The concept of pairs trading is pretty straightforward. As described in [Gatev et al. (2006)], we first find two stocks that have moved together historically and then monitor the spread between these stocks. If the prices of the two stocks diverge, we short the winner and go long on the loser, hoping that these prices converge in the future. If the spread is mean reverting, it will revert to its historical mean. Then, the positions are reversed and a profit can be made.
There are various frameworks that could be used to identify a pair of stocks and build pairs trading strategies. In this article, we will be discussing a couple of papers related to stochastic control based approaches, which had the highest impact in this domain. We will not be discussing pairs selection techniques here, and interested readers can refer to the Stock Selection Methods using Copula and Machine Learning for Pairs Selection articles. The objective of these methods is to identify the optimal portfolio holdings in the legs of a pairs trade compared to other available assets. Stochastic control theory is used to determine value and optimal policy functions for this portfolio problem. It does sound a bit complicated, but, I’ll try to keep things simple and explain the intuition behind how and why these methods work.