Currency Correlation: How to Avoid Doubling Your Risk
Currency correlation explained: what the coefficient means, why pairs move together, and how to count your net exposure so three trades are not one bet.
You risk 1% on EUR/USD, 1% on GBP/USD and 1% on a short USD/CHF. Three careful trades, each sized by the book. Then one strong US data release lifts the dollar and all three stops are hit within the hour. Currency correlation is the reason, and it is easy to miss until it costs you.
Quick answer: Currency correlation measures how closely two pairs move together, on a scale from −1 (always opposite) to +1 (always together). Pairs that share a currency or a driver often move as one, so several correlated trades can be a single bet. Count your net exposure to each currency and cap the risk on it.
What is currency correlation?
Every currency pair has two currencies in it, and many pairs share one. EUR/USD, GBP/USD and USD/CHF all contain the US dollar. When the dollar moves sharply, it pushes all three at once, whatever is happening to the euro, the pound or the franc. Currency correlation is the name for that tendency: the degree to which two pairs rise and fall together.
Traders measure it with the correlation coefficient, a single number between −1 and +1. You do not need the formula to use it. You need to know what the ends of the scale mean:
- +1 means the two pairs always move in the same direction. When one rises, so does the other.
- 0 means there is no consistent link. Knowing what one pair did tells you nothing about the other.
- −1 means they always move in opposite directions. When one rises, the other falls.
Real pairs sit somewhere in between, and they rarely reach either end. As a rough rule of thumb, readings above +0.7 or below −0.7 are treated as strong, readings between 0.3 and 0.7 either way as moderate, and anything closer to zero as weak. These bands are conventions, not laws, but they give you a working vocabulary.
Two details matter more than the formula. First, correlation is measured on changes, usually daily percentage moves, not price levels: two pairs that both trended up for a year can look related while their day-to-day moves have little in common. Second, correlation describes direction, not size. A high reading means two pairs tend to move the same way on the same days, not by the same number of pips.
The glossary entry for correlation has the one-line definition. For risk purposes, the question is simpler: if this trade loses, how likely is it that my other open trades lose at the same time? That is what forex correlation answers.
Why currency pairs move together
Pairs do not correlate by coincidence. There are two main reasons, and once you know them you can predict most relationships without looking anything up.
Reason 1: a shared currency
The most common cause is a currency that appears in both pairs. EUR/USD and GBP/USD both have the US dollar as the quote currency. When the dollar strengthens broadly, both fall, because both are “something against the dollar”. The euro and the pound have their own stories, but a large dollar move can swamp them.
The direction depends on which side the shared currency sits. In EUR/USD the dollar is the quote currency; in USD/CHF it is the base currency. A stronger dollar pushes EUR/USD down and USD/CHF up. That is why those two pairs usually have a strongly negative correlation: the same dollar move, seen from opposite sides of the quote. Our lesson on currency pairs and quotes covers reading base and quote if it is not yet second nature.
The US dollar is the most common shared currency by far. According to the BIS Triennial Survey, the dollar is on one side of close to nine in every ten currency trades worldwide. That makes “how much am I betting on the dollar?” the first exposure question for most traders.
Reason 2: a shared driver
Pairs can also correlate when they share no currency at all, because the same force pushes both. The Australian and New Zealand dollars are the classic example: both economies export commodities, both trade heavily with China, and both currencies respond to the global mood. AUD/USD and NZD/USD therefore tend to move together more closely than the shared dollar alone would explain. The Canadian dollar’s link to oil, and the yen’s and Swiss franc’s role as safe havens, work the same way.
Market mood is the biggest shared driver of all. When investors are confident, they tend to buy higher-yielding and commodity currencies; when they are frightened, they tend to buy the dollar, the yen and the franc. Switch the mood below to see which currencies usually rise and fall together.
Tap a currency to see why it behaves this way.
Risk-on and risk-off moods move groups of currencies together. Tendencies, not rules: a currency with its own big news can ignore the mood.
That shared mood is why a portfolio of long AUD/USD, long NZD/USD and long AUD/JPY can feel diversified while being one bet on optimism. Our guide to risk-on vs risk-off explains the mechanism in full, and the Academy lesson on commodities and currency correlations goes deeper into the commodity links.
