Research On Forecast Of Corn Futures Price

Agriculture is China’s primary industry.The production of crops not only determines the food and clothing of 1.4 billion people,but also provides a solid foundation for the steady development of the national economy.Entering the new era,the prosperity of the country has brought scientific and technological progress,institutional reform and rapid economic development.In this process,the position of agricultural futures gradually improved and occupied the main position in the whole market.In recent years,the market share of corn futures has soared in the domestic futures market.Due to its large demand,high turnover,and the price is greatly affected by the fluctuations of external environment and policies,it is of great theoretical and practical significance to study the prediction of corn futures prices.This paper,with corn futures closing price as the research object,select 243 trading days data in 2020(except holiday period),respectively,set up three kinds of single forecasting model and four kinds of combination forecast model,by comparing the predicted value and actual value of the indicators to determine the effectiveness and accuracy of forecasting model,and analyzes the applicability of the model.Specific results are as follows:(1)The research status of futures price prediction at home and abroad is summarized and commented,and the prediction theory of futures price is mainly introduced from the aspects of technical analysis,traditional time series analysis and neural network.However,due to the limitations of theories and methods,the prediction results also deviate greatly from the expected results.(2)By constructing three single prediction models of ARIMA,GM(1,1)and BP neural network,and four combined prediction models based on dominance matrix method,entropy weight method,variance-covariance method and IOWA operator method,the first 233 data of 2020(from January 2to December 17)were taken as the training set.The actual values of the last 10 data(from December 18 to December 31)were used as test sets to test and evaluate the prediction effect of the model.The gray prediction model GM(1,1)reduced the range of training sets according to the level ratio.MAE,MSE,RMSE and MAPE were selected to evaluate the prediction effect.(3)Through the empirical analysis of corn futures closed prediction and evaluation index of comparison,the paper established four combination forecast model than the other three kinds of single forecasting model of high precision,and because the IOWA operator method of the combination forecast model can according to each single forecast model prediction accuracy of the orderly,at different moments,empowerment,Thus more scientific and practical.Therefore,the combination model based on IOWA operator method is the most satisfactory.(4)In order to verify the applicability of the portfolio model under different time spans,the corn futures closing price of 20 trading days from January 4,2021 to January 31,2021 is taken as the research object.Based on the prediction theory and method of long-time span,the short-term prediction research is conducted.From the prediction results,under any time span,the prediction performance of the combined model has been greatly improved compared with that of the single model,and the combined prediction model using IOWA operator method to calculate the weight has the most outstanding performance among the four combined models,which confirms the strong applicability of the combined model.In conclusion,when the futures market runs smoothly and the volatility is small,research on the prediction model of corn futures price portfolio can find a prediction method with small error,provide a reliable basis for investors to make situation judgment and investment behavior,and reduce investment risks.