Currency Correlations: How Pairs Move Together and Diverge
How Do Currency Correlations Drive Forex Trading Decisions?
Currency correlations measure the degree to which two currency pairs move together or in opposite directions. A correlation of +1.0 means two pairs move in lockstep; a correlation of -1.0 means they move in opposite directions; a correlation of 0.0 means movements are independent. Understanding currency correlations is essential for forex traders seeking to diversify portfolios, manage position risk, and identify mispricings. A trader holding long positions in both EUR/USD and GBP/USD might assume they are taking different bets, but if the correlation between these pairs is +0.90, the positions are nearly redundant—both are essentially long the euro and pound against the dollar, creating concentrated risk.
Currency correlations are not static. They shift with market regimes, central bank policy changes, and geopolitical events. During risk-off periods, correlations tend to increase as investors simultaneously sell risky assets. During risk-on periods, correlations often decrease as individual country-specific factors exert stronger influence. Sophisticated traders exploit correlation instability by identifying pairs that are currently highly correlated but likely to decorrelate, allowing them to construct hedges or capture statistical reversions.
Quick definition: Currency correlation is a statistical measure of how closely two currency pairs move together, expressed as a coefficient ranging from -1.0 (perfectly inverse) to +1.0 (perfectly aligned), with values shifting based on market conditions.
Key Takeaways
- EUR/USD and GBP/USD are highly correlated (typically +0.80 to +0.95) because both pairs express bets on dollar strength, creating portfolio concentration risk.
- Commodity currencies (AUD/USD, CAD/USD) are highly correlated with equity indices: when stocks rise, commodity prices and commodity-exporting currencies typically rise together.
- Safe-haven pairs like USD/JPY and USD/CHF often show negative or low correlations with risk-on pairs, offering portfolio diversification benefits.
- Correlation matrices calculated over different time periods (1-month, 3-month, 12-month) reveal that historical correlations are poor predictors of future correlations.
- Traders use correlation breakdowns to identify early warnings of regime shifts (e.g., when historically correlated pairs suddenly diverge, a volatility or liquidity crisis may be imminent).
The Mathematics of Correlation
Currency correlations are computed using the Pearson correlation coefficient, a statistical measure of linear relationship between two variables. The formula is:
Correlation = Cov(X, Y) / (StdDev(X) * StdDev(Y))
Where Cov(X, Y) is the covariance between two returns, and StdDev(X) and StdDev(Y) are the standard deviations of each pair's returns.
In practical terms, if EUR/USD has moved up 2% and GBP/USD has moved up 1.8% on the same day, the correlation is positive. If EUR/USD has moved up 2% and USD/JPY has moved up 1.8% (meaning the yen weakened), the correlation is also positive because both moves benefit euro and pound longs. However, if EUR/USD moved up 2% while USD/CHF fell 1.5%, the correlation would be negative because franc strength (USD/CHF decline) typically occurs when safe-haven demand rises.
Correlation coefficients are typically calculated over rolling windows (1-month, 3-month, 6-month, 12-month). A 1-month rolling correlation reveals short-term pair relationships, while a 12-month correlation captures longer-term trends. Professional traders monitor multiple timeframes because short-term correlations can diverge significantly from longer-term correlations. A pair might show +0.50 correlation over one month but +0.85 correlation over 12 months.
Common Currency Pair Correlations
EUR/USD and GBP/USD correlation: typically +0.85 to +0.95
The euro and British pound are both European currencies, and both pairs express directional bets on dollar strength. When the dollar weakens broadly, both EUR/USD and GBP/USD typically rise together. When the dollar strengthens, both pairs typically fall together. This high correlation means a trader who is long both EUR/USD and GBP/USD is not diversifying; rather, they are doubling down on dollar weakness. If the trader's thesis is that the dollar will weaken, this concentrated position might be intentional. But if the trader believed they were making independent bets, they would be taking unintended risk.
The correlation between EUR/USD and GBP/USD does occasionally break down. During Brexit discussions (2016-2020), GBP/USD sometimes diverged from EUR/USD because Brexit-specific uncertainty affected the pound independently of the euro. Similarly, during periods of U.K. monetary policy divergence from ECB policy (e.g., the Bank of England tightening while the ECB eased), the pairs could diverge. However, these periods are exceptional; the baseline 12-month correlation remains very high.