Typical forex correlations between the major pairs
You do not need to memorise numbers, and you should be wary of any precise figure quoted as if it were fixed. What you need is a feel for the usual relationships, so you notice when two trades are really one. Over long periods, these tendencies have been fairly consistent:
- EUR/USD and GBP/USD are usually positively correlated, often strongly. The EUR/USD GBP/USD correlation is the one most traders meet first, because both are among the most traded pairs and both are quoted against the dollar.
- EUR/USD and USD/CHF are usually strongly negative. The dollar is on opposite sides, and the euro and the franc are close European neighbours.
- AUD/USD and NZD/USD are usually strongly positive, through the shared dollar plus shared drivers.
- EUR/USD and USD/JPY are usually negative but weaker. The yen has strong drivers of its own, especially interest rates and risk mood, which dilute the dollar link.
- USD/CAD usually moves against the “something/USD” pairs, and also responds to oil.
- Cross pairs such as EUR/GBP, which contain no dollar, tend to have weaker links to the dollar pairs.
The grid below shows illustrative long-run correlations for six major pairs. Tap a cell to see what it means for two open positions, then use the counter underneath to tick a basket of trades and see how many times you are really betting on the dollar.
Typical correlation between pairs
| EUR/USD | GBP/USD | AUD/USD | USD/JPY | USD/CHF | USD/CAD | |
|---|---|---|---|---|---|---|
| EUR/USD | ||||||
| GBP/USD | ||||||
| AUD/USD | ||||||
| USD/JPY | ||||||
| USD/CHF | ||||||
| USD/CAD |
Tap any cell to see what that correlation means for two open positions.
How many times are you really betting on the dollar?
Illustrative long-run correlations between six major pairs, not live data. Tick positions in the counter to see your real USD exposure across a basket of trades.
Pairs with the dollar on the same side show positive readings, and pairs that cross from “something/USD” to “USD/something” show negative ones. That block structure is the shared-currency effect at a glance. Our EUR/USD trading guide covers what drives the euro side of these relationships.
Correlations change over time
A correlation is a summary of the recent past, not a property of the pair. It is usually calculated over a rolling window, such as the last 50 or 100 trading days, and it drifts as that window moves. A figure you read in an old article, or in a table someone made last year, may no longer describe the market you are trading.
Two things push correlations around. The first is one currency getting its own story. If the Bank of England surprises the market while the European Central Bank is quiet, GBP/USD can move sharply while EUR/USD barely reacts, and their correlation weakens for a while. The most dramatic example in recent history came in January 2015, when the Swiss National Bank removed its minimum exchange rate against the euro and the franc jumped within minutes. Any trader relying on the usual EUR/USD and USD/CHF relationship that morning found it did not hold.
The second is a panic. When fear takes over, the market tends to trade everything as a dollar or safe-haven decision. Correlations between dollar pairs tighten, and positions that looked independent start losing together. Correlations often rise exactly when you most need your trades to be independent.
The practical lesson is to treat any correlation reading as a tendency to check, not a fact to lean on. Many charting platforms and financial websites publish correlation tables with selectable periods, and you can build your own in a spreadsheet: put the daily percentage changes of two pairs in two columns and use the CORREL function over the last 50 or 100 rows. Comparing a short window with a long one tells you whether the relationship is holding or drifting.
A simpler default: assume two pairs that share a currency or a driver are correlated unless you have a recent reason to think otherwise. Being wrong that way costs you a smaller position; being wrong the other way costs you a doubled loss.
Net currency exposure: count currencies, not trades
Correlation tables are useful, but there is a simpler way to see what you are betting on, and it needs no statistics. Break every open position into its two currencies, note which one you are long and which you are short, then add them up. This is the net currency exposure method.
The rule for splitting a trade is mechanical. Buying a pair means you are long the base currency and short the quote currency. Selling a pair means the opposite. So:
- Buy EUR/USD = long euro, short dollar.
- Buy GBP/USD = long pound, short dollar.
