Stationarity Of Bias Under Jump-diffusion Model And Its Application In High-frequency Trading

High-frequency trading is a trading method that automatically buys and sells assets in a very short time through the computer algorithm,and the premise of high-frequency trading is that it must be supported by theoretical analysis and trading strategies.In high-frequency trading,one buying and selling circle only costs a short time,so the trading process is mainly based on technical indicators while the fundamental aspects have rarely changed.On the other hand,asset prices often jump in the financial market which result in extreme risk.The thesis attempts to construct a stationary technical indicator for the financial market with jumping behavior as the core of highfrequency trading strategy.Bias is a technical indicator that is often used to construct Bias strategy in high-frequency trading.Firstly,this paper proves the stationarity of the Bias under the jump-diffusion model and combines the result with the Lee?Mykland jump test to construct a new Bias-J strategy.After that,we select high-frequency trading data by one-minute from January to December 2020 to conduct an empirical analysis of the Bias-J strategy,which verifies the effectiveness of the strategy and the stationarity of returns.The trading assets include CSI300,SSE50,and the stocks of Ping An Insurance Company of China and Kweichow Moutai.Furthermore,by comparing the results of the Bias strategy and the Bias-J strategy,it is found that considering the impact of jumps in highfrequency trading strategies can promote the Sharpe ratio,and a new idea of using jumps to increase return is proposed.