Currency Correlation
A currency pair like EUR/USD is just a price: how many US dollars it takes to buy one euro. Currency correlation measures whether two of these pairs tend to move in the same direction, opposite directions, or with no relationship at all, expressed as a number between -1 and +1. A reading near +1 means the two pairs rise and fall together almost in lockstep; near -1 means when one goes up the other reliably goes down; near 0 means they move independently, with no useful pattern between them.
In forex, EUR/USD and GBP/USD are a classic positive pair, both hovering around +0.85 to +0.90 most years, because the euro and the British pound often react to the same European economic news while both trading against the same US dollar. USD/CHF, by contrast, sits close to -0.90 against EUR/USD, since the Swiss franc and the euro tend to strengthen or weaken against the dollar in mirror image.
The practical problem correlation solves is hidden risk: opening a EUR/USD buy and a GBP/USD buy at the same time can feel like two separate trades, but with a +0.90 correlation they are really one large bet on the dollar weakening, doubling the potential loss if that view is wrong. Checking a correlation table before stacking positions is how traders avoid accidentally tripling their exposure to a single economic event while believing they've diversified.
The same idea shows up outside forex under different names. In crypto, most altcoins move with a correlation to Bitcoin often above 0.80, meaning a Bitcoin crash tends to drag the whole market down regardless of any single coin's own news. In stocks, it's called sector correlation - airline stocks or bank stocks tend to rise and fall together because they share the same economic exposure, like fuel costs or interest rates.
The most important thing about these relationships is that they are regimes rather than constants. In calm conditions pairs respond to their own economies and correlations reflect genuine economic links. When risk appetite turns, an entirely different structure asserts itself: the dollar, yen and franc strengthen together as havens, commodity currencies fall together, and pairs that had been unrelated for months suddenly move in lockstep because a single factor is driving everything. A correlation table calculated over the previous year describes the first regime and tells you very little about the second, which is when the number is most needed.
The practical response is to recalculate over several windows rather than trusting one. A 20-day reading captures the regime in force now, a 120-day one captures the structural relationship, and the interesting information is in the gap between them - two pairs that are structurally uncorrelated but are currently moving together are being driven by something temporary and shared, which is exactly the situation where a position that felt diversified is not. Any spreadsheet will calculate it from two columns of closing prices, which is worth doing occasionally rather than accepting a broker's table built over an unstated period.