Abstract by Matthew Zackrison
Pairs Trading Algorithm
There are many strategies that investors use to decide when to buy and sell stocks. One popular strategy is called pairs trading, which is designed to exploit fluctuations from the long-run equilibrium between two stocks. This project sets out to assess how sensitive the performance of this strategy is to the many decisions that an investor must make when implementing the pairs trading algorithm, such as choosing a threshold level, and determining a correlation level between the two stocks. In this project we are choosing a threshold value based on the number of standard deviations that the ratio moves above or below the long term average, and will also test different correlation values between the two stocks. We will perform a simulation study to determine a method to find the optimal threshold value to maximize profits, to see if that threshold value is related to the correlation between the two stocks, and to see how the correlation between stocks affects the final profit. Finally, we will apply the algorithm to three pairs of stocks and measure the performance of the algorithm.