Currency pairs do not move in isolation. Because every forex transaction involves two currencies, and many pairs share a common currency (such as the US dollar), the movements of different pairs are often mathematically related. Understanding these relationships, measured through correlation, helps traders avoid hidden risk concentration, identify confirmation signals, and build more diversified portfolios.
John Murphy's Intermarket Technical Analysis established the broader principle that all financial markets are connected. Within forex, these connections are particularly direct: when you understand how EUR/USD and GBP/USD tend to move together, or how EUR/USD and USD/CHF tend to move in opposite directions, you add a layer of analytical context that pure chart analysis alone cannot provide.
The Pearson Correlation Coefficient
The formula involves calculating the covariance of the two price series divided by the product of their individual standard deviations. Most trading platforms and analytical tools calculate this automatically, you do not need to compute it manually. What matters is understanding how to interpret the output.
Interpretation scale:
| Coefficient Range | Interpretation |
|---|---|
| +0.80 to +1.00 | Strong positive correlation, pairs move together closely |
| +0.60 to +0.79 | Moderate positive correlation, generally move together |
| +0.40 to +0.59 | Weak positive correlation, some tendency to move together |
| -0.39 to +0.39 | Negligible or no meaningful correlation |
| -0.59 to -0.40 | Weak negative correlation, some tendency to move oppositely |
| -0.79 to -0.60 | Moderate negative correlation, generally move opposite |
| -1.00 to -0.80 | Strong negative (inverse) correlation, pairs move opposite |
Reading a Correlation Matrix
A correlation matrix displays the correlation coefficients between multiple currency pairs in a grid format. Each cell shows the correlation between the pair listed in its row and the pair listed in its column.
Most forex correlation tools present this as a color-coded table. Strong positive correlations are shown in green (or dark shading), strong negative correlations in red, and weak correlations in neutral tones. The diagonal of the matrix is always +1.00, every pair is perfectly correlated with itself.
Correlation matrices are available on platforms like Myfxbook, Mataf, and OANDA's correlation tool. TradingView also offers correlation indicators that overlay the relationship between two instruments on a chart.
When viewing a matrix, remember that correlations are calculated over a specific lookback period. A matrix using a 20-day window will show different values than one using a 120-day window for the same pairs. Always check which lookback period the tool is using, and consider viewing multiple windows for a more complete picture.
Major Pair Correlations
Why do these correlations exist?
The primary driver is the shared currency. EUR/USD and GBP/USD both rise when the US dollar weakens, hence their positive correlation. EUR/USD rises when the dollar weakens, while USD/CHF falls when the dollar weakens (because USD is the base currency in USD/CHF), hence their strong inverse correlation.
Secondary drivers include economic similarity. Australia and New Zealand are both commodity-exporting economies in the Asia-Pacific region with similar trade relationships and monetary policy cycles, which explains the persistently high positive correlation between AUD/USD and NZD/USD.
Rolling Correlation Windows
Correlation is not static. The relationship between two pairs can shift significantly over weeks and months, driven by changes in monetary policy, economic divergences, geopolitical events, or shifts in risk sentiment.
Common rolling periods and their use:
| Window | Purpose |
|---|---|
| 20-day | Short-term trading correlations, day trading reference |
| 60-day | Medium-term swing trading reference |
| 120-day | Longer-term portfolio and position trading reference |
| 250-day | Annual correlation, used for strategic allocation |
A practical approach is to monitor both a short window (20-day) and a longer window (120-day). When the short-term correlation diverges significantly from the long-term norm, it signals a potential temporary dislocation, either a reversion opportunity or an early signal of a structural shift.
Correlation Breakdown During Crisis
One of the most important (and dangerous) characteristics of correlation is that it can break down precisely when you need it most, during market crises and extreme volatility events.
During periods of extreme risk aversion (financial crises, geopolitical shocks, pandemics), correlations across many asset classes tend to converge toward +1.0 as traders simultaneously flee to safe-haven assets. Currency pairs that are normally uncorrelated may suddenly move in lockstep as broad "risk-off" or "risk-on" flows dominate individual pair dynamics.
For example, during the 2008 financial crisis and the initial COVID-19 shock in March 2020, the US dollar strengthened broadly against nearly all currencies simultaneously, causing correlations between USD-denominated pairs to shift dramatically.
This phenomenon means that relying on historical correlation for risk management requires caution: the diversification benefit you expect from holding uncorrelated pairs may evaporate during the precise market conditions when you need it most.