USD/JPY and USD/CHF correlation: typically +0.40 to +0.70
Both the yen and franc are safe-haven currencies, and both USD/JPY and USD/CHF reflect dollar weakness during risk-off periods. However, the correlation between these pairs is moderate rather than high because the Bank of Japan and Swiss National Bank sometimes pursue different policies. When the BoJ eases aggressively while the SNB tightens, USD/JPY might weaken (as yen strength lags) while USD/CHF strengthens (as franc appreciates). Additionally, the yen is more sensitive to Japanese equity market performance (the correlation between USD/JPY and the Nikkei is strongly negative), while the franc is more sensitive to geopolitical risk.
AUD/USD and equity indices correlation: typically +0.50 to +0.75
The Australian dollar is a classic "commodity currency" because Australia is a major exporter of iron ore, coal, copper, and agricultural products. When global equities rise, investors become more optimistic about global economic growth, driving up demand for commodities and commodity-exporting currencies. Conversely, when equities fall and risk appetite declines, the Australian dollar falls. This correlation means AUD/USD can serve as an early warning signal: a sharp decline in AUD/USD often precedes broader equity weakness by days or weeks.
The correlation between AUD/USD and equities is not constant. During periods dominated by risk sentiment (2008, 2020), the correlation is strong and positive. During periods dominated by commodity supply shocks (e.g., unexpected droughts reducing agricultural output, or mining strikes reducing metal supply), the correlation can weaken because commodity prices and equities move independently. Traders who assume AUD/USD always correlates with equities will be surprised during commodity-specific supply shocks.
USD/CAD and oil prices correlation: typically +0.60 to +0.80
Canada is a major oil exporter, making the Canadian dollar sensitive to crude oil prices. When oil prices rise, Canadian export revenues and foreign currency inflows increase, typically strengthening the loonie (Canadian dollar). The correlation between USD/CAD and crude oil prices (expressed in dollars) is negative: when oil rises, the pair falls (the loonie strengthens). When oil falls, the pair rises (the loonie weakens). This relationship is robust over long periods and provides traders with a clear trading signal: if oil is rising but USD/CAD is not falling, it could signal a divergence warranting investigation.
Why Correlations Change: Market Regimes
Currency correlations shift as the dominant market driver changes. During risk-off periods (financial crises, geopolitical shocks), correlations among risky currency pairs typically increase. This is because the common driver—fear, falling risk appetite—overwhelms individual country-specific factors. In March 2020 during the COVID crash, virtually all emerging market currencies fell simultaneously against the dollar, creating correlations near +1.0. Individual Brazilian, Mexican, and South African economic conditions became irrelevant; all three countries were hit by the same panic.
Conversely, during periods when monetary policy divergence is the dominant driver, correlations may weaken. If the Federal Reserve is tightening while the ECB is easing, USD/EUR may move differently than historical correlations would suggest. The euro weakens not because of broad risk-off sentiment but because of interest rate differentials. In this regime, traders can exploit carry trades (borrowing euros and investing in higher-yielding dollar assets), creating distinct EUR/USD behavior.
Correlations also shift when commodity prices are the dominant driver. If copper prices collapse due to Chinese economic slowdown, the Australian dollar weakens, the Canadian dollar weakens, and the Chilean peso weakens—a coordinated move. However, if these countries' central banks respond differently (the Reserve Bank of Australia eases sharply while the Bank of Canada only modestly eases), their currencies may diverge. The initial correlation spike (all commodity currencies falling together) is followed by correlation breakdown as policy divergence takes effect.
Trading Applications of Correlation Analysis
Pairs trading: A pairs trader identifies two historically correlated pairs that have recently diverged, then bets that the correlation will revert to historical norms. For example, if EUR/USD and GBP/USD typically correlate at +0.90 but have diverged (EUR/USD rising while GBP/USD falls) due to a temporary shock, the trader might buy GBP/USD and short EUR/USD, betting they will reconverge. If the pairs revert toward historical correlation, the trader profits.
This strategy requires identifying when a correlation breakdown is temporary versus permanent. If GBP/USD is diverging because new information about UK growth has emerged, the breakdown may be permanent. But if GBP/USD is diverging due to temporary liquidity issues or a data release, convergence is likely. Experienced traders develop intuition for distinguishing signal from noise.
