Research On The Volatility Of Soybean And Cotton Futures Prices Based On MS-GARCH Model

Agricultural products are essential for human survival.The price of agricultural products is closely related to the everyone.Agricultural futures market has gradually become an important way to optimize the agricultural structure,improve the degree of agricultural modernization,and ensure the income of farmers.It is also a safe haven tool for agricultural enterprises and farmers to arrange production rationally.Therefore,the study of agricultural futures market has practical significance.Agricultural futures prices show periodicity and volatility,which has a certain inherent regulation.In recent years,the price of agricultural futures market fluctuates frequently in China.Taking soybean futures market as an example,it experienced "bull before bear" in 2008,and the price fluctuated dramatically.Agricultural futures prices are affected by many factors,occasionally showing abnormal volatility characteristics,not only will affect the preferences of investors,but also will have a significant impact on enterprises' production hedging.Therefore,the study of the mechanism of agricultural futures price fluctuation can provide a reference for investors to hedge risks and hedge.This paper focuses on the agricultural futures market,aiming at the yield series of soybean and maize futures prices,and uses Markov regime-switching GARCH model to model,in order to study its inherent volatility mechanism.Firstly,we describe the statistic characteristics of the sequence,such as stability and correlation,and then use the estimation of GARCH family model to further prove its asymmetry and thick-tailed distribution.Finally,we use MS-GARCH model and use maximum likelihood estimation to analyze the fluctuation of agricultural futures prices in China.The empirical results show that the fitting effect of MS-GARCH model is better than that of traditional single-regime GARCH model.The number of optimal transfer mechanisms of soybean futures price return series is k=3.The optimal fitting model is MS(3)-GARCH(1,1)-SSTD model.In the low fluctuation state,the duration is longer,16.98 days;in the medium fluctuation state,soybean lasts the longest,41.67 days;in the high fluctuation state,the duration is the shortest,1.47 days.The duration of high volatility of soybean futures price return is the shortest,and it is easier to change from high volatility to low and medium volatility.The number of optimal transfer mechanisms of cotton futures price return series is k=3.The optimal fitting model is MS(3)-GARCH(1,1)-STD model.The duration is shorter in low fluctuation state(25.58 days);51.02 days in medium fluctuation state;and 98.04 days in high fluctuation state.Cotton futures prices are in a high volatility state for a longer period of time,and it is easier to change from a low volatility state to a high volatility state.In addition,we compared and analyzed the smoothing probability of soybean and cotton,and did the Kendall synergy coefficient test.The test results show that the correlation of soybean and cotton futures in a fluctuating state is low,the commodity attributes of soybean and cotton futures are strong,and the financial attributes are weak.The impact of the financial market on the volatility of soybean and cotton futures prices is relatively small.Investors can invest in soybean and cotton futures to avoid risks in financial markets.