Pair Trading and VAR Analysis Applied to Energy Stocks

33 Pages Posted: 24 Sep 2019

See all articles by Carlos Salas Najera

Carlos Salas Najera

University of London; New York City Data Science Academy; CQF Institute

Date Written: June 2019

Abstract

This project is based on developing a straightforward application to apply a simple mean-reversion strategy tailored to Energy stocks (XOM, COP, CVX) and two ETFs (XLE and SPY). Firstly, a VAR model is built using security price returns as first step before conducting IRF (Impulse Response Function) analysis and Granger Causality tests to obtain preliminary clues about the most interesting pair trading combinations. Secondly, 2-step Engle Granger is used to provide a formal framework to shortlist the best pair combinations, with the first stage based on testing each pair for cointegration, and the second step allowing to test for the existence of an ECM (Error Correction Mechanism) that brings the cointegrated pair ultimately towards a long term equilibrium. The result obtained downsized significantly the pair combinations more likely to be successful with COP-XOM as one of the best choices. The last section delves into backtesting of the COP-XOM pair starting by identifying key strategy parameters using an Ornstien-Uhlenbeck (OU) process to model the spread. Finally, the backtested strategy returns are compared against a sample of Fama-French risk factors in order to ascertain whether or not the alpha delivered by the cointegration exit/entry points is significant.

Keywords: Pairs Trading, Energy, Multi-Dimensional, VAR, Backtesting, IRF, Energy

JEL Classification: C01, C02, C22

Suggested Citation

Salas Najera, Carlos, Pair Trading and VAR Analysis Applied to Energy Stocks (June 2019). Available at SSRN: https://ssrn.com/abstract=3458559 or http://dx.doi.org/10.2139/ssrn.3458559

Carlos Salas Najera (Contact Author)

University of London

Senate House
Malet Street
London, WC1E 7HU
United Kingdom

New York City Data Science Academy ( email )

500 8th Avenue #905
, New York, NY NYC 10018
United States

CQF Institute ( email )

30 N Colonnade, Canary Wharf
London, Greater London E14 5GN
United Kingdom

Here is the Coronavirus
related research on SSRN

Paper statistics

Downloads
74
Abstract Views
353
rank
331,006
PlumX Metrics