Forex correlation: the currency correlation table, daily and weekly
You buy EURUSD in one trade and GBPUSD in another, and you think you spread the risk. Then one morning the dollar turns and you lose on both at once. This table tells you, on the daily and the weekly, which pairs move together and which move the opposite way, so you don't double your risk without knowing it.
By JH Grandgerard · Data from October 28, 2026
Correlated currency pairs, daily and weekly
Among the 28 main pairs, on the daily and over 3 months, the ones that move most together are EURJPY and GBPJPY (0.95) and the ones that move most in opposite directions, NZDUSD and GBPNZD (−0.81). On the weekly and over 1 year, the most together are EURJPY and GBPJPY (0.88) and the most opposite, NZDUSD and EURNZD (−0.84).
Daily: ECB fixings up to October 28, 2026. Weekly: Friday fixings up to October 23, 2026.
| Pair | EURUSD | GBPUSD | AUDUSD | NZDUSD | USDCAD | USDCHF | USDJPY | EURGBP | EURAUD | EURNZD | EURCAD | EURCHF | EURJPY | GBPAUD | GBPNZD | GBPCAD | GBPCHF | GBPJPY | AUDNZD | AUDCAD | AUDCHF | AUDJPY | NZDCAD | NZDCHF | NZDJPY | CADCHF | CADJPY | CHFJPY |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| EURUSD | 1 | 0.83 | 0.50 | 0.61 | −0.55 | −0.55 | −0.46 | 0.34 | 0.25 | 0.00 | 0.47 | 0.23 | 0.06 | 0.10 | −0.16 | 0.27 | 0.06 | −0.05 | −0.22 | 0.11 | −0.03 | −0.09 | 0.37 | 0.20 | 0.05 | −0.18 | −0.20 | −0.09 |
| GBPUSD | 0.83 | 1 | 0.56 | 0.68 | −0.56 | −0.53 | −0.39 | −0.25 | 0.05 | −0.22 | 0.28 | 0.10 | 0.05 | 0.17 | −0.12 | 0.44 | 0.22 | 0.12 | −0.24 | 0.17 | 0.04 | 0.01 | 0.45 | 0.30 | 0.18 | −0.14 | −0.11 | −0.02 |
| AUDUSD | 0.50 | 0.56 | 1 | 0.58 | −0.48 | −0.29 | −0.15 | −0.08 | −0.71 | −0.35 | 0.01 | 0.10 | 0.13 | −0.72 | −0.34 | 0.06 | 0.14 | 0.14 | 0.30 | 0.71 | 0.62 | 0.50 | 0.38 | 0.43 | 0.33 | 0.10 | 0.11 | 0.05 |
| NZDUSD | 0.61 | 0.68 | 0.58 | 1 | −0.66 | −0.62 | −0.34 | −0.08 | −0.16 | −0.79 | −0.06 | −0.19 | −0.03 | −0.13 | −0.81 | −0.01 | −0.14 | 0.00 | −0.60 | 0.11 | −0.01 | 0.07 | 0.79 | 0.62 | 0.48 | −0.15 | 0.01 | 0.09 |
| USDCAD | −0.55 | −0.56 | −0.48 | −0.66 | 1 | 0.62 | 0.42 | −0.01 | 0.10 | 0.41 | 0.48 | 0.25 | 0.16 | 0.11 | 0.45 | 0.50 | 0.25 | 0.15 | 0.30 | 0.27 | 0.10 | 0.07 | −0.06 | −0.19 | −0.13 | −0.17 | −0.11 | −0.01 |
| USDCHF | −0.55 | −0.53 | −0.29 | −0.62 | 0.62 | 1 | 0.43 | −0.06 | −0.13 | 0.35 | 0.08 | 0.68 | 0.17 | −0.11 | 0.40 | 0.12 | 0.70 | 0.18 | 0.44 | 0.18 | 0.58 | 0.20 | −0.31 | 0.24 | −0.08 | 0.67 | 0.12 | −0.27 |
