Correlation and exposure: when several trades are really one

You follow the 1% rule to the letter. This morning you bought EUR/USD with 1% risk, then GBP/USD with 1%, then AUD/USD with 1%. Three well-managed trades, each at 1%. Then a US data release comes out better than expected, the dollar takes off, and all three stops are hit in the same minute. You have just lost 3% on what was, in reality, a single bet.

What is correlation?

Correlation measures how closely two markets move together. It is expressed as a coefficient between −1 and +1. Close to +1, the two markets rise and fall at the same time. Close to −1, they move in opposite directions. Around 0, their moves have no consistent link.

Pairs that share a currency are often correlated. EUR/USD and GBP/USD both have the dollar as their quote currency: when the dollar strengthens sharply, both tend to fall together. Conversely, EUR/USD and USD/CHF often move in opposite directions, since the dollar is on one side of the pair in one and on the other side in the other.

To find the recent correlation between two markets, many platforms and websites offer matrices updated over different periods: a week, a month, six months. Pick a period that matches your holding time. A swing trader who holds positions for several days will look at correlation over one to three months; an intraday trader will care about the last few weeks.

Stacked risk

When two positions are highly correlated, their losses tend to arrive together. You are not diversifying anything: you are doubling your bet on the same idea. So you need to count your risk per market idea, not per line in your platform.

The opposite trap exists too. Buying EUR/USD and buying USD/CHF at the same time often means taking two positions that largely cancel each other out: when one wins, the other loses. You pay the spread twice for almost no exposure. That is not clever hedging; it is usually a sign of muddled analysis.

The same reasoning applies to other markets. A buy on the DAX and a buy on the S&P 500 often rest on the same appetite for risk. A gold buy and a short on the dollar against the yen can both profit from a weaker dollar. Before opening a position, ask yourself which idea you are really trading.

Measuring your exposure by currency

A simple method is to split each pair into two legs. Buying EUR/USD means being long the euro and short the dollar. Selling USD/JPY means being short the dollar and long the yen. Then add up the risk by currency.

This small table shows that three of the four positions are playing the same thing: a weaker dollar. The fourth, EUR/GBP, is of a different nature, but it adds to the euro exposure already there through EUR/USD.

Setting a cap on total risk

The solution is to add two caps to your per-trade risk rule. The figures below are a cautious starting point, to be adapted to your strategy and your results.

• Risk per trade: 1% maximum.
• Risk per correlated group (same currency, same theme): 2% maximum in total.
• Total open risk across the whole account: 3 to 5% maximum.
• If a new trade breaks a cap: reduce its size, close an existing position or skip it.

Go back to the opening example with these caps. The three buys against the dollar break the 2% limit per correlated group. So you have three options: keep only the best of the three signals at 1%, take all three at about 0.66% each, or take two at 1%. Either way, if the dollar takes off, your loss stays below 2% instead of 3%.

Good habits

1. Before each trade, look at the positions already open and the currency they share.
2. Check the recent correlation of the markets you combine, over a period close to your trading horizon.
3. Add up risk by currency or by theme, not just by position.
4. Be careful with news releases that hit a currency present in several of your trades, such as US jobs data for the dollar.