Study On Risk Measurement Of Chinese Soybean Futures Market

With the globalization of the world economy,the futures market has been developing rapidly and the risk in the futures market is also increasing.The unique leverage effect of futures makes the futures market more risky.The Chinese soybean futures price is affected by many factors,including US soybean futures price,US dollar index,China domestic soybean output and demand and so on,which has many risk channels.At the same time,the Chinese soybean futures price is related to the interests of upstream and downstream participants such as soybean growers,consumers and processors,and the interests of participants in soybean futures market.Therefore,the risk management in the futures market is the focus of participants of futures market,soybean growers,soybean processors,soybean consumers and market regulators.China is a big country in soybean consumption,which has an important impact on international prices of soybean.The Chinese Number 1 soybean Futures(It is called the Chinese soybean futures in this thesis)occupies an important position in the agricultural products market.March 15,2002 is a special day.On this day,the Chinese Number 1 soybean Futures is officially listed on the Dalian Commodity Exchange.The listing of soybean contracts has promoted the development of the Chinese agricultural products market.To make Chinese soybean futures market run smoothly,we must make accurate risk measurement to it.In this thesis,the closing price of the Chinese soybean futures contracts of the recent 5 years is selected as the sample data.And the most widely used risk measurement method that is VaR method is selected for risk measurement.And the most convenient method of VaR method,variance covariance method,is used for empirical research.Since futures have the nature of two-way trading,this thesis gives different VaR formulas to short position and long position of soybean futures.Normality test shows that the Chinese soybean futures don't obey normal distribution,with high apex and heavy tail.So this thesis introduces the GARCH family model into the empirical test to reflect the volatility characteristics of soybean futures return rates,making the empirical results more accurate.The representative models of GARCH(1,1),TARCH(1,1),EGARCH(1,1)models in the GARCH family model are selected to make an empirical analysis.And the parameters of the three models are respectively estimated under normal distribution,T distribution and GED distribution.Using the AIC and SIC minimum criteria,the parameters of GARCH model that obeys normal distribution,the TARCH model that obeys GED distribution and the EGARCH model that obeys GED distribution are best estimated.The VaR values of the long position and the short position of soybean futures based on the three models of soybean futures are calculated respectively.It is concluded that when the soybean futures are selling,the VaR-TARCH-GED model has the lowest failure rate and the best risk measurement.When the soybean futures are buying,the VaR-EGARCH-GED model has the lowest failure rate and the best risk measurement.Then the accuracy of VaR value is tested.The result shows that both the soybean futures short position and long position models of empirical analysis pass through the test.Finally,some suggestions on risk management in Chinese futures market are put forwarded according to the results of empirical analysis.