Statistical Arbitrage Of Commodity Futures Research By High Frequency Data

For statistical arbitrage, due to the lack of market system, such as short selling mechanism, make its slow development, it was not until2010that the launch of margin trading on March31, April16, the csi300index futures contract officially listed trading-market shorting mechanism will be officially formed in our country. Subsequently, margin target range expanding evening, Treasury futures market, commodity trading, simulation as well as the recent launch of the stock index futures and options trading, to statistical arbitrage bring opportunities and challenges in the development of our country. And use of statistical arbitrage strategy body, also is quietly budding in our country. However, the current domestic use of statistical arbitrage strategy of quantitative hedge funds also very few.Along with our country shorting mechanism gradually formed and quantitative investment in domestic financial circles gradually heating up, how to make good use of programming trading in financial markets this sword, become China’s large financial institutions workers hot issues of concern. In this paper, according to the liquidity and stability principle, selection of the Shanghai futures exchange rebar RBI305and RB1310two contracts for statistical arbitrage, through the cointegration test and error correction model, found that5seconds data the most optimal average regeneration. So eventually choose RBI305and RBI310two contracts on December26,2012to February26,2013a total of36days5seconds to establish commodity futures under the high frequency data of four kinds of statistical arbitrage strategy:GARCH model with internal conditions of stop-loss strategy, GARCH model with external stop-loss conditions, with internal stop conditions of standard deviation, standard deviation with external conditions of stop-loss strategy.Through the analysis:(1) choose the liquidity and stability of good varieties can help statistical arbitrage strategy of risk control;(2) both of GARCH model strategy and the standard deviation has advantages and disadvantages;(3) the strategy with an external stop-loss strategy is better than that of with internal stop condition of strategy;(4) the policy model meets the strategy your basic assumptions, strategy is an important factor to effect is good or bad.