An Empirical Study On Pair Trading Strategy In Chinese A Share Market

In recent years, with the purpose to get absolute returns and trading strategy which are very flexible and diversified, hedge funds increasingly attract the investing public's attention. Pair trading strategy is just one of the key strategies hedge funds used.This paper choose the pair trading strategy as the research content. Firstly this paper reviewed in detail on pair trading strategy domestic and international, and found most of pair trading strategy rearches domestic and international are based on the convergence test, and try to analyze similar extent of the history price movements between two stocks in order to find matching pairs. However, H. Nicolas (2009) proposed a pair trading strategy which is completely different from the previous match methods, which establish the stock portfolio through returns prediction and ranking. This paper empirically studies on these two pair trading strategies using Chinese A-Share market data between 2007 and 2009, in order to test the feasibility and profitability pair trading strategies are used in the Chinese A-Share market, and more makes a further comparative research between two trading strategies.The results shows that these two strategies both can get a higher return rate than the broader market. Through further analysis we found that in the bear market and the adjustment market the return is significantly higher than the broader market, and arbitrage is better. But in the bull market period, the return rate is slightly less than the broader market, the effect of arbitrage is worse. From the comparison of two pair trading strategies, average return of the trading strategy based on convergence test is higher. However, due to the return volatility of the trading strategy based on earnings forecasts and ranking is even smaller, so we get that the Sharpe Index of the trading strategiy based on earnings forecasts and ranking is higher. It shows that the performance of the trading strategy based on earnings forecasts and ranking is better than the trading strategy based on convergence test. In addition, we find that the trading strategy based on earnings forecasts and ranking got low correlation with the broader market.This paper is one to one matching method, so in future research we could study the trading strategy with portfolio which contains more pairs, and a resultant strategy consists of pair trading strategy and other trading strategies.