper shows that the Markov modified model can improve the prediction accuracy and the rate of return of strategic transactions,but it still needs to consider the investor sentiment,policy factors and so on in reality,and trading strategy should be constructed reasonably using the forecast results.