Correlation: Avoiding Hidden Concentration Risk
By Chad Smith, Director of Research & Analysis · Updated
On this page
In this lesson you will learn
- Explain positive and negative correlation and the correlation coefficient
- Identify common correlations between currency pairs
- Calculate net currency exposure across positions
- Set rules that cap correlated risk
The leverage risk lesson showed how to measure total exposure. This lesson looks at a hidden form of exposure: several positions that look separate but move together. You will learn how correlation works, which pairs tend to move together, how to work out your net currency exposure and how to cap correlated risk.
What correlation is
Correlation measures how closely two prices move together. It is expressed as a coefficient between −1 and +1:
- +1: perfect positive correlation. They move in the same direction every time.
- 0: no consistent relationship.
- −1: perfect negative correlation. They move in opposite directions every time.
Real markets fall in between. A coefficient of +0.8 means two pairs have tended to move in the same direction most of the time over the period measured. Correlation is calculated over a chosen period, such as the last 50 or 100 days, and it changes over time.
Why currency pairs correlate
Pairs share currencies. EUR/USD and GBP/USD both have the US dollar as the quote currency. When the dollar weakens broadly, both pairs tend to rise. The euro and the pound are also linked through their close economic ties.
Typical tendencies include:
- Positive: EUR/USD and GBP/USD; AUD/USD and NZD/USD; USD/CHF and USD/JPY at times.
- Negative: EUR/USD and USD/CHF; GBP/USD and USD/JPY at times.
- Cross-market: AUD/USD with metal prices; USD/CAD inversely with oil, as covered in the commodity currencies lesson.
These are tendencies that change with conditions. Check current correlations rather than relying on lists. Many platforms offer a correlation indicator or table.
The hidden concentration problem
The danger is believing you have spread your risk when you have not.
Worked example (illustrative numbers): a trader opens three trades, each risking 1%:
- Buy EUR/USD
- Buy GBP/USD
- Sell USD/CHF
Each looks like a separate setup. But all three are bets against the US dollar. If strong US data lifts the dollar, all three may hit their stops together. The real risk is close to 3% on one idea, not three separate 1% risks.
Net currency exposure
A practical way to see this is to break each position into its two currencies.
| Position | Long | Short |
|---|---|---|
| Buy EUR/USD | EUR | USD |
| Buy GBP/USD | GBP | USD |
| Sell USD/CHF | CHF | USD |
Count the exposure per currency: short USD three times; long EUR, GBP and CHF once each. The dominant risk is clearly the dollar.
Doing this for all open trades before adding a new one shows whether the new trade spreads risk or concentrates it.
Negative correlation and hedging
Two positions in negatively correlated pairs can offset each other.
Worked example (illustrative scenario): buying EUR/USD and buying USD/CHF means long EUR against USD, and long USD against CHF. The dollar exposure roughly cancels, leaving you mainly long EUR against CHF, similar to holding EUR/CHF, while paying two sets of spreads and swaps.
Some traders use offsetting positions as a deliberate hedge. For most, accidental offsets simply add costs and make results harder to understand. If you want a view on EUR/CHF, trading EUR/CHF directly is usually cleaner.
Rules to cap correlated risk
Common approaches include:
- Group cap. Treat highly correlated positions as one group, and cap the group's total risk, for example at 1.5 to 2 times your normal per-trade risk.
- One per theme. Allow only one trade per underlying idea, such as "weaker US dollar", and choose the best setup among the candidates.
- Split risk. If you want exposure through two correlated pairs, split your normal risk between them, for example 0.5% each instead of 1% each.
- Currency cap. Limit total risk in any single currency, such as no more than 2% of the account tied to the US dollar at once.
Worked example (illustrative numbers): a trader with a 1% per-trade rule and a 2% cap on any single currency already has 1% risk short USD through EUR/USD. A GBP/USD buy signal appears. They can take it at 1%, reaching the 2% cap, but would then skip a further AUD/USD buy signal until one of the trades closes.
Checking correlation yourself
You do not need special software. A simple check is to open two charts side by side on the same timeframe and compare their swings over the last few months. If the peaks and troughs line up, the pairs are positively correlated; if one peaks when the other troughs, they are negatively correlated. Many platforms and charting tools also offer a correlation coefficient indicator, where you choose the two markets and the lookback period. Check both a short period, such as 20 days, and a longer one, such as 100 days. If the two readings differ a lot, the relationship is unstable and you should be more cautious.
Correlation beyond forex
Correlation also links different markets. Long positions in a stock index, AUD/USD and a risk-sensitive cross such as AUD/JPY may all depend on a risk-on mood, as covered in the risk-on and risk-off lesson. In a sudden risk-off move, they can fall together.
Correlations also tend to rise in crises. Assets that usually move independently can fall together when investors rush to safety, so diversification is often weakest exactly when it is needed most. Leave a margin for this in your caps.
Correlation and drawdown
Correlated positions increase the size of losing streaks, because several trades lose together. That makes drawdowns deeper than your per-trade risk suggests. Counting correlated trades together is one of the simplest ways to keep your actual drawdown close to what your plan expects.
Key takeaways
- Correlation runs from −1 to +1 and measures how closely two prices move together over a period.
- Pairs that share a currency, or depend on the same driver, often correlate.
- Several trades on one theme are one large bet; break positions into currencies to see net exposure.
- Cap risk per correlated group or per currency, or split risk between correlated trades.
- Correlations change and often rise in crises, so check them regularly and leave a margin.
Check your understanding
1. A correlation coefficient of −0.9 between two pairs means what?
- A. They tend to move strongly in opposite directions
- B. They tend to move strongly in the same direction
- C. They are unrelated
- D. They never move
Show answer
A. They tend to move strongly in opposite directions
2. You buy EUR/USD and buy GBP/USD, each risking 1%. What is the main shared exposure?
- A. Short US dollar
- B. Long US dollar
- C. Long Japanese yen
- D. None
Show answer
A. Short US dollar
3. You buy EUR/USD and buy USD/CHF. Since these pairs are often negatively correlated, what tends to happen?
- A. The positions tend to offset each other
- B. The risk doubles
- C. Both always win
- D. Both always lose
Show answer
A. The positions tend to offset each other
Where to practice this
Put this into practice on a demo account first, with no money at risk.
- Featured broker
Pepperstone4.7 / 5Visit
IC Markets4.9 / 5Visit
GO Markets4.8 / 5Visit