Granger Causality And Its Application In Futures Market

The study of causality always plays an important role in financial analysis,Granger causality test is a simple and practical method,which is widely used in financial analysis.Research on the causality between the market and the market of non-ferrous metals can reveal the transmission mechanism of the cross-product and transnational communication mechanism of market information,especially domestic and international financial risks,which can enhance the understanding of the micro and macro structure of futures.At present,scholars at home and abroad have little research on the two different futures markets,and generally study the trading situation of two futures exchanges in Shanghai and London.It is necessary to study metal futures trading in other markets.Since the yield of nonferrous metals is influenced by many different factors,the causal relationship between variables is not simple.Therefore,it is not comprehensive enough to study the causal relationship between variables only from the linear perspective,and it is particularly important to test the nonlinear Granger causalityThe article has been working on the existing Granger theory of causality,and discovers that the phenomena of multiple colinear often occur in the process of testing.Therefore,the article considers that it is possible to combine the Granger test with the Lasso algorithms which can solve the multicollinearity problem to establish an improved causal verification model.The article has established an improved linear Granger causality model by introducing the Lasso algorithm to improve the traditional linear Granger causality test firstly,and then the article has established an improved nonlinear Granger causality model by introducing the Lasso algorithm to improve the traditional nonlinear Granger causality test based on Taylor expansion.The paper analyzes the actual data of nonferrous metal futures by using the modified causal inspection model,and the objects of examination are the data yield sequences of silver in Shanghai and silver in New York and the exchange rate of RMB and dollar.Firstly,the paper uses the traditional linear Granger method to test the sequences.Secondly,the paper uses the linear model based on the Lasso algorithm to do the causality test.Thirdly,the paper uses the nonlinear model based on the Lasso algorithm to test the sequences.Finally,the paper summarizes and forecasts the improvement of causality test,and puts forward some Suggestions on the development of futures market and investors' investment.