This forex correlation matrix shows how eight major dollar instruments have moved together over years of daily data. Each cell holds a long-run average correlation of daily returns, rounded to two decimals. A value near +1 means two pairs usually close in the same direction. A value near -1 means they usually close in opposite directions. Values near zero mean no steady link. The grid covers all 28 unique relationships between these eight symbols. Use it to spot doubled-up risk before you place a trade, not to predict the next candle.
Long-run averages from historical daily returns. Reviewed quarterly — last review July 2026. Correlations drift; check a live tool before relying on any single value.
| EURUSD | GBPUSD | USDJPY | USDCHF | AUDUSD | USDCAD | NZDUSD | XAUUSD | |
|---|---|---|---|---|---|---|---|---|
| EURUSD | 1.00 | +0.85 | -0.25 | -0.90 | +0.60 | -0.55 | +0.55 | +0.40 |
| GBPUSD | +0.85 | 1.00 | -0.20 | -0.80 | +0.60 | -0.55 | +0.55 | +0.35 |
| USDJPY | -0.25 | -0.20 | 1.00 | +0.45 | +0.10 | +0.25 | +0.10 | -0.40 |
| USDCHF | -0.90 | -0.80 | +0.45 | 1.00 | -0.50 | +0.50 | -0.45 | -0.50 |
| AUDUSD | +0.60 | +0.60 | +0.10 | -0.50 | 1.00 | -0.70 | +0.85 | +0.55 |
| USDCAD | -0.55 | -0.55 | +0.25 | +0.50 | -0.70 | 1.00 | -0.65 | -0.45 |
| NZDUSD | +0.55 | +0.55 | +0.10 | -0.45 | +0.85 | -0.65 | 1.00 | +0.50 |
| XAUUSD | +0.40 | +0.35 | -0.40 | -0.50 | +0.55 | -0.45 | +0.50 | 1.00 |
Green cells mark strong positive relationships, at +0.70 or higher. Red cells mark strong negative relationships, at -0.70 or lower. Gray cells are moderate. Plain cells are weak, with long-run values between -0.30 and +0.30. The weak cells matter as much as the strong ones. They show where real diversification lives. The values are a rounded consensus of published long-window tables from Mataf, Myfxbook and broker education pages, not the output of any single feed.
The pair relationships that matter most
A handful of relationships in this table drive most of the correlated risk retail traders carry. Here are the ones worth memorizing, and the reasons they hold.
| Relationship | Long-run value | Why it holds |
|---|---|---|
| EURUSD and GBPUSD | +0.85 | Two large European economies priced against the same dollar. Dollar flows move both at once. |
| EURUSD and USDCHF | -0.90 | The franc shadows the euro, but USD sits on the opposite side of the quote. The pairs mirror each other. |
| AUDUSD and NZDUSD | +0.85 | Neighboring commodity exporters with linked trade flows and similar rate cycles. |
| AUDUSD and USDCAD | -0.70 | Both track commodities against the dollar, but CAD is the quote currency's inverse side. |
| XAUUSD and AUDUSD | +0.55 | Australia is a top gold producer, and both quotes weaken when the dollar strengthens. |
| XAUUSD and USDCHF | -0.50 | Gold rises when the dollar weakens; USDCHF rises when the dollar strengthens. |
| EURUSD and USDJPY | -0.25 | Mildly negative on average, but unstable. Yield cycles flip this one for years at a time. |
Three blocs explain almost everything here. The European bloc ties EURUSD, GBPUSD and USDCHF together through the shared EUR leg and the franc's inverse-USD quoting. The commodity bloc links AUDUSD, NZDUSD and USDCAD through raw-material exports and risk appetite. Gold behaves as an anti-dollar asset, so it leans with AUDUSD and against USDCHF. The yen stands apart, driven more by rate differentials than by either bloc.
Beyond the grid: the crosses
The eight symbols above generate 28 unique pairings, and the same logic extends to every cross. When two dollar pairs correlate near +0.85, their cross moves slowly. EURGBP ranges for months because its two legs cancel. AUDNZD behaves the same way. EURCHF barely breathes outside policy shocks because the -0.90 mirror leaves little residual. The reverse also holds. Crosses built from weakly linked legs, like GBPJPY or EURAUD, inherit two independent risk sources and move fast. Read any cross as the difference between its two dollar legs in this table.
How to read a forex correlation matrix
The correlation coefficient measures how two return series move together. It runs from +1 to -1.
- +1 — the two pairs close in the same direction every day.
- -1 — they close in opposite directions every day.
- 0 — knowing one tells you nothing about the other.
These values come from daily returns, meaning day-over-day percentage changes, not from raw price levels.
A reading of +0.85 does not mean move-for-move on every candle. It means the daily closes agree far more often than not, over a long sample. On any single day the two pairs can diverge sharply. On lower timeframes the link is looser still. Intraday correlation is noisier than daily, and a five-minute chart can show the two pairs fighting each other for hours.
Correlation also says nothing about size. GBPUSD often travels further than EURUSD in a session even while agreeing on direction. Correlation measures agreement, not magnitude. If you need magnitude, compare average daily ranges separately. And remember the quote direction: USDCHF and USDCAD carry USD as the base, so their signs flip against pairs where USD is the quote. That is a quoting artifact, not an economic mystery.
