Vol: 10, Issue: 3 2020

Page: 47-61

Pairs Trading Based Statistical Arbitrage Using Cointegration Approach and Kalman Filter

Lavaneesh Sharma

In this paper we explore the pairs trading based statistical arbitrage technique. The proposed pairs trading methodologies was employed to equity trading systems to find the stocks and their underlying ETF’s and was able to identify the relative statistical mispricing between the prices of stock-ETF pairs, using regression residuals, and to modeled them as natural mean-reversion processes with a short holding period in the U.S. equities market under any market cycle conditions. The strategy follows a twostep process. First in the formation phase we select and identify securities that have long term mean reverting spread; the spread having high standard deviation to allow a profitable strategy. In the actual trading phase, we define rules for trading entry and exit points as prices diverge and converge. Finally, we the back test the strategy using back trader.