| This paper uses VAR and its extension model to conduct empirical research,trying to analyze the fluctuation of RMB exchange rate to have a transmission effect on China’s industry income fluctuation.The industry and RMB daily data from January 2008 to December 2018 were analyzed,and the volatility was calculated on a weekly basis.The ability of RMB exchange rate fluctuations to explain the fluctuations of China’s industry returns was reported from within and outside the sample.Secondly,taking the energy industry as an example,the error reduction value ΔROOS2 and its P value are calculated from the exchange rates of two kinds of CNY/USD and CNY/EUR respectively from the recursive and rolling window prediction methods outside the sample,and then the model is extended to Two asymmetry models and the prediction of the energy industry for a longer period of the future 1 will lead to the following conclusions:1.Whether it is within the sample or outside the sample,the linear model is either a nonlinear model,or a longer period is predicted.The predicted values of the two RMB exchange rates and their P values are very significant,indicating the fluctuation of the RMB exchange rate.It has a strong ability to explain fluctuations in China’s industry earnings.2.No matter what model or exchange rate,there is a common feature,that is,as the sub-interval shrinks,the predictive power of the independent variable exchange rate is also rising,and the prediction ability is the strongest in the nearest sub-interval.3.It is found that in the linear model,the CNY/EUR exchange rate forecasting ability is stronger than the CNY/USD exchange rate forecasting ability;the SSE comprehensive index forecasting result is slightly higher than the energy industry;the recursive method’s forecasting ability is better than the rolling method forecasting ability.Strong.4.In the analysis of the results of the nonlinear model,there is a special point outside the commonality with the linear model.In the conclusion of the asymmetric model of the recursive prediction method,there is no evidence that the result of the asymmetric model is better than linear.The fitting effect of the model,and the analysis of the asymmetry model under the rolling prediction method,the fitting effect of the linear model is better than that of the nonlinear model.5.When comparing the long-term predictions,it is different from the linear ones:(1)In the long-period prediction process,the forecasting ability of the RMB exchange rate is significant,and the shorter the forecast future period,the stronger the exchange rate has an impact on the auto-regressive prediction.(2)CNY/USD has strong ability to interpret income fluctuations in the energy industry under the recursive window forecasting method.The CNY/EUR has a stronger ability to interpret industry income fluctuations under the rolling window forecasting method.(3)Whether in the recursive window or the rolling window prediction method,the CNY/USD interpretation ability is better than the CNY/EUR interpretation ability in predicting long-term industry returns. |