Statistical arbitrage pairs : can cointegration capture market neutral profits?

Abstract

We back-test a statistical arbitrage strategy, pairs trading, over the ten year period 01.01.2003 – 31.12.2012 at the Oslo Stock Exchange. We construct an unbiased dataset, where stocks are matched into pairs using a cointegration approach and traded according to a set of pre specified rules. The strategy yields consistent negative returns independent of parameterisation of entryand exit thresholds. Our findings are in line with previous literature, where we support the view that absence of profits is not necessarily due to increased activity among hedge funds, but rather changes in fundamental factors governing the relationships between stocks.