Forex pair correlation

Frequently asked questions

What does a coefficient of +0.8 mean?

It means strong positive correlation — the two pairs tend to rise and fall together. It doesn't guarantee future correlation, only the selected period.

Why is the diagonal 1.00 everywhere?

A pair is perfectly correlated with itself: r = 1.00. The diagonal is a visual reference, not useful trading information.

What data is the calculation based on?

On daily ECB reference rates, distributed for free via Frankfurter. Frankfurter publishes data after the European market close (~16:00 CET). Our page refreshes every 12 hours.

How do correlations change around major macro events?

Around Fed decisions, inflation prints and crises, correlations typically spike — pairs that are normally independent start moving together because they all respond to the same risk-on / risk-off shock. Recompute the matrix before opening large positions around scheduled events, and reduce position sizes if the pairs you trade become strongly correlated.

What is the formula?

For each pair we compute daily log returns r_t = ln(c_t / c_{t-1}). Then we apply the Pearson coefficient between the return series. The logarithm provides symmetry between upward and downward moves.

What do the arrows (↑↑, ↑, →, ↓, ↓↓) in the matrix mean?

The arrows duplicate the colour-coding so the matrix is readable even without colours: ↑↑ — strong positive correlation (r ≥ +0.7, pairs move almost identically); ↑ — moderate positive (+0.3 ≤ r < +0.7, similar trend but weaker); → — negligible (−0.3 < r < +0.3, pairs are effectively independent); ↓ — moderate negative (−0.7 < r ≤ −0.3, weak inverse trend); ↓↓ — strong negative (r ≤ −0.7, pairs move in opposite directions).