| USDJPY | −0.46 | −0.39 | −0.15 | −0.34 | 0.42 | 0.43 | 1 | −0.15 | −0.21 | 0.07 | −0.04 | 0.10 | 0.86 | −0.15 | 0.15 | 0.05 | 0.17 | 0.87 | 0.25 | 0.18 | 0.23 | 0.78 | −0.11 | 0.02 | 0.66 | 0.14 | 0.86 | 0.75 |
| EURGBP | 0.34 | −0.25 | −0.08 | −0.08 | −0.01 | −0.06 | −0.15 | 1 | 0.36 | 0.36 | 0.34 | 0.23 | 0.03 | −0.12 | −0.09 | −0.27 | −0.27 | −0.30 | 0.02 | −0.09 | −0.11 | −0.18 | −0.12 | −0.16 | −0.20 | −0.06 | −0.16 | −0.12 |
| EURAUD | 0.25 | 0.05 | −0.71 | −0.16 | 0.10 | −0.13 | −0.21 | 0.36 | 1 | 0.39 | 0.36 | 0.07 | −0.09 | 0.89 | 0.25 | 0.15 | −0.11 | −0.20 | −0.52 | −0.71 | −0.71 | −0.63 | −0.13 | −0.32 | −0.32 | −0.25 | −0.29 | −0.13 |
| EURNZD | 0.00 | −0.22 | −0.35 | −0.79 | 0.41 | 0.35 | 0.07 | 0.36 | 0.39 | 1 | 0.43 | 0.41 | 0.08 | 0.24 | 0.90 | 0.22 | 0.22 | −0.04 | 0.58 | −0.06 | −0.01 | −0.16 | −0.72 | −0.63 | −0.56 | 0.05 | −0.16 | −0.18 |
| EURCAD | 0.47 | 0.28 | 0.01 | −0.06 | 0.48 | 0.08 | −0.04 | 0.34 | 0.36 | 0.43 | 1 | 0.50 | 0.23 | 0.22 | 0.30 | 0.81 | 0.33 | 0.11 | 0.08 | 0.40 | 0.08 | −0.03 | 0.32 | 0.01 | −0.08 | −0.36 | −0.32 | −0.10 |
| EURCHF | 0.23 | 0.10 | 0.10 | −0.19 | 0.25 | 0.68 | 0.10 | 0.23 | 0.07 | 0.41 | 0.50 | 1 | 0.25 | −0.04 | 0.33 | 0.38 | 0.87 | 0.16 | 0.32 | 0.31 | 0.65 | 0.15 | −0.04 | 0.45 | −0.05 | 0.63 | −0.03 | −0.40 |
| EURJPY | 0.06 | 0.05 | 0.13 | −0.03 | 0.16 | 0.17 | 0.86 | 0.03 | −0.09 | 0.08 | 0.23 | 0.25 | 1 | −0.11 | 0.07 | 0.22 | 0.23 | 0.95 | 0.15 | 0.26 | 0.24 | 0.83 | 0.09 | 0.13 | 0.78 | 0.06 | 0.85 | 0.79 |
| GBPAUD | 0.10 | 0.17 | −0.72 | −0.13 | 0.11 | −0.11 | −0.15 | −0.12 | 0.89 | 0.24 | 0.22 | −0.04 | −0.11 | 1 | 0.31 | 0.30 | 0.02 | −0.07 | −0.56 | −0.71 | −0.70 | −0.58 | −0.08 | −0.27 | −0.24 | −0.24 | −0.23 | −0.08 |
| GBPNZD | −0.16 | −0.12 | −0.34 | −0.81 | 0.45 | 0.40 | 0.15 | −0.09 | 0.25 | 0.90 | 0.30 | 0.33 | 0.07 | 0.31 | 1 | 0.36 | 0.37 | 0.10 | 0.61 | −0.02 | 0.04 | −0.09 | −0.71 | −0.59 | −0.51 | 0.08 | −0.09 | −0.14 |
| GBPCAD | 0.27 | 0.44 | 0.06 | −0.01 | 0.50 | 0.12 | 0.05 | −0.27 | 0.15 | 0.22 | 0.81 | 0.38 | 0.22 | 0.30 | 0.36 | 1 | 0.50 | 0.30 | 0.07 | 0.47 | 0.15 | 0.08 | 0.40 | 0.11 | 0.04 | −0.33 | −0.23 | −0.03 |
| GBPCHF | 0.06 | 0.22 | 0.14 | −0.14 | 0.25 | 0.70 | 0.17 | −0.27 | −0.11 | 0.22 | 0.33 | 0.87 | 0.23 | 0.02 | 0.37 | 0.50 | 1 | 0.31 | 0.30 | 0.35 | 0.70 | 0.24 | 0.02 | 0.53 | 0.05 | 0.65 | 0.05 | −0.33 |
| GBPJPY | −0.05 | 0.12 | 0.14 | 0.00 | 0.15 | 0.18 | 0.87 | −0.30 | −0.20 | −0.04 | 0.11 | 0.16 | 0.95 | −0.07 | 0.10 | 0.30 | 0.31 | 1 | 0.14 | 0.28 | 0.27 | 0.85 | 0.13 | 0.18 | 0.81 | 0.08 | 0.86 | 0.79 |
