The dollar at the center, and correlations between pairs
Lesson 19 · about 10 min
Nearly nine in ten FX trades have the US dollar on one side. That makes the dollar the hub of the whole market, and it means that "separate" pairs are often the same bet wearing different clothes. Understanding this keeps you from taking three trades that are really one.
Why the dollar is the hub
The US dollar is the world's reserve currency, the currency most commodities are priced in, and the currency most international debt is issued in. When a bank in Tokyo needs Swiss francs, it usually sells yen for dollars and buys francs with the dollars, because the USD/JPY and USD/CHF markets are far deeper than JPY/CHF. Crosses are priced off their dollar legs (Module 2), and dollar liquidity is what the whole system runs on.
Consequences for a trader:
- A "dollar move" moves every major at once. Strong US data pushes EUR/USD, GBP/USD, AUD/USD and NZD/USD down and USD/JPY, USD/CHF and USD/CAD up in the same minute.
- US news dominates. The FOMC and US CPI move EUR/USD more than most ECB decisions do.
- The dollar is also a safe haven. In a global panic, the dollar tends to strengthen even when the trouble started in the US, because the world needs dollars to pay dollar debts.
The dollar index
The US Dollar Index (DXY) is a weighted average of the dollar against six currencies: EUR 57.6%, JPY 13.6%, GBP 11.9%, CAD 9.1%, SEK 4.2%, CHF 3.6%. It is a useful one-line summary of dollar strength, with a caveat: it is more than half euro, so it is largely an inverted EUR/USD. If EUR/USD falls 1% and the other five are unchanged, DXY rises about 0.58%.
Traders use it to answer "is this a dollar move or a euro move?" If EUR/USD is falling and DXY is rising, the dollar is strong. If EUR/USD is falling and DXY is flat while EUR/GBP and EUR/JPY are also falling, the euro is weak. That distinction tells you which side's news to look at.
Correlation between pairs
Because of the shared dollar leg, pairs move together in predictable ways. The correlation coefficient runs from +1 (always the same direction) to −1 (always opposite). Illustrative values over a typical month:
| Pair 1 | Pair 2 | Typical correlation | Why |
|---|---|---|---|
| EUR/USD | GBP/USD | +0.7 to +0.9 | Both are "something versus dollar" and Europe-linked |
| EUR/USD | USD/CHF | −0.8 to −0.95 | Dollar on opposite sides; EUR and CHF are tied |
| AUD/USD | NZD/USD | +0.8 to +0.95 | Both commodity, risk-on, Pacific |
| USD/CAD | Crude oil | −0.5 to −0.8 | CAD strengthens with oil |
| USD/JPY | US 10-year yield | +0.5 to +0.8 | JPY sold when US yields rise |
| EUR/USD | USD/JPY | Unstable, −0.3 to +0.5 | Dollar leg says negative; risk sentiment says positive |
| AUD/JPY | Equity indices | +0.5 to +0.8 | The risk-on/off barometer |
These drift. A correlation that held at +0.9 for six months can break to +0.2 in a week when one country has its own crisis. Check them monthly rather than memorising them.
Correlated positions are one trade
This is where the risk management course and this one meet again. Suppose you risk 1% on each of three trades: long EUR/USD, long GBP/USD, short USD/CHF. Each looks like a separate idea. All three are short the dollar. If the dollar rallies on a hot CPI print, all three lose at once, and your "1% per trade" has become 3% on one event.
| Positions | Nominal risk | Correlated risk if the dollar moves |
|---|---|---|
| Long EUR/USD (1%) | 1% | 1% |
| Long EUR/USD + long GBP/USD (1% each) | 2% | About 1.8% (correlation ~0.8) |
| + short USD/CHF (1%) | 3% | About 2.7% |
| + long AUD/USD (1%) | 4% | About 3.5% |
The approximation is rough; the point is not the decimals but the direction: the dollar exposure adds up nearly one for one. The rule of thumb is to count total exposure to each currency, not the number of trades. Four trades that are all short USD get the risk budget of one trade split four ways, or one of them gets taken and the others are left alone.
Key idea: Most pairs share a dollar leg, so most positions share a dollar bet. Count your exposure by currency: if three open trades are all short USD, you have one 3% trade, not three 1% trades.
Using correlation instead of being used by it
Correlation is also information:
- Confirmation. If EUR/USD is breaking a level but GBP/USD and DXY are not confirming, the break is more likely a euro-specific move or a false one.
- Choosing the cleanest pair. If you think the dollar will strengthen, pick the pair where the other currency is also weak (for example, sell AUD/USD if Australia's data is soft) rather than the pair where the other side is strong.
- Hedging by accident. Long EUR/USD and long USD/CHF are close to offsetting; you pay two spreads and two swaps for almost no net exposure. Beginners do this without noticing.
A quick check before opening any trade: list the open positions, write the currency each one is long and short, and total the net exposure per currency. If the new trade adds to a currency you are already heavily exposed to, size it as an addition to that exposure, not as a fresh trade.
Try it: Take the daily closes for EUR/USD, GBP/USD and USD/CHF for the last 20 trading days (most platforms will export them). Compute the daily percentage changes and the correlation between each pair of series in a spreadsheet. Then write down what your risk would have been on the worst single day if you had been long the first two and short the third at 1% risk each.
Recap
- The dollar is on one side of the great majority of FX trades; dollar moves shift every major at once, and US data dominates.
- DXY is a weighted dollar basket, 57.6% euro, useful for separating dollar moves from single-currency moves.
- EUR/USD and GBP/USD are strongly positive, EUR/USD and USD/CHF strongly negative, AUD/USD and NZD/USD strongly positive; correlations drift, so check monthly.
- Correlated positions are one trade; total exposure by currency, not by number of positions.
- Use correlation for confirmation and for picking the cleanest pair, and to spot accidental hedges that cost two spreads for nothing.