- Sell USD/CHF = short dollar, long franc.
Laid out side by side, they make one large short-dollar trade with three different partners.
| Position | What you are long | What you are short |
|---|---|---|
| Buy EUR/USD | EUR | USD |
| Buy GBP/USD | GBP | USD |
| Sell USD/CHF | CHF | USD |
| Net | EUR, GBP, CHF (one each) | USD, three times |
You can go one step further and count the actual amounts. At 0.10 lots, each trade controls 10,000 units of the base currency. Using illustrative prices of 1.0800 for EUR/USD, 1.2700 for GBP/USD and 0.8800 for USD/CHF, the dollar legs add up as shown below.
The shape matters more than the amounts. The euro, pound and franc columns each have one entry; the dollar column has three, all on the same side. That column is your real position.
The same method catches exposures that a correlation table hides. Buying EUR/JPY and buying USD/JPY share no obvious pair in common, but both are short the yen. Buying EUR/USD and selling EUR/GBP look related, but split them and you find the euro legs cancel: you are left long the pound and short the dollar, which is just a GBP/USD trade by another route. The lesson on correlation risk has quiz questions that drill exactly this splitting step.
Three trades, one bet: what correlated pairs do to your risk
Now connect the exposure to money. Suppose each of the three trades above is sized so that hitting its stop loses 1% of the account. Per trade, that looks disciplined. But the three stops are not independent events. A single dollar rally, from a strong jobs report, a hawkish speech or a sudden flight to safety, can push all three to their stops together. The realistic worst case is a 3% loss from one idea that turned out to be wrong.
A risk per trade rule assumes each trade is a separate event, and correlated pairs quietly break that assumption.
The effect compounds over a losing run. Set the ladder below to 1%, then to 3%, and compare what ten losses in a row do to the account.
For comparison: at 1% you keep 90%, at 5% you keep 60%, at 10% you keep 35%.
Ten losses in a row at the chosen risk. Three correlated trades at 1% each behave like one trade at 3%: ten losses leave about 74% of the account, against about 90% at a true 1%.
At a genuine 1% per idea, ten consecutive losses leave about 90% of the account. At an accidental 3%, they leave about 74%, and recovering that needs a gain of roughly 36%. Ten losses in a row sounds extreme, but for any method that wins around half its trades, a losing streak of that length is something you should expect to meet eventually.
The simulator below runs 100 random trades with your chosen risk, win rate and reward-to-risk. Run it a few times at 1%, then at 3%, with the other settings unchanged. The method is the same; only the effective size differs.
Each run uses the same settings with different luck. Run it several times: the shape changes, the drawdowns don't go away.
The drawdowns lesson explains why deep drawdowns are so much harder to climb out of than shallow ones. Correlation is one of the main ways traders reach them without ever breaking a written rule.
Rules for managing currency correlation risk
None of this means you should only ever hold one trade. Your risk rules just need to see what the market sees. Four rules cover most of it.
Rule 1: cap the total risk on any one currency
Alongside your per-trade limit, set a limit on how much you can lose on any single currency, adding up every open trade that is long or short it in the same direction. A common setting is around 1.5–2% of the account per currency. If you already have 1% riding on a weaker dollar and your cap is 1.5%, the next dollar-short trade can risk only 0.5%, or must wait.
Rule 2: cap total open risk across the account
The second limit is the sum of what you would lose if every open position hit its stop at once. We suggest something around 4–6%. When you reach it, nothing new opens until something closes. This is the backstop for correlations you did not spot, including the ones that appear suddenly in a panic. Our forex risk management guide treats this as one of the core layers of a risk plan.
Rule 3: size correlated trades as a group
When setups share a currency, decide the total risk for that currency first, then divide it. Three dollar-short setups under a 1.5% cap become three trades at 0.5% each, or one at 1% and one at 0.5%, or just the best one. Position sizing then turns each of those risk figures into a lot size from the stop distance, exactly as it would for a lone trade. The sizing lesson in Forex Foundations walks through the arithmetic.
