Research On Microstructure Of Futures Market Based On Parallel Processing Technology

Market microstructure theory is one of the three fastest growing area of finance. This paper belong to the scope of market microstructure. We research the market microstructure of China future market from three dimensions:market mechanism, investor behavior and market operating quality. In the part of market mechanism we compare the market-maker market with bidding market. Introduce the risk management system of the China future market. Then, We analysis the structure of China’s futures market investors, that the high proportion of individual investors led market speculative. The computer processing capacity constraint the analysis of high frequency finance data, We propose to use cluster parallel processing technology to solve this bottlenecks. Introduce the method of parallel processing technology in detail for the first time. Apply the parallel processing technology in empirical research of future market. Use "tick to tick" data analysis the "calendar effect" of cotton, copper, zinc and gold futures, pointed out that the futures price volatility, spreads and other indicators show the "L" type intra-day fluctuations. The intraday fluctuations comes from gap between the inter-day information amount and the intra-day information amount. For the uneven distributed "tick to tick" data, we use the ACD model to test the feature of clustering of cotton future. The test show that transaction duration is effect by the one duration before him. We compare the estimate result between the EACD model and the WACD model. The WACD and the EACD dose not show obvious differences in the estimate outcomes. In the final of the article, we discussed the high-frequency automated trading, which is the application of market microstructure in the area of financial practice.