| AUDNZD | −0.22 | −0.24 | 0.30 | −0.60 | 0.30 | 0.44 | 0.25 | 0.02 | −0.52 | 0.58 | 0.08 | 0.32 | 0.15 | −0.56 | 0.61 | 0.07 | 0.30 | 0.14 | 1 | 0.57 | 0.62 | 0.41 | −0.55 | −0.30 | −0.24 | 0.27 | 0.11 | −0.05 |
| AUDCAD | 0.11 | 0.17 | 0.71 | 0.11 | 0.27 | 0.18 | 0.18 | −0.09 | −0.71 | −0.06 | 0.40 | 0.31 | 0.26 | −0.71 | −0.02 | 0.47 | 0.35 | 0.28 | 0.57 | 1 | 0.76 | 0.60 | 0.37 | 0.32 | 0.26 | −0.02 | 0.04 | 0.05 |
| AUDCHF | −0.03 | 0.04 | 0.62 | −0.01 | 0.10 | 0.58 | 0.23 | −0.11 | −0.71 | −0.01 | 0.08 | 0.65 | 0.24 | −0.70 | 0.04 | 0.15 | 0.70 | 0.27 | 0.62 | 0.76 | 1 | 0.59 | 0.07 | 0.57 | 0.21 | 0.63 | 0.20 | −0.18 |
| AUDJPY | −0.09 | 0.01 | 0.50 | 0.07 | 0.07 | 0.20 | 0.78 | −0.18 | −0.63 | −0.16 | −0.03 | 0.15 | 0.83 | −0.58 | −0.09 | 0.08 | 0.24 | 0.85 | 0.41 | 0.60 | 0.59 | 1 | 0.15 | 0.29 | 0.79 | 0.19 | 0.82 | 0.69 |
| NZDCAD | 0.37 | 0.45 | 0.38 | 0.79 | −0.06 | −0.31 | −0.11 | −0.12 | −0.13 | −0.72 | 0.32 | −0.04 | 0.09 | −0.08 | −0.71 | 0.40 | 0.02 | 0.13 | −0.55 | 0.37 | 0.07 | 0.15 | 1 | 0.67 | 0.53 | −0.33 | −0.08 | 0.12 |
| NZDCHF | 0.20 | 0.30 | 0.43 | 0.62 | −0.19 | 0.24 | 0.02 | −0.16 | −0.32 | −0.63 | 0.01 | 0.45 | 0.13 | −0.27 | −0.59 | 0.11 | 0.53 | 0.18 | −0.30 | 0.32 | 0.57 | 0.29 | 0.67 | 1 | 0.50 | 0.48 | 0.13 | −0.16 |
| NZDJPY | 0.05 | 0.18 | 0.33 | 0.48 | −0.13 | −0.08 | 0.66 | −0.20 | −0.32 | −0.56 | −0.08 | −0.05 | 0.78 | −0.24 | −0.51 | 0.04 | 0.05 | 0.81 | −0.24 | 0.26 | 0.21 | 0.79 | 0.53 | 0.50 | 1 | 0.02 | 0.80 | 0.77 |
| CADCHF | −0.18 | −0.14 | 0.10 | −0.15 | −0.17 | 0.67 | 0.14 | −0.06 | −0.25 | 0.05 | −0.36 | 0.63 | 0.06 | −0.24 | 0.08 | −0.33 | 0.65 | 0.08 | 0.27 | −0.02 | 0.63 | 0.19 | −0.33 | 0.48 | 0.02 | 1 | 0.25 | −0.34 |
| CADJPY | −0.20 | −0.11 | 0.11 | 0.01 | −0.11 | 0.12 | 0.86 | −0.16 | −0.29 | −0.16 | −0.32 | −0.03 | 0.85 | −0.23 | −0.09 | −0.23 | 0.05 | 0.86 | 0.11 | 0.04 | 0.20 | 0.82 | −0.08 | 0.13 | 0.80 | 0.25 | 1 | 0.83 |
| CHFJPY | −0.09 | −0.02 | 0.05 | 0.09 | −0.01 | −0.27 | 0.75 | −0.12 | −0.13 | −0.18 | −0.10 | −0.40 | 0.79 | −0.08 | −0.14 | −0.03 | −0.33 | 0.79 | −0.05 | 0.05 | −0.18 | 0.69 | 0.12 | −0.16 | 0.77 | −0.34 | 0.83 | 1 |
Each cell is the correlation between two pairs. 1 means they move the same; 0, that they have no relation; −1, that they move opposite. The name of each pair is the usual one in the market (EURUSD, not USDEUR).
- 0.80 or more: they move almost the same
- from 0.20 to 0.79: they go together in part
- near 0 (from −0.19 to 0.19): each goes its own way
- from −0.79 to −0.20: they go opposite in part
- −0.80 or less: they move almost opposite
