A Study Of C-SVM Investment Strategy For Treasury Futures In China

With the development of market economy in our country,the futures market has gradually played more important role in the financial market,which is irreplaceable to the futures price discovery and risk aversion in the field of economic.As the winner of the Nobel Prize for economics Merton Miller said " The economic system without futures market can't be deserved to be called market economy".Treasury bonds futures in our country launched a little late.Just in September,2013,our country's five-year Treasury bonds futures traded in China.But in a short span of two years,the average daily volume rose remarkably.At present,the theoretical study of the futures price discovery at home and abroad is very rich.However,due to the complex internal structure and various factors affected by the market interaction,it is hard for investors to make informed investment decisions with correct futures price prediction.After the application of the traditional technical analysis and fundamental analysis in futures market is limited.The artificial intelligence and statistical learning theory,represented by the support vector machine,which has provided a new direction for futures price discovery and transaction decision.Support vector machine is based on statistical learning theory,which can effectively excavate the inherent law of high dimensional time series under the condition of small sample.Support vector machine in the data processing makes it very suitable for use in the initial stage of the national debt futures market.In recent years,support vector machine has been gradually applied to the stock market,and its use in the national debt futures market is still blank.This paper will take the current national debt futures market as the research object,based on the support vector machine construction trading strategy model,and optimize the relevant parameters.This paper first introduces the related theoretical background,including the basic theory of support vector machine(SVM)and the futures market.It not only studies the Treasury futures varieties of characteristics and the method to predict the price of transactions,and introduces the development of the theory of support vector machine,then using support vector institutions to make C-static simulation for the SVM model,and building more practical dynamic model based on the optimal parameters of the static simulation,stop-loss mechanism will be joined in which to improve its accuracy of trading signal recognition.Finally to promote SVM model constructed in the overall market stably make profit.