High Frequency Trading Analysis Based On Stable Distribution

This paper studies how to improve the return of the portfolio of high-frequency trading strategies by constructing the trading assets portfolio of non-normal sta-ble distributions.According to Efficient Market Hypothesis(EMH),asset returns are normally distributed in efficient or weak efficient market. Comparatively,behavioral finance researchers have found that the market is frequently inefficient because of the participants’over-reaction to information,and that assets return exhibits the severe phenomena of excess kurtosis and heavy tails.In general,trading strategies are usually designed to capture certain market efficiency or heavy tail phenomena. On the oth-er hand,compared with the low-frequency data,the high-frequency data exhibit the feature of heavier tail, so in this paper we mainly focus on high frequency trading strategies.Then, we first select a portfolio of different assets such that the portfolio re-turn follows a stable distribution with a high degree of heavy tail,then according to the proportion of the portfolio we allocate investment capital in every portfolio of high-frequency trading strategies. Thus, we establish a relatively efficient portfolio of HFT strategies.The empirical study shows that, the greater the degree of fat-tail,the more efficient the high-frequency trading strategies.So we may improve the efficiency of high-frequency trading strategies by selecting a fat-tailed distribution portfolio.