What it means
Correlation measures how two return series move together, from −1 (opposite) through 0 (unrelated) to +1 (together). A rolling correlation recalculates it over a moving window (for example the last 20 or 60 bars), so you can see how the relationship changes.
Why traders use it
Intermarket ideas such as "gold moves against the dollar" are only useful while they hold. A rolling correlation shows whether they hold now, not just in the past.
Chart examples
Correlation that flips sign
How to identify it objectively
- 1Use returns (percentage changes), not prices.
- 2Pick a fixed window length.
- 3Plot the rolling correlation with a zero line.
- 4Note when it crosses zero or changes sign.
Bullish use
- 1Use a related market as confirmation only while the correlation is strong and stable.
Bearish use
- 1When the correlation flips, stop relying on the old relationship.
What confirms it
- Correlation stable across two window lengths.
What invalidates it
- A flip in sign.
- Correlation near zero.
How a hypothetical setup is built
Entry
Not an entry rule: a filter for when an intermarket idea applies.
Stop
From the trade setup.
Target
From the trade setup.
A hypothetical illustration of how the concept is turned into a plan, not a recommendation. Practise it on a paper account first.
Related concepts
- Intermarket AnalysisUS Dollar and Gold
- Risk Management & StatisticsLook-Ahead Bias and Overfitting
- Intermarket AnalysisRisk-On and Risk-Off
Historical testing
Not tested on The Tradeskill. The drawing is schematic: a strongly negative correlation weakens and turns positive.
Limitations: when it fails
- Correlation is not causation.
- Short windows are noisy; long windows react late.
- Using prices instead of returns gives misleading values.
Knowledge check
Pick an answer to see the explanation. Nothing is saved.