Cointegration, Stationarity, and the Logic of Pairs Trading
Summary
The document introduces cointegration as a long-run relationship in which individually nonstationary series can have a linear combination that is stationary. It explains stationarity through stable statistical properties over time and contrasts it with series whose mean shifts. Cointegration is presented as an alternative way to model nonstationary financial and economic data, alongside approaches such as differencing.
For pairs trading, the proposed intuition is that the spread between two cointegrated securities fluctuates around a stable level. A trader may buy the relatively cheaper asset and short the more expensive one when the spread widens, then reverse those positions if it converges. The article emphasizes that correlation alone does not establish cointegration and mentions unit-root testing, including the Dickey-Fuller test, as a stationarity check. It is an introductory explanation rather than a full implementation: it supplies no detailed test procedure, trading thresholds, costs, or performance results.
Key ideas
- Cointegration occurs when a linear combination of nonstationary series is stationary.
- A stationary series has statistical properties that remain stable over time.
- Pairs trading can use a mean-reverting spread between cointegrated securities.
- Correlation and cointegration describe different statistical relationships.
- Unit-root tests such as Dickey-Fuller are used to assess stationarity.
Tags
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.