Practical Applications
Avoiding Hidden Risk Doubling
If you take a long position on EUR/USD and simultaneously take a long position on GBP/USD, you have not diversified, you have roughly doubled your exposure to a weaker US dollar. Because these pairs are strongly positively correlated, both positions will likely profit or lose together.
Similarly, going long EUR/USD and short USD/CHF is essentially the same directional bet expressed twice, because these pairs are strongly inversely correlated.
Before entering multiple trades, check the correlation matrix to ensure you are not inadvertently concentrating your risk.
Trade Confirmation
If your analysis produces a buy signal on EUR/USD, checking whether GBP/USD is also showing bullish signals (given their positive correlation) can provide additional confidence. If EUR/USD looks bullish but GBP/USD looks decisively bearish, the conflicting signal warrants caution.
Portfolio Diversification
To achieve genuine diversification, seek pairs with low or zero correlation. For example, EUR/USD and USD/JPY often have weak or variable correlation, meaning they respond to different fundamental drivers. Holding positions in weakly correlated pairs reduces the portfolio's overall volatility compared to holding positions in strongly correlated pairs.
Correlation Divergence as a Signal
When two normally highly correlated pairs suddenly diverge, one rallying while the other falls, it can signal an opportunity. Either one pair has moved too far and will revert, or a fundamental shift is occurring that warrants investigation. For example, if EUR/USD is rising strongly but GBP/USD is falling, it suggests EUR-specific strength or GBP-specific weakness rather than broad USD movement.
Cross-Asset Correlations
Currency correlations extend beyond forex pairs to other asset classes. Understanding these intermarket relationships adds another dimension to your analysis:
USD and Gold (XAU/USD): Gold is traditionally inversely correlated with the US dollar. When the dollar weakens, gold tends to rise, and vice versa. This relationship has been consistent over long periods, though it can temporarily decouple during extreme market stress when both assets are sought as safe havens.
Commodity currencies and commodity prices: The Australian dollar (AUD) is correlated with iron ore and copper prices. The Canadian dollar (CAD) is correlated with crude oil prices. The New Zealand dollar (NZD) is correlated with dairy prices. When the underlying commodity rises, the associated currency tends to strengthen because commodity exports boost the country's trade balance and economic outlook.
USD/JPY and US equity indices: USD/JPY often shows a positive correlation with the S&P 500 and other US equity indices. When risk appetite is high (equities rising), traders tend to sell the safe-haven yen, pushing USD/JPY higher. When equities fall (risk aversion), traders buy yen, pushing USD/JPY lower.
These cross-asset correlations are not as stable as same-market pair correlations, but they provide valuable context. If oil prices are surging and your analysis shows a buy signal on CAD, the commodity backdrop supports the trade. If oil is collapsing while your technicals are bullish on CAD, the fundamental headwind is a warning worth heeding.
Common Mistakes with Correlation Analysis
Treating correlation as causation. Two pairs being correlated does not mean one causes the other to move. They may be responding to a common third factor (such as USD strength). Misinterpreting correlation as causation leads to flawed analysis.
Using a single lookback period. A 20-day correlation snapshot can be misleading if the longer-term relationship is different. Always compare multiple time windows.
Ignoring correlation changes. Traders who check correlation once and never revisit it may be blindsided when relationships shift. Build a habit of reviewing your correlation matrix at least weekly.
Over-relying on correlation for hedging. Some traders attempt to "hedge" a EUR/USD long by going short on GBP/USD (since they are positively correlated, this would be a partial offset). However, because the correlation is imperfect, this hedge can fail, both positions may lose simultaneously if GBP and EUR diverge.
Key Takeaways
- The Pearson correlation coefficient measures the linear relationship between two currency pairs on a scale from -1.0 (perfect inverse) to +1.0 (perfect positive).
- Strong correlations (above +0.7 or below -0.7) mean pairs tend to move together or inversely. Weak correlations (between -0.4 and +0.4) mean little systematic relationship.
- Common correlations: EUR/USD and GBP/USD are positively correlated; EUR/USD and USD/CHF are inversely correlated; AUD/USD and NZD/USD are positively correlated.
- Rolling correlation windows (20, 60, 120 days) are essential because correlation is not static, it shifts over time.
- Correlation breaks down during crises. Risk-off events can cause normally uncorrelated pairs to move in lockstep, reducing the diversification benefit precisely when it is needed most.
- Check correlations before entering multiple trades to avoid hidden risk concentration. Two strongly correlated positions are effectively one doubled position.
- Use correlation divergence as a signal, when normally correlated pairs move in opposite directions, it reveals pair-specific rather than broad market forces.
This lesson is for educational purposes only. It does not constitute financial advice. Trading forex involves significant risk of loss and is not suitable for all investors.