How to read it
Near 1, buying both pairs is almost opening the same trade twice. Near −1, buying one and selling the other is too. Near 0, each goes its own way. Since correlation changes with the period, the table lets you see three periods: 1, 3 and 6 months on the daily, and 6 months, 1 year and 2 years on the weekly.
A worked example
Take EURUSD on the daily, over 3 months: 63 daily returns up to October 28, 2026. Of the other 27 main pairs, the ones that move most with it are GBPUSD (0.83), NZDUSD (0.61), AUDUSD (0.50), EURCAD (0.47), NZDCAD (0.37). The ones that move the opposite way are USDCAD (−0.55), USDCHF (−0.55), USDJPY (−0.46), AUDNZD (−0.22), CADJPY (−0.20).
With a pair that moves with EURUSD, buying both is betting twice on the same thing, and with one that moves the opposite way, buying one and selling the other is too. And look at the period, because today's correlation may not be the one from six months ago.
What to do with this
Before you open a second trade, look up the pair of that trade and see how much it moves with the pair you already have open. If the correlation is over 0.8 buying both, or under −0.8 buying one and selling the other, you have one idea with the risk of two.
The table tells you which pairs look alike, never when to get in. That's decided on the chart: in the LDP (Lenguaje del Precio) method I wait for the daily and the weekly to be on my side, and then for the trigger on the close of a candle, with a relevant Level nearby. Until the chart confirms, there's no entry, however far apart two pairs have moved. Waiting is what costs most of us, and that patience is emotional muscle: you build it by speculating, with a little more weight each week.
How we calculate it and how to check it
The price of each pair
The table comes from the European Central Bank's euro reference rates, the ECB fixing. They're set around 14:10 and published around 16:00 Frankfurt time on every business day, and they say how many units of each currency one euro buys. With them you get the price of any pair. GBPUSD, for example, is the dollars one pound buys, which is the dollars per euro divided by the pounds per euro.
The return and the correlation
For each pair we take the log return, which is ln(today's price ÷ the previous day's price). On the weekly we use the fixing of each Friday, or the last business day before it if that Friday had no fixing, and we compare each Friday with the one before. It's logarithmic because a rise and a fall of the same size cancel out and the changes can be added.
