The Research Of China Market Pair Trading Based On Markov Regime Switching Model

Since the concept of statistical arbitrage was put forward,pair trading has been favored by domestic and foreign investors.As a market-neutral strategy,paired trading hedges a pair of highly related stocks or other assets to eliminate systemic market risks and obtain stable excess returns.Pair trading believes that the spread sequence meets the mean recovery process,that is,no matter how the spread changes,it will return to the long-term mean level.The most commonly used method of pair trading is the cointegration method.By looking for a combination that has a long-term cointegration relationship,buying securities with an underestimated price and selling securities with an overestimated price,waits for the price difference to return to the average value for reverse operation Get revenue.In recent years,more and more studies have shown that traditional cointegration pair trading strategies cannot completely avoid market risks,especially during periods of market volatility,cointegration strategies often perform poorly and even show negative returns.In order to solve the problem that the traditional cointegration model is not effective when the market structure changes,this paper introduces a Markov Regime Switching Model based on the Ornstein-Uhlenbeck(O-U)process to describe the spread sequence.Because the background of the pair trading strategy is that the spread sequence has the mean recovery characteristic,this article describes the mean recovery process of the spread sequence through the O-U process.The Markov state transition model is a method for processing variable structure data.It was first used by Hamilton to classify the US market state.It uses a Markov chain to describe the current market state implied by the current spread sequence.The state at each moment is only related to the state at the previous moment.Through modeling,this paper divides the spread sequence into two states:high volatility and low volatility.It is found that it is the existence of high volatility that affects the benefits of traditional cointegration models.Subsequently,based on the Markov Regime Switching Model,this paper gives the calculation process of the optimal trading signal under the O-U process.The approximate analytical solution of the trading signal is obtained by maximizing the expected return per unit time.Compared with traditional fixed trading signals,this method has a theoretical basis and is more reliable.In the empirical research part,this article takes the daily closing dataof 202 financial real estate stocks as the research object,and the research time is from January 1,2014 to January 1,2019.This article compares the paired trading strategy based on the state transition model with the traditional cointegration pair trading strategy,and finds the advantages of the strategy by analyzing indicators such as cumulative return rate,Sharperate,trading win rate,and maximum drawdown.Finally,it shows that the investment portfolio constructed based on the state transition model has higher average returns and better risk control capabilities.At the same time,in the selection of trading signals,this method also has certain advantagescompared with the traditional fixed threshold method.