Long EURUSD, short USDCHF and long AUDUSD look like three separate ideas — but across a central-bank week they can behave like one trade at triple size.
You place three trades on a Monday morning. Long EURUSD. Short USDCHF. Long AUDUSD. Three different pairs, three different charts, three different reasons. The account feels diversified.
Then the dollar moves, and all three positions move together — in the same direction, at roughly the same time, by roughly the same proportion. The account didn't hold three ideas. It held one idea, expressed three times.
That is correlation. It is one of the few concepts in trading that changes the risk of a position without changing anything you typed into the order ticket.
Correlation measures how two instruments move together — nothing more
Correlation is a single number between -1 and +1. It describes how closely the returns of two instruments have moved in the same direction over a chosen period.
- +1 — the two moved in lockstep, same direction, every period measured.
- 0 — no consistent relationship.
- -1 — they moved in opposite directions, consistently.
Two details matter more than the number itself.
First, correlation is backward-looking. It is calculated over a window — 20 days, 100 days, 12 months — and each window gives a different answer. A pair showing +0.85 on a 20-day window may show +0.30 on a one-year window. Neither is wrong. They are measuring different stretches of history.
Second, correlation says nothing about size of move. Two instruments can be correlated at +0.95 while one moves four times as far as the other. Direction agreement and magnitude are separate questions. Correlation answers only the first.
In FX, most correlation comes from one shared leg
Every currency pair is a ratio of two things. EURUSD is the euro priced in dollars. USDCHF is the dollar priced in francs. AUDUSD is the Australian dollar priced in dollars.
Look at what they share: the US dollar appears in all three. So when the dollar itself repricesss — broadly, across the board — it pushes every one of those quotes at once.
Long EURUSD is a short-dollar position wearing a euro label. Long AUDUSD is a short-dollar position wearing an Australian dollar label. Short USDCHF is a short-dollar position with no disguise at all. Three tickets. One factor.
This is why a trader can hold five positions and still be running a single bet. The instruments are diversified. The driver is not.
The same logic applies outside FX. Gold and silver share a precious-metals factor and a dollar factor. Two index CFDs tracking overlapping large-cap universes share most of their constituents. Crude and an energy-heavy equity index share the oil price. Diversification is about drivers, not about the number of symbols on your watchlist.
A worked example: three trades, one exposure
The following numbers are illustrative only — they are not live prices, forecasts, or expected outcomes.
Assume an account with 10,000 USD of equity. Assume, for simplicity, that each position carries a pip value of roughly 10 USD per 1.00 standard lot.
The trader opens:
- Long 1.00 lot EURUSD
- Short 1.00 lot USDCHF
- Long 1.00 lot AUDUSD
Each position has a 40-pip stop. Looked at individually, each trade risks about 400 USD — 4% of equity. That sounds disciplined.
Now assume a broad dollar move of a size that produces roughly 40 pips of adverse movement in each pair. Because all three positions are short the dollar, they do not offset. They stack.
- EURUSD: −400 USD
- USDCHF: −400 USD
- AUDUSD: −400 USD
- Total: −1,200 USD, or 12% of equity
The trader intended 4% risk per position. What actually existed was a single 12% dollar position split across three tickets. Margin tells the same story from the other side: three lots consume three lots' worth of margin while delivering one lot's worth of diversification.
Flip the correlation and the arithmetic flips with it. Had one of the three been a dollar-long position of similar size, part of the exposure would have netted off — reducing both the loss and the potential gain, and leaving a smaller net factor bet than the ticket count suggests.
Neither outcome is good or bad. The point is that the real risk number was never on any individual ticket.
Correlation tightens exactly when you need it loosest
Here is the part that catches experienced traders as well as new ones. Correlations are not stable. They tend to rise — sometimes sharply — around scheduled macro events and during stress.
On a quiet day, EURUSD trades on European data, AUDUSD on commodity flows, USDCHF on its own risk dynamics. Local stories dominate, and the pairs drift apart. Measured correlation falls.
In the minute after a major release, the local stories stop mattering. One common factor — the dollar, or global rate expectations — moves every quote simultaneously. Correlation converges toward 1. A portfolio that looked spread out on Friday behaves like a single leveraged position on Wednesday.
This week offers a clear calendar of those moments. German Ifo Business Climate lands Monday 26 October at 09:00 GMT+3 with a forecast of 89.8 against 89.9 previously. US Durable Goods Orders follow Tuesday 27 October at 12:30 GMT+3, forecast 0.6% versus 0% prior. ECB President Lagarde speaks Tuesday at 17:00 GMT+3. Australian inflation data arrives Wednesday 28 October at 00:30 GMT+3, with the monthly rate forecast at 0.7% against 0.4% previously. A Federal Reserve rate decision also falls in this window, with commentary widely framed around a pause.
Each of those is a moment when separate charts can start answering to the same question. Spreads typically widen through those windows too, which affects every open position at once rather than one at a time.
The takeaway: size the factor, not the ticket
Before adding a position, ask one question: what is this trade actually a bet on?
If the honest answer is the same as an existing position — the dollar, the rates path, the oil price, risk appetite — then the new trade is not a fourth idea. It is an increase in the size of an idea you already own. Treat the group as one position and size the group, not each line.
A practical habit: list open trades, write the dominant driver beside each, and total your intended risk per driver rather than per symbol. The number that comes out is usually larger than expected. That number, not the per-ticket stop, is what a release at 12:30 GMT+3 tests.
At GCC Brokers — founded in 2016 and regulated by the FSC Mauritius under licence GB22200739 — STP is our primary execution model, with client positions hedged with liquidity providers and execution fully automated: no requotes, no filtering by profitability, symmetric slippage. Consistent fills matter most in precisely those windows where correlations converge and several positions reprice at once.
If you want to test the maths above, run it on your own open positions before this week's releases rather than after them.
For information only. This is market commentary and education, not investment advice. Figures are as at the time of writing.