The correlation of two pairs is Pearson's, between their two columns of returns: the covariance of the two divided by the product of their standard deviations, which gives a number from −1 to 1. The periods are exact: on the daily, 21 business days (1 month), 63 (3 months) and 126 (6 months); on the weekly, 26 weeks (6 months), 52 (1 year) and 104 (2 years).
We don't offer 1 month on the weekly, because it would be 4 points, and with 4 points a correlation says nothing.
A worked example with the real numbers of EURUSD and GBPUSD on the daily, over 3 months:
- ECB fixing of October 28, 2026: 1 euro = 1.1133 dollars and 1 euro = 0.83785 pounds. The business day before, October 27, 2026: 1.115 dollars and 0.83658 pounds.
- EURUSD is the dollars per euro, so its return for the day is ln(1.1133 ÷ 1.115) = −0.1526%. GBPUSD is the dollars per pound, which is 1.1133 ÷ 0.83785 = 1.328758; the day before, 1.332807; its return is ln(1.328758 ÷ 1.332807) = −0.3043%, with all the decimals of the rates.
- The same is repeated over the 63 business days of the period: two columns of 63 returns. The averages of EURUSD and GBPUSD are −0.0494% and −0.0159%.
- Pearson, with x the returns of EURUSD and y those of GBPUSD: the sum of the products (x − mean of x) times (y − mean of y), divided by the square root of [the sum of (x − mean of x) squared times the sum of (y − mean of y) squared]. It comes out 0.8292, which rounded is 0.83, the same number as in the table.
The ECB rates we use are in the public file correlacion.json, under "serie", so you can redo any cell with a spreadsheet: the logarithm of each rate, the difference from one day to the day before, and the CORREL function over the two columns.
Worth knowing
The ECB fixing is from 14:10 in Frankfurt and it isn't your broker's close, so the figure won't match your broker's to the decimal. What matters is the sign and the strength of the relationship.
Correlation measures the past: it says how the pairs moved together, not how they will move. When the short period and the long one say different things, the relationship is usually changing, and that's information too.
Two pairs that share a currency look alike in part for that reason: EURJPY and GBPJPY both carry the yen. And a pair that joins the Hong Kong dollar with the US dollar, or the Danish krone with the euro, moves very little, because those two currencies are pegged, the first to the dollar and the second to the euro, so its correlation measures mostly noise. The Chinese yuan (CNH) uses the figure for the onshore yuan of mainland China (CNY), because the ECB doesn't publish the CNH.
And gold?
The ECB doesn't publish gold, and we haven't found an official, free source of the daily price of gold that we're allowed to reuse. That's why it isn't in this table, and we won't fill the gap with data of doubtful license.
Frequently asked questions
What is correlation in forex?
It's how much two currency pairs move together. It's measured with a number from −1 to 1: near 1 they move the same, near −1 the opposite, and near 0 each goes its own way.
What is currency correlation?
It's the same idea seen from the currencies. Two pairs correlate when they share a currency, or when the same news moves both. Here you get the number for 28 main pairs, daily and weekly, and the finder gives it for any pair among the 23 currencies.
Which forex pairs are most correlated?
Pairs that share a currency tend to look alike, because that currency moves both. Those that carry the dollar as the quote currency, like EURUSD, GBPUSD, AUDUSD and NZDUSD, tend to move together, and those that carry it as the base, like USDCHF, USDCAD and USDJPY, tend to move the opposite way to them. The table gives you today's number, and the finder gives it for any pair.
Which currency pairs are not correlated?
Look for the cells near 0 in the table, from −0.19 to 0.19: each of those pairs goes its own way. It changes with the period, so check the three periods before you count on it.
How to use correlation in forex?
Before you open a new trade, look up the pair of that trade and see how much it moves with the pairs you already have open. If the correlation is over 0.8 buying both, or under −0.8 buying one and selling the other, you're repeating the same trade. Then the risk you take is the risk of two, even if you think they're two different ideas.