Forex correlation and position risk
Here is the core lesson of this page. Long EURUSD plus long GBPUSD is not two positions. It is one oversized EUR-bloc position wearing two tickets. At +0.85, a dollar rally hits both trades on the same day. Both stops can trigger together, and a planned 1% loss quietly becomes 2%.
The fix is simple: count correlated exposure as one position when you size. If your rule is 1% risk per idea, then EURUSD and GBPUSD together get 1%, split between them, not 1% each. Run the numbers in the position size calculator before entry, and stress the downside with the drawdown calculator. Correlated losing streaks are the fastest route to deep drawdowns, and the risk of ruin calculator shows how sharply repeated doubled losses raise the danger.
The same trap hides in short baskets. Short EURUSD, short GBPUSD and long USDCHF is one large dollar-long bet, three spreads wide. Check every open trade against this table before adding the next one.
Download the complete indicator database
Manage correlated positions with tested tools on your chart. One email unlocks the full library of 1,380+ indicators with compiled MT4 and MT5 files, plus my TradingView scripts. No paywall, no spam, unsubscribe any time.
Get free access to my indicator database
One email unlocks 1,380+ free MT4, MT5 and TradingView indicators — the complete library. No single-tool download; you get the whole database.
The hedging illusion
Many traders discover the -0.90 cell and decide to hedge. Long EURUSD, short GBPUSD, opposite exposure, safety. Except that is not a hedge. It is a synthetic EURGBP position paying two spreads instead of one. You have not removed risk; you have swapped dollar risk for euro-versus-pound risk, at double the transaction cost.
True offsetting positions mostly pay costs twice while earning nothing. If the two legs cancel perfectly, your net exposure is zero and you are paying spread and swap on both sides for a flat line. If they cancel imperfectly, you hold a cross position you never analyzed. Either way, the honest question is: what exposure do I actually want? If the answer is EURGBP, trade EURGBP directly for one spread.
Negative correlation still has one legitimate defensive use. If you already hold a position you cannot exit, a negatively correlated pair can soften a shock temporarily. Treat that as an emergency brake, not a strategy.
Why correlations drift
Every number in this table is an average across regimes, and regimes change. Rate cycles are the biggest driver. When central banks diverge, pairs that shared a dollar story start trading their own yield story. USDJPY spent years mildly negative against EURUSD, then tracked US yields so tightly in 2022 that old relationships inverted for months. The interest rate tracker shows where each bank sits in its cycle right now.
Risk sentiment is the second driver. In calm markets, each pair trades its own fundamentals and correlations loosen. In stress, everything collapses into one risk-on or risk-off trade and correlations spike toward the extremes. Commodity shocks are the third. An oil squeeze can pull USDCAD away from its bloc, and a gold run can tighten the XAUUSD link to AUDUSD well beyond its average.
Policy can break a relationship outright. The Swiss National Bank's franc cap, and its sudden removal in January 2015, made EURUSD versus USDCHF readings meaningless for stretches. This is exactly why the page is labeled long-run and reviewed quarterly. For today's reading, use a live chart-based tool, and watch the forex heatmap to see which currencies are actually driving the session.
What this table cannot tell you
Honest limits, stated plainly. First, an average hides the swings that produced it. A long-run +0.55 between gold and AUDUSD contains years near +0.8 and stretches near zero. Second, these are daily-return figures. Your five-minute scalp lives in a different statistical world, and this table says little about it. Third, correlation is not causation. EURUSD does not move GBPUSD; shared dollar flows move both.
Fourth, every published table depends on its sample window. A one-year window and a ten-year window disagree, sometimes by a lot. That is why conflicting published values were rounded to a consensus here rather than quoted to false precision. Treat every cell as a tendency with error bars, not a constant. The correlation-aware dashboards in the best MT4 indicators guide can put live readings next to these baselines, and everything in the library is tested per the Editorial and Testing Policy.
Want the full toolkit? Get the complete database.
FAQ
Are these correlation values live?
No. They are long-run averages of daily returns, compiled from published multi-year tables and reviewed quarterly. A live correlation tool recalculates over recent windows and will often disagree with this page. Use this table for structural risk decisions and a live tool for current conditions.
What timeframe and data are the values based on?
Daily closing returns over multi-year windows, cross-checked across public sources such as Mataf, Myfxbook and broker education pages. Where sources conflicted, the value shown is a rounded consensus. Weekly correlations run similar; intraday correlations run noticeably weaker.
Does a strong correlation ever break?
Yes, and sometimes violently. The SNB's removal of the franc cap in 2015 shattered the EURUSD and USDCHF mirror for a period. USDJPY decoupled from the majors during the 2022 yield surge. Brexit headlines repeatedly split GBPUSD from EURUSD. Strong long-run links resume eventually, but never on a schedule.
Why do published correlation tables disagree?
Because each one uses a different sample window and timeframe. A table built on one year of daily data will not match a table built on ten years, and neither will match a weekly-return table. Both can be correct for their window. This page rounds conflicting published figures to a consensus instead of quoting one source to false precision.
Which pairs should I avoid trading together?
Avoid stacking same-direction trades inside one bloc: EURUSD with GBPUSD, or AUDUSD with NZDUSD, doubles your exposure. Opposite-sign doubles count too, like long EURUSD with long USDCHF pointing the same economic way. If you take both anyway, size the combination as a single position. Results are not guaranteed; past performance is not indicative of future results.
Related tools: trading journal, bank holidays 2026 and atr position size calculator, plus the full free forex tools directory.
External references