Font Size: a A A

Study On Pairs Trading Strategy In China's A-Share Market

Posted on:2020-05-23Degree:MasterType:Thesis
Country:ChinaCandidate:Y R FengFull Text:PDF
GTID:2439330572471589Subject:Applied statistics
Abstract/Summary:PDF Full Text Request
The pairs trading strategy incorporates statistical and econometric research methods and it's a branch of quantitative investment trading strategies.This concept was originated in Wall Street in the mid-1980s and was widely used in mature capital markets in the West.It was mainly used by hedge funds to generate revenue and achieved success.The pairs trading strategy is a market-neutral strategy.The basic idea of the strategy is to find stocks with the same or similar historical trend,when the spread deviates from the historical average,they will short the relatively strong stocks recently while long the relatively weak stocks.When the spread returns to equilibrium level,it's time to perform the opposite operation and obtain the spread income.One of the preconditions for the operation of the pairs trading strategy is that the market has a short-selling mechanism.But China's securities market has always been a unilateral operating mode,which means the securities only can be long but not short.With the implementation of securities margin trading and the listing of the CSI 300 stock index futures in 2010,China's securities market has a short-selling mechanism,which provides a good platform for the use of pairs trading strategy.At present,China's research on pairs trading is still in its infancy,and most studies are concentrated in the futures market.In this paper,the pairs trading strategy is applied to the A-share market,and 948 underlying stocks of securities margin trading are selected as the research objects.We select the daily closing prices of stocks as sample data.In the setting of the sample interval,the whole year of 2017 is selected as the in-sample interval and the whole year of 2018 as the out-of-sample interval.The implementation of pairs trading strategy mainly includes two important aspects:(1)choose the pairs stock;(2)set the trading strategy.When screening the pairs portfolio,we firstly make logarithmic processing of stock price,and use the correlation analysis and cointegration test to verify the long-term cointegration relationship between the two stocks of Ningbo Bank and Ping'an Bank.Then,we use the error correction model to estimate the short-term deviation of the stock spread between Ningbo Bank and Ping'an Bank relative to its long-term equilibrium,and determine the trading position of the two stocks.When setting up the trading strategy,the traditional pairs trading strategy considers that the standard deviation of the spread of the pairs portfolio is stable,so a certain multiple of the fixed standard deviation is used as the trigger signal for the trading of building,closing and stopping losses.However,the research of econometrics shows that most financial time series will show heteroscedasticity.Due to the superiority of GARCH model in analyzing and simulating the volatility of time series,in this paper we improve the strategy based on the traditional model by using GARCH model.In this paper,we use two types of trading signals:(1)the fixed standard deviation of the time series of the stock spread under the traditional model;(2)the time-varying standard deviation of pairs portfolio spread calculated by GARCH model.We use one time standard deviation as the building signal,three times standard deviation as the stop loss signal,and the spread recovery or crossing the mean as the closing signal.We conduct a backtest analysis of the trading strategy under the traditional model and the GARCH model.The results show that:under the premise of considering the trading cost,the pairs trading strategy under the two models get good returns both in and out of samples,indicating the effectiveness of the pairs trading strategy in the A-share market.At the same time,the results also show that GARCH model can better simulate the volatility of the spread sequence compared with the traditional model,and the obtained returns are more robust.
Keywords/Search Tags:Pairs Trading, Cointegration, Traditional Model, GARCH Model
PDF Full Text Request
Related items