How to trade correlated forex pairs?
You can build a strategy around it, but I use correlation for something simpler: to know how much I'm risking in total before I open the second trade. The entry is always decided on the chart, and on the chart you wait.
How to read a correlation matrix?
Each row and each column is a pair, and each cell is the correlation between the row pair and the column pair. The diagonal is always 1, because a pair moves exactly the same as itself. Look for the dark green cells, where they move almost the same, and the dark gold ones with a minus sign, where they move almost opposite.
How is correlation calculated?
With Pearson's correlation between the log returns of the two pairs, which are the change of each day (or each week) measured with the logarithm so the changes can be added. The calculation is explained step by step, with real numbers, in "How we calculate it and how to check it."
What does a correlation of 0.8 or −0.8 mean?
With 0.8, the two pairs move in the same direction most of the time; with −0.8, in the opposite one. In practice, buying both (with 0.8), or buying one and selling the other (with −0.8), is almost opening the same trade twice.
Which period should I use to measure forex correlation?
Look at the three periods. If the short one and the long one say the same, the relationship holds, and if they don't match, it's usually changing. That's why the table has 1, 3 and 6 months on the daily, and 6 months, 1 year and 2 years on the weekly.
Why does the correlation between pairs change?
Because what moves each currency changes with time: central banks, inflation, jobs and the news of each country. Two pairs that went together six months ago can split apart today, so you have to look at the correlation of now.
Which pairs have a negative correlation?
The ones that move in opposite directions, which show in gold with a minus sign in the table. A typical case is EURUSD and USDCHF, because in one the dollar is the quote currency and in the other the base. If you buy both, part of it cancels out; if you buy one and sell the other, both are betting on the same thing.
Which timeframe should I use to look at correlation?
For this table, the daily and the weekly, which is where we speculate calmly. On minute charts correlation moves faster, that's true, but we decide with what holds for weeks and months.
Is there a forex correlation calculator?
This page is one: you type a pair and it tells you which pairs move with it and which move the opposite way, with the period you choose. There's no sign-up and no ads, and you can see how it's calculated.
Is the correlation table updated in real time?
No. The daily is redone every business day after the ECB publishes its fixing, around 16:00 in Frankfurt, and the weekly changes every Friday. For those of us who trade the daily and the weekly, the movement of every minute adds nothing.
Can I see the correlation of a pair that isn't one of the main pairs?
Yes: type its name in the finder and tick the box for the 253 crosses of the 23 currencies, and the pairs with the Norwegian krone, the rand or the Mexican peso come in too.
MetaTrader tools
After you pick pairs that don't repeat each other, there's still the lot of each order, and doing it by hand wears you out
In MetaTrader, the Order Manager in our tools calculates that lot with your broker's data (the contract size, the minimum lot and the lot step), and for several orders at once. You decide the risk, and the Order Manager sends nothing other than what you ask for.
Where the data comes from
- Source of the prices: the European Central Bank's euro reference exchange rates, set around 14:10 and published around 16:00 Frankfurt time on every business day, for information only. They're free to use with credit to the source (ECB terms). TendenciasFX calculates the crosses and the correlations from them.
- The Chinese yuan (CNH) uses the figure for the onshore yuan of mainland China (CNY), because the ECB doesn't publish the CNH. The two move very close together.
- Open data: correlacion.json (the matrices of the 28 main pairs over the six periods and the series of ECB rates to redo any cell or to calculate the 253 crosses) and correlacion.csv (the 378 combinations of two main pairs, over the six periods). The field names are in Spanish. Free to use with credit to TendenciasFX and the European Central Bank.
The table updates itself: the daily every business day after the ECB fixing, and the weekly every Friday.
Everything on this page is information, not investment advice, and it doesn't say when to enter any pair. The page doesn't use cookies and doesn't ask you for any data.
Embed the forex correlation table on your website
Copy this code and paste it into your site. You get the summary sentence, the table of the 28 main pairs (with the change of timeframe and period) and the date of the data, always up to date. It's free and has no ads.