Portfolio hedging: A portfolio manager holding a long position in EUR/USD might want to hedge this exposure. If the manager buys a short position in GBP/USD, they have not hedged effectively because the correlation is +0.90—both positions are long euros/pound against the dollar. A better hedge would be to short USD/JPY, which has lower correlation with EUR/USD and provides true diversification. Alternatively, the manager could simply buy dollar-denominated assets, providing the most direct hedge.
Volatility prediction: Currency correlations are useful for predicting volatility. When correlations among currency pairs are high (near +0.80 to +1.0), portfolio volatility is elevated because moves in individual pairs are not offsetting. When correlations are low or negative, portfolio volatility is reduced because some pairs gain while others lose. Traders monitoring correlation levels can anticipate volatility shifts and adjust position sizing accordingly.
Liquidity assessment: During liquidity crises, correlations among currency pairs spike toward +1.0 because traders are forced to liquidate all positions indiscriminately. Conversely, during normal market conditions with healthy liquidity, correlations are moderate. A sudden spike in correlation across many pairs can signal emerging liquidity stress, warning traders to reduce leverage and tighten stop losses.
Real-World Examples of Correlation Breakdown
The 2016 U.S. presidential election: Before and after the election, correlations among currency pairs shifted dramatically. During the pre-election period, risk uncertainty spiked and correlations increased. After the election result was announced, correlations decreased as investors re-assessed individual currency and equity markets. The dollar strengthened broadly against developed-market currencies (USD/EUR, USD/GBP, USD/JPY all appreciated), but by different amounts, creating correlation breakdown.
The March 2020 COVID crash: During the initial COVID panic, correlations among virtually all currency pairs approached +1.0. Safe-haven pairs (USD/JPY, USD/CHF) and risk-on pairs (AUD/USD, NZD/USD) temporarily moved in sync as the common factor—extreme panic—dominated. However, within weeks, correlations normalized as central banks deployed support and investors distinguished between countries with different fiscal capacity and pandemic severity.
The 2022-2023 Fed hiking cycle: As the Federal Reserve raised rates aggressively while the ECB and BoJ lagged, correlations between USD/EUR and other dollar-strength pairs decreased. This is because the dollar's strength was driven by U.S.-specific monetary policy divergence, not broad risk sentiment. EUR/GBP briefly became more strongly correlated than EUR/USD because European policy divergence (ECB tightening but slower than Fed) was a key driver.
The June 2016 Brexit referendum: The pound crashed immediately after the vote, losing 10% against the dollar in days. However, EUR/GBP also moved sharply—the euro strengthened against the pound. This created a temporary breakdown in the usually-high correlation between EUR/USD and GBP/USD because pound-specific news (Brexit) overwhelmed the common dollar-strength factor. Traders who relied on the historical +0.90 correlation were caught off-guard.
Computing and Monitoring Correlations
Professional traders use correlation matrices—tables showing pairwise correlations among multiple currency pairs—to monitor their portfolios. A typical matrix for major pairs looks like this:
| Pair | EUR/USD | GBP/USD | USD/JPY | AUD/USD | USD/CAD |
|---|---|---|---|---|---|
| EUR/USD | 1.00 | 0.90 | -0.45 | 0.55 | -0.30 |
| GBP/USD | 0.90 | 1.00 | -0.40 | 0.58 | -0.25 |
| USD/JPY | -0.45 | -0.40 | 1.00 | -0.75 | 0.35 |
| AUD/USD | 0.55 | 0.58 | -0.75 | 1.00 | 0.20 |
| USD/CAD | -0.30 | -0.25 | 0.35 | 0.20 | 1.00 |
This matrix reveals several insights. EUR/USD and GBP/USD are highly correlated (+0.90), meaning a diversified portfolio should not hold both without offsetting positions. USD/JPY is negatively correlated with risk-on pairs (EUR/USD, GBP/USD, AUD/USD), confirming the yen's safe-haven status. USD/CAD correlations are weak and mixed, suggesting Canadian dollar moves are driven by commodity and policy factors distinct from broader dollar strength.
Traders update correlation matrices monthly using the previous 12 months of daily returns. This rolling approach captures both structural correlations and temporary regime shifts. When a correlation deviates significantly from its historical average, it signals opportunity or risk.
