Pairs trading is a quantitative trading strategy that is often discussed in the academic as well as practitioners’ literature. We have written about this trading strategy extensively from different perspectives. In this post, we’re going to look at the risk/PnL drivers of the pairs trading strategy.
Reference [1] pointed out that the profit of pairs trading comes from the spreads between small- and large-cap stocks and between value and growth stocks in addition to the spread between high- and intermediate-grade corporate bonds and shifts in the yield curve.
These profits are uncorrelated to the S&P 500; however, they do exhibit low sensitivity to the spreads between small and large stocks and between value and growth stocks in addition to the spread between high- and intermediate-grade corporate bonds and shifts in the yield curve. In addition to risk and transaction cost, we rule out several explanations for the pairs trading profits, including mean reversion as previously documented in the literature, unrealized bankruptcy risk, and the inability of arbitrageurs to take advantage of the profits because of short-sale constraints
However, in a recent publication [2], other authors linked the profitability of pairs trading to market liquidity,
Our results have shown that a simple pairs trading strategy building on an unsupervised machine learning approach does not generate sufficient excess returns to cover a conservative estimate of explicit transaction costs on the S&P500. Conversely, the same trading strategy appears to be profitable on OSE even when adjusting for both explicit and implicit transaction costs. We have shown that the profitability of pairs trading appears to be closely related to the market liquidity of the stocks that are traded, which might explain why the trading strategy appears to be more profitable at OSE.
In other words, using their proposed approach, which consists of first using Principal Component Analysis to divide stocks into clusters and then using cointegration to select the pairs, the strategy is not profitable in SP500, but in OSE, and the profitability is closely related to the market liquidity.
These articles raised some interesting questions,
- Is pairs trading still profitable in the US markets,
- What truly is the source of excess return, if any,
- Does the profitability depend on strategy design?
Let us know what you think.
References
[1] E. Gatev, WN. Goetzmann, KG. Rouwenhorst, Pairs Trading: Performance of a Relative-Value Arbitrage Rule, The Review of Financial Studies, Volume 19, Issue 3, Fall 2006
[2] A. Høeg, EK. Aares, Statistical Arbitrage Trading using an unsupervised machine learning approach: is liquidity a predictor of profitability?, BI Norwegian Business School, 2021.
Further questions
What's your question? Ask it in the discussion forum
Have an answer to the questions below? Post it here or in the forum

COLOGNE, Germany — In a historic breakthrough for sustainable filmmaking, Green Toolkit and A24 have achieved what no other collaboration has before: ISO 14067 certification for a major international film. The upcoming feature film Mother Mary has officially become the first production worldwide to meet…

1NCE completes its largest funding round to date; the company has raised the equivalent of $160 million USD in total funding since 2017. 1NCE intends to broaden its product portfolio with a next generation Software as a Service (SaaS) platform, add new AI tools, continue…

MARKHAM, Ontario — Canada’s leading defence and military magazine, CANADIAN DEFENCE REVIEW, has released its annual Survey & Ranking of Canada’s Top 100 Defence Companies. For 2025 CDR has named CAE its Top Defence Company and it’s the third time for the company at the…

BOSTON & PARIS — Earnix, the leading Dynamic AI platform for Insurance, today announced that it has signed a definitive agreement to acquire Zelros, a pioneer in Generative AI for insurers and banks. This innovative collaboration will bring the power of predictive, generative, and agentic…