Why it is not simple division
Risking 2% of $25,000 is $500. Fifty of those is $25,000, so fifty losses empties the account — except that is wrong, and the way it is wrong matters.
Each loss is 2% of what remains, not of the original balance:
balance after n losses = balance × (1 − risk)ⁿ
losses to lose share X = ln(1 − X) ÷ ln(1 − risk)
At 2% per trade it takes 35 consecutive losses to halve the account, and 114 to lose 90%. The account approaches zero without ever reaching it.
That sounds reassuring and is not, because the damage that ends a trading career happens long before zero.
| Risk per trade | Losses to −25% | To −50% | To −90% |
|---|---|---|---|
| 0.5% | 58 | 139 | 460 |
| 1% | 29 | 69 | 230 |
| 2% | 15 | 35 | 114 |
| 5% | 6 | 14 | 45 |
| 10% | 3 | 7 | 22 |
The number the table does not show
A streak length means nothing without its frequency, which is why the win rate field is here.
At a 45% win rate, each trade loses 55% of the time. Ten losses in a row is 0.55¹⁰ ≈ 1 in 400 — so across a few thousand trades it is not unusual, it is expected. At a 35% win rate it is roughly 1 in 75, which is several times a year for an active trader.
Put those together and the picture changes. Ten losses at 5% per trade leaves 60% of the account, and it is an event that arrives on a schedule rather than a catastrophe that might. Sizing has to survive the streak that will happen, not the average that will not.
Why the survivable number is smaller than the arithmetic one
Two things end accounts before the mathematical limit:
Recovery is asymmetric. Down 50%, you need +100% to return — a larger run than the one that caused it, and the recovery calculator shows how fast that gap widens.
Behaviour breaks first. Deep drawdowns change how people trade: sizes go up to recover faster, stops widen, rules that held for months stop holding. The mathematical hole is deep; the behavioural one is what finishes the account, and it opens somewhere around 20–30% rather than 90%.
Which is why the honest planning number is not "how many losses until zero" but "how many until I stop trading the plan".