Common Mistakes When Using Correlations
Assuming historical correlations predict future correlations. The correlation between EUR/USD and GBP/USD has been +0.90 for a decade, but this does not guarantee it will remain +0.90. During political shocks affecting one country more than another, correlations can drop to +0.70 or lower. Traders who mechanically assume correlations are constant will be surprised.
Confusing correlation with causation. AUD/USD and equity indices are correlated, but this does not mean equities cause the Australian dollar to move. Rather, both are driven by a common factor: risk sentiment. Assuming causation might lead a trader to short AUD/USD when equities are strong, expecting the Australian dollar to weaken in lockstep. But if commodity prices spike (strengthening AUD despite equities), the trade fails.
Ignoring regime changes. Correlations are regime-dependent. A pair that shows +0.40 correlation over 12 months might show +0.80 correlation over the most recent month if market regime has shifted. Traders should monitor both long-term and short-term correlations and distinguish between them.
Holding multiple highly-correlated positions without realizing it. A trader might believe they are diversified by holding EUR/USD long, GBP/USD long, and AUD/USD long. But if all three pairs are moving at +0.85 correlation, the portfolio is concentrated on dollar weakness, not diversified. A single adverse move in dollar strength could wipe out all three positions simultaneously.
Using correlation to predict price movements. High correlation does not imply future price moves; it merely describes how pairs have historically moved together. A trader who observes that EUR/USD and GBP/USD have +0.95 correlation and then shorts GBP/USD when EUR/USD rises is making a logical error. Both pairs could rise together (if the dollar weakens further), eliminating any profit.
FAQ
How is correlation different from cointegration?
Correlation measures linear statistical relationship between returns of two pairs. Cointegration measures whether two pairs move together in the long term, even if they diverge in the short term. A correlation of zero does not imply two pairs are unrelated; they could be cointegrated, meaning they drift back together over time. Cointegration is more useful for pairs trading because it identifies pairs that will revert to historical relationships.
What is the relationship between correlation and volatility?
Generally, when correlations among risky assets increase, portfolio volatility increases (assuming the portfolio is long multiple correlated pairs). Conversely, when correlations decrease, diversification improves and portfolio volatility falls. However, correlation and volatility are independent phenomena: a pair can show low correlation with others while displaying high individual volatility.
Can correlations be negative?
Yes. Negative correlations occur when two pairs move in opposite directions. USD/JPY and AUD/USD often show negative correlation: when the yen appreciates (USD/JPY falls), the Australian dollar typically depreciates (AUD/USD falls) because both reflect safe-haven demand during risk-off periods. A perfectly negative correlation (-1.0) is rare in forex, but negative correlations of -0.40 to -0.60 are common.
How do traders use correlation breakdowns as trading signals?
When two pairs that are historically highly correlated (e.g., EUR/USD and GBP/USD at +0.90) suddenly diverge, it signals new information or regime shift. A trader might interpret the breakdown as an opportunity: if the divergence is temporary, the trader can bet on reversion. Alternatively, the breakdown might signal an emerging risk factor (political upheaval in one country, or monetary policy divergence) that warrants caution.
Is correlation the same across all timeframes?
No. A pair might show +0.40 correlation on a 1-day basis but +0.85 correlation on a 12-month basis. This occurs because short-term noise and temporary shocks cause daily divergences, while longer-term structural factors (shared economic drivers) reassert themselves over months. Traders should specify which timeframe they are referencing when discussing correlation.
Why is correlation important for position sizing?
Correlation determines portfolio diversification benefits. If a trader holds two uncorrelated positions, position sizing can be larger (e.g., 5% per position) because the portfolio is truly diversified. If two positions are highly correlated, position sizing should be smaller (e.g., 2% per position) because they represent concentrated risk. Professional risk managers adjust position limits based on correlation matrices.
Related Concepts
Summary
Currency correlations measure how closely two forex pairs move together, with coefficients ranging from -1.0 to +1.0. Understanding correlations is essential for portfolio construction, risk management, and identifying trading opportunities. EUR/USD and GBP/USD are highly correlated (+0.85-0.95), while safe-haven pairs like USD/JPY show negative or low correlations with risk-on pairs, providing true diversification. Correlations are not static; they shift with market regimes, central bank policy, and geopolitical events. Traders exploit correlation breakdowns through pairs trading and monitor correlation matrices to ensure their portfolios achieve intended diversification. A critical mistake is assuming historical correlations will persist; correlations are regime-dependent and can shift dramatically during market crises.