forex-scalping.com

Total Open Risk: The Number That Matters More Than Per-Trade Risk

You risk 1 percent per trade, which sounds safe. But what if you have six trades open at once?

Total open risk is the sum of the potential losses across all current positions, assuming all stops are hit. Six trades at 1 percent each is 6 percent of the account at risk simultaneously. If those trades are correlated, such as several dollar pairs, the true risk is closer to a single large bet.

Setting a cap on total open risk is a basic safeguard. Many traders limit it to 3 to 5 percent, depending on their approach. When the cap is reached, no new trades open until some close or stops move to breakeven.

Another subtlety concerns trades whose stops have moved. A position with a stop at breakeven carries no risk of loss, though it still has opportunity cost, and it could be counted as zero risk. But be careful: slippage can still produce a loss on a stop.

Total risk should also account for event risk. Holding multiple positions into a major news release concentrates the danger.

The right cap depends on correlation, strategy, and your tolerance for drawdowns.

Pay attention to a worst-case scenario: if everything goes wrong, how much do you lose, and can you accept that?

Practical step: add a total open risk cell to your trade tracking and record the maximum you reach each week.

A quick way to track it: add a column in your spreadsheet that sums the risk in percent of every open trade. Check it before each new entry. If the total is already at your max, the new trade waits, no matter how good it looks. Most people set their max somewhere between 2 and 5 percent depending on experience. I'd also count correlated trades as one block. Two dollar trades at 1 percent each in the same direction are really 2 percent on one idea. When a stop moves to breakeven, that trade's open risk drops to zero, which frees up room. That gives a natural reason to manage winners properly before adding new trades.

Correlation makes it worse. Four trades at 1% each that are all dollar shorts is closer to 4% on one idea than four separate 1% risks. Total open risk and correlation have to be looked at together. A simple rule is to count highly correlated pairs as one trade for your limit. Long EURUSD and long GBPUSD together count as one slot, sized at 1 percent combined, so each gets 0.5 percent. It feels restrictive when both setups look good, but on days when the dollar moves hard after US data at 1:30pm Lagos, both trades move together anyway.

Total open risk is the right number to cap, and the correlation point is key. A practical way to count it: group open trades by the currency they share. Six trades at 1% each on USD pairs can behave like one 4-6% position when the dollar moves, so the effective open risk is much higher than the six separate stops suggest.

Some traders cap risk per currency as well as total, e.g. no more than 2% on any one currency across all positions. That catches the concentration before the total cap does.

Two things I count that the spreadsheet versions here leave out. The first is pending orders, because three resting limits on dollar pairs can all fill inside the same spike, and at that point total open risk jumps from 1 percent to 4 before I can react.

So any pending order counts at full risk from the moment it is placed. If the cap is full, the limit does not go in, which removes the temptation to park orders everywhere and hope only one triggers.

The second is breakeven stops. Fairman is right that slippage can still bite, and I would rather put a number on it than count zero.

I carry each breakeven trade at 0.2R against the cap, and gold at 0.3R because its stops slip further on my fills. Five breakeven trades on majors then still use about 1 percent of room, which feels right when a US release can gap price straight through all of them.

On FXS Prop Desk's per currency limit, I apply it to gold as well and treat XAUUSD as mostly a short dollar position. Long gold and long EURUSD together is one dollar bet in my book, even if the symbols look unrelated.