Empirical Research On Future-Spot Arbitrage Of Treasury Bond Futures Based On Cointegration

Along with the rapid advance of the interest rate marketization,and the development of financial markets,the requirement of investors for risk hedging is getting stronger and stronger.Treasury bond butures as an important varieties of financial derivatives as well as an important risk management tool to interest rate,will play an important role in establishing of the pricing marketization, in improving of the treasury bond issuance system, in furthering of the reform for the interest rate marketization and in leading of the resource optimal allocation.The 5-year treasury bond futures has officially listed in China Financial Futures Exchange on September 6 in 2013.With the growing scale of treasury bond futures, China Financial Futures Exchange has published the 10-year Treasury Future on March 20 in 2015。 As the second national debt futures varieties in china’s domestic market, the 10-year treasury bond futures listed means that China’s national debt futures market becomes further perfect..As arbitrage trading is one of the three primary trading strategies in the futures markets, the arbitrage activities will play a very important role in treasury futures market。 But we are still lacking of systemic study of arbitrage strategies on treasury futures.Up to now treasury bond futures was officially listed more than one and a half years,and the market has gradually become active.The obvious correlation of prices between futures and the spot,makes future-spot arbitrage rational. Moreover, a large number of researches have shown that both at home and abroad,the method of statistical arbitrage based on the cointegration is a sound way of investment because it combines high security and stable profit. Obviously, it is both of the theoretical value and great practical significance to make empirical research between treasury bond futures and the spot on the statistical arbitrage strategy based on the cointegration.This paper firstly introduces the fundamental knowledge of treasury bond futures, reviews the development course of China’s treasury bond futures, analyzes the experiences and lessons from the "327" event, and introduces the relevant knowledge of treasury ETF and its significance of launch. Secondly,this article uses the closing price datas of TF1503 and treasury ETF(511010) from June 16,2014 to March 13,2015 to make empirical analysis,and uses the closing price datas of TF1506 and treasury ETF(511010) from September 15,2014 to March 25,2015 to test.we use constant variance and time-varying variance to research the statistical arbitrage strategies of future-spot arbitrage in the treasury bond futures. First of all, we analysis the relevance of the price series through the correlation test and the stability through unit root test. Secondly,we use cointegration test and error correction model to build statistical arbitrage model. We use the sample data to build the model, and the data out of the sample to forecast.And we found the problem of time-varying variances. We finally build The GARCH model to improve the first model, and concluded that statistical arbitrage method based on GARCH model is inferior to the constant variance model of statistical arbitrage.Finally it is concluded from the empirical results that statistical arbitrage method based on cointegration on the future-spot arbitrage of treasury bond futures in China is feasible and profitable.