Rule 4: be suspicious of hedges
Correlation also tempts traders in the opposite direction: opening a trade that is meant to cancel another. Most of these “hedges” are illusions. Buying EUR/USD and buying USD/CHF looks like two opposing dollar trades, and on the dollar it roughly is. But the dollar legs cancel and something else is left behind.
What remains is long the euro and short the franc, which is a EUR/CHF position. You may or may not want that trade, but it is a trade, not protection. You paid two spreads to open it, and if you hold it overnight you pay or receive swap on both legs. The same logic applies to buying EUR/USD and selling GBP/USD: the dollar cancels and you are left with EUR/GBP.
Opening the exact opposite trade on the same pair is a different illusion. It freezes the loss rather than removing it, costs a second spread, and postpones the real decision: whether your idea is still valid. If it is not, closing the trade is cheaper.
Worked example: long EUR/USD, long GBP/USD and short USD/CHF
The example below puts the whole method together on a realistic account. All prices are illustrative.
Nothing in steps 2 to 4 was wrong on its own. The mistake only becomes visible at step 5, when you stop counting tickets and start counting currencies.
Common mistakes with correlated pairs
Counting tickets instead of currencies
Your platform lists positions by pair, but the market moves currencies. Before every new trade, split your open positions into currency legs and check whether the new one adds to a column that is already heavy.
Trusting an old correlation figure
A correlation table from last year, or from an article written in a different market, is a description of the past. Relationships weaken when one currency has its own news and tighten in a panic. Check a recent window, or assume pairs that share a currency are correlated until you know otherwise.
Mistaking offsetting trades for a hedge
Positions that cancel on one currency leave a position in another, with extra costs. If you want the cross, trade it directly.
Adding a correlated trade to “confirm” a loser
When EUR/USD is going against you, buying GBP/USD because “it looks even better” doubles the same idea at the moment the market is telling you it may be wrong. Correlation can support a setup before you enter; it should never be used to add risk to a losing one.
Forgetting the crosses and the yen
Exposure hides in pairs without the dollar. EUR/JPY, GBP/JPY and AUD/JPY all share the yen and respond to risk mood, so a basket of long yen crosses is one risk-on bet. Count currencies on every pair you hold.
Frequently asked questions
What is currency correlation in forex?
Currency correlation is a measure of how closely two currency pairs move together, from −1 (always opposite) through 0 (unrelated) to +1 (always together). Pairs correlate when they share a currency, such as the US dollar, or a common driver, such as commodity prices or market mood. It matters because correlated trades tend to win and lose together.
Which currency pairs are most correlated?
Pairs that share a currency on the same side are usually the most positively correlated, such as EUR/USD and GBP/USD, or AUD/USD and NZD/USD. Pairs with the dollar on opposite sides, such as EUR/USD and USD/CHF, are usually strongly negative. Exact readings change over time, so check a recent period rather than relying on a fixed number.
Is it bad to trade correlated pairs at the same time?
Not necessarily, but you must size them as one position. Two or three trades that share a currency in the same direction can lose together on a single news release. Decide how much you are willing to lose on that currency in total, often 1.5–2% of the account, and divide that amount between the trades.
How do I calculate currency correlation?
Take the daily percentage changes of two pairs over a period, such as the last 50 or 100 trading days, and calculate the correlation coefficient. In a spreadsheet the CORREL function does this in one step. Use price changes rather than price levels, and compare a short window with a long one to see whether the link is holding.
Can I use correlation to hedge forex trades?
Usually not in the way people hope. Buying EUR/USD and buying USD/CHF roughly cancels the dollar but leaves you long the euro and short the franc, which is a EUR/CHF trade with two spreads. A genuine reduction in risk normally means a smaller position or a closed trade, not a second trade.
Why do correlations between currency pairs change?
Correlations shift when one currency has its own news, such as a surprise interest rate decision, which weakens its link to others. They also tend to tighten during market panics, when traders treat almost everything as a dollar or safe-haven decision. That is why a correlation is a tendency to check, not a fixed rule.
Free course
Learn this step by step
“Risk & Money Management” is a free BrokerCut Academy course: 7 short lessons (about 59 minutes), quizzes, interactive tools and a certificate when you finish.
Enrol free
