Losses and the gains that undo them are not symmetric, because the gain is measured on the smaller balance that is left. −10% needs +11%, −30% needs +43%, −50% needs +100%. The practical consequence is that per-trade risk size is decided long before the losing streak arrives, and it is what determines whether the streak is an inconvenience or the end of the account.
Lose 10%, make 10% back, and you are still down.
$10,000 becomes $9,000. A 10% gain on $9,000 is $900, so you end at $9,900. The loss was 10% of the larger number and the recovery was 10% of the smaller one, and that single fact — a gain is always measured on what is left — is the most consequential piece of arithmetic in risk management.
The curve, and where it turns
required gain % = drawdown ÷ (100 − drawdown) × 100
| Drawdown | Gain needed |
|---|---|
| −10% | +11.1% |
| −20% | +25% |
| −30% | +42.9% |
| −40% | +66.7% |
| −50% | +100% |
| −70% | +233% |
| −90% | +900% |
Read down the right column and the character of the problem changes about a third of the way in. Up to roughly 20%, recovery is a normal stretch of trading — the same edge, applied for longer. Past 30% it needs a better run than the one that lost the money, which is a strange thing to require of a trader who has just been losing. Past 50% you are not recovering an account; you are building a new one out of what remains.
Most traders know that 50% needs 100%. Fewer have looked at what sits between, which is where the actual decisions live.
Where the number gets decided
Not during the drawdown. Before it — in the position size.
Two traders with identical edges and an identical run of ten losing trades:
- Risking 1% per trade. Down about 10%. Needs +11% to be whole. That is a decent quarter.
- Risking 5% per trade. Down about 40%. Needs +67%. That is a career year, required immediately after the worst stretch of their year.
Same skill, same streak, same market. The difference was set before the first of those ten trades was placed, by a number typed into a box.
This is why "risk 1-2% per trade" survives as advice despite sounding timid. It is not a claim about how confident you should be. It is a claim about what a losing streak costs, and losing streaks arrive on their own schedule regardless of confidence.
The half the arithmetic does not show
There is a second cost that no calculator prints, and it is usually the one that finishes the account.
Deep drawdowns change behaviour. Position sizes go up to make it back faster. Stops widen, because a stop that would have been fine at the start of the month now closes a trade the account "needs". Rules that held for six months stop holding in the seventh, and the trader can usually explain why this particular situation is different.
The mathematical hole at −50% is +100%. The behavioural hole is that the person climbing out is not trading the way the person who dug it did.
Which means the number worth deciding is not the one you use during a drawdown — it is the one you set while the account is fine. A drawdown level at which you stop and review is arithmetic when chosen at zero and a negotiation when chosen at −40%.
What to do with this
Three things, in order of how cheap they are:
- Look up your own number. Take your worst historical drawdown and check the gain it required. Most traders have never done this and are surprised by the size.
- Work backwards from a streak you can accept. Decide the losing run you want to survive without changing how you trade, then see what each risk percentage costs over that run. That gives you a size, rather than a feeling.
- Write down the review level now. Not a level where you quit — a level where you stop adding and look at what changed.
The drawdown recovery calculator does the first one on any number you give it, including the money and the number of winning trades. The position size calculator handles the second.
Neither is complicated. The arithmetic here is one line long. It is just that almost nobody looks at it until the number is already large.
See this in your own trades
Reading about a pattern is one thing; finding it in your own history is another. Import a statement from MT4/MT5, a broker CSV or a crypto export and the journal shows you where your intent and your execution went apart.
Start free