Why Price Correlation Alone Does Not Establish a Pairs Trade
Summary
The discussion explains why high correlation between two price series does not by itself show that a pairs trade will converge. A nonlinear example, where one series is the square of another, can produce strong correlation even though their relationship does not imply a stable spread that returns to a common level. Correlation measures co-movement, not a mean-reverting trading relationship.
It also highlights that correlating raw prices can imply share-count-based position sizing, leaving the portfolio unbalanced in dollar exposure. For strategies weighted by invested value, percentage returns may be a more relevant basis for comparison. A further response warns that apparent relationships during a rising market may break down during corrections, while waiting for convergence can impose substantial carry costs. The answers offer cautions rather than a tested strategy; they do not establish that any particular alternative measure guarantees profitable pairs trading.
Key ideas
- High price correlation does not prove that the spread between two assets is stable or mean reverting.
- A nonlinear relationship can produce strong correlation while the series continue to diverge.
- Correlating raw prices can amount to sizing positions by share count and create imbalanced dollar exposure.
- Percentage changes may better match pair analysis when positions are weighted by dollar value.
- Relationships observed in rising markets may weaken during corrections, and convergence can take costly time.
Tags
Full text
# Why isn't it appropriate to use correlation between prices in a pairs trade strategy?
# Why isn't it appropriate to use correlation between prices in a pairs trade strategy?
I've already seen this question and read through all the answers, but I'm still confused about why you shouldn't use price correlation in a pairs trade strategy.
For example, if I'm looking at the price ratio between two stocks, and I find the current price ratio deviates from the norm, doesn't a high price correlation tell me that the two prices are bound to converge to the same movements and so I have an opportunity to sell high and buy low?
## Answer by madilyn (score 5)
https://quant.stackexchange.com/a/34836
You could, and it doesn't hurt for you to test this yourself. Some of my best work has come from drawing the opposite conclusion to conventional wisdom or stylized "facts" in publications.
That said, it's trivial to construct an example where you won't be able to spread a correlated pair. Suppose the underlying data generation process is $y_t = x_t^2$, you will find very high correlation between the two but obviously $x_t \in \mathbb{o}\left(y_t\right)$ and the two will diverge.
## Answer by amdopt (score 4)
https://quant.stackexchange.com/a/34837
If you are correlating prices that would imply that you are sizing positions based on the number of shares in each position. This can result in a book that is very biased in terms of dollars invested. This is not conventional and actually, makes little sense--most of the time.
Most pair trading strategies weight positions by dollar value which is why normally, percentage changes would be used to correlate.
## Answer by FaceInstitute (score 2)
https://quant.stackexchange.com/a/34903
Whatever correlations appear during a bull direction market (the slow upward grind with rare down days and pallid implied relativity), these correlations undergo deposition to near zero when the market does correct. Their is your price extreme indicator, but often the cost to carry associated with the prolonged time it takes for market correction is too much to bear for most people. It is much like going long volatility in those market conditions (Valid concept, difficult and expensive in reality).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.