Pairs trading using cointegration in pairs of stocks

Abstract

The aim of this project is to implement pair trading strategy, which aims to generate profits in any market conditions by examining the cointegration between a pair of stocks. Pair Trading, also known as a relative spread trading, is a strategy that allows a trader to benefit from the relative price movements of two stocks. A trader can capture the anomalies, relative strength or fundamental differences in the two stocks to create profit opportunities. Pair Trading primarily involves finding correlated stocks and exploiting the volatile market conditions, which lead to a diversion in their correlation. A trader takes a short position in one stock and simultaneously takes a long position in the other. If the market goes down, the short position makes money. On the other hand, if the market goes up, the long position makes money. Creating such a portfolio enables the investor to hedge the exposure to the market. Furthermore, by taking a long-short position on this pair, when prices diverge, and then closing the position when the spread retreats to its mean or a threshold, a profit is earned. In this project, we implement pair trading strategy using an Ornstein-Uhlenbeck (OU) process based spread model, is applied on stocks from three different sectors - Energy, HealthCare and Banking of the NYSE. Stocks were selected based on a combination of Distance Test, ADF Test and Granger-Causality Test. The paper concludes by summarizing the performance of this strategy and offers possible future enhancements and applying it to more complex scenarios.