Counting a drawdown in stop-outs instead of percent

[Fairman]

Counting a drawdown in stop-outs instead of percent

When people talk about drawdown limits they usually use percentages, like 10% max drawdown. I find it more useful to translate that into the number of full losses.

With fixed fractional risk, each loss is a percentage of the current equity, so the losses get slightly smaller as the account falls. The number of consecutive losses to reach a given drawdown is log of (1 minus drawdown) divided by log of (1 minus risk).

At 1% risk, a 10% drawdown takes about 10.5 losses in a row.
At 0.5% risk, about 21.
At 2% risk, about 5.2.

For a 20% drawdown:
At 1%, about 22.
At 0.5%, about 44.
At 2%, about 11.

The drawdown doesn't need consecutive losses, of course. A mix of wins and losses with more losses than usual gets there too. But thinking in number of losses gives a feel that percentages don't. A 10% limit at 2% risk is five bad trades. That can happen in one morning.

For prop accounts with a fixed dollar limit rather than percentage, it's simpler: limit divided by risk per trade. $10,000 limit, $500 per trade is 20 losses.

I keep a note on my screen with how many losses I can take before my personal stop and before the account limit. On a bad day it helps to know I've used 4 of 10 rather than a vague 4%.

[FXS HongKong Scalper]

Counting drawdown in losses is a very practical way to think about it. It makes the risk per trade choice clearer too. If a trader knows they can handle a run of 15 losses without losing confidence, the risk per trade can be set so that 15 losses stays within the drawdown limit. That links the number to personal tolerance, not just a percentage. It also helps when explaining risk to someone new, since a number of losses is easier to picture than a percentage drawdown.

[Fairman]

The recovery side is the reverse. After a 10% drawdown, you need an 11.1% gain to get back to the start. At 1% risk, with a system that makes 0.2R per trade on average, that's about 55 trades on average, assuming the risk stays fixed in dollars, and more if it's fixed in percent of a smaller account. Knowing how long recovery takes on average makes the drawdown limit feel more concrete. If the risk is a fixed percent of a smaller account, the dollar gain per win is smaller too, so recovery takes longer. That's a tradeoff of fixed fractional sizing that isn't often discussed. It protects the account on the way down and slows it on the way back up.

[Fairman]

Costs add to the count. If each loss also carries commission and slippage of 0.1R, then 10 losses at 1% risk is closer to 11% drawdown. The number of losses to reach a limit is smaller than the formula suggests. I'd recalculate with your actual average loss, from your trade history, rather than the planned 1R. A trade history export gives the average loss in money. Dividing by the planned risk shows how much bigger the real loss is than 1R, usually 5% to 15% for scalpers once costs and slippage are included. That ratio is the number to use in the formula. It brings the count closer to reality and makes the limits feel tighter.

[Fairman]

Some traders cut risk after a certain number of losses, for example halving it after five. That changes the count. At 1% for the first five losses and 0.5% after, a 10% drawdown takes about 16 losses instead of 10.5. It slows the fall, and it also slows recovery, but it's a reasonable way to buy time when things go wrong. The rules for when to return to full risk matter as much as when to cut. One option is to return to full size once half the drawdown is recovered. Without a clear rule, people tend to stay small too long after losses or jump back too soon after a couple of wins.

[Fairman]

For multiple strategies on one account, the count has to include all of them. If two systems each risk 1% and can lose at the same time, a bad day might bring two losses per signal. Counting stop-outs per strategy hides that. A combined count, total risk lost so far today, is the number that matches what the account sees. A daily count across all strategies also makes it obvious when two systems are more correlated than you thought. If both lose on the same days most of the time, they're really one system with two names. Knowing that changes how much risk each should carry.