Assessing Whether Cointegration-Based Pairs Are Spurious

Summary

The document describes a search across many NYSE ticker pairs, testing each combination for cointegration or stationarity. One pair, whose companies appear unrelated and belong to different sectors, showed a strong one-year backtest. The author worries that the apparent relationship may be accidental and asks whether candidate pairs should instead be selected within sectors or by another systematic process.

Its central lesson is the risk of mistaking a result found through broad repeated testing for a durable relationship. The reported backtest is only an observation from the search; the document provides no validation sample, multiple-testing correction, economic explanation, or trading-cost analysis. It also contains no answer recommending a particular candidate-generation method. The pair’s performance should therefore not be read as evidence that the strategy will persist, and the document leaves the practical screening and validation process unresolved.

Key ideas

• Testing many ticker combinations can produce apparently successful pairs by chance.
• A high backtest return alone does not establish a reliable cointegrating relationship.
• An unexplained relationship between unrelated companies prompts concerns about spurious findings.
• The document asks about sector-based screening but gives no recommended method or validation results.

Tags

# How do we know if cointegration is down to luck?


# How do we know if cointegration is down to luck?












I have done some analysis of various pairs of tickers on the NYSE. I did a brain dead algorithm to come up with all combinations of pairs and then checked all pairs for cointegration/stationarity.

I found one pair that backtested to 40% returns over the last year. But the components of the pair are totally unrelated, not even in the same sector, I cannot dream up a scenario that explains the cointegration. So I guess it's down to luck and I should leave it alone.

How should I have gone about my search? Start with a particular sector? Is there a systematic way of finding candidate pairs?

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.