Risk of Ruin Calculator

Expectancy tells you the destination. Risk of ruin tells you whether you arrive

Two traders take the same signals from the same strategy for a year. One finishes up 60%. The other is out of the market by March. The strategy was identical — a 45% win rate at 1:2, a solid +0.35R per trade. The only difference between them was how much they put on each trade.

That is the gap this calculator exists to close. Expectancy is a statement about the average of an infinite number of trades. Your account does not get an infinite number of trades; it gets a specific, finite, badly-ordered sequence of them, and somewhere in that sequence is the worst losing streak you will ever have. Risk of ruin is the probability that streak arrives before the edge has paid for it.

Why a winning strategy still blows accounts up

Because losing streaks are longer than they feel. At a 45% win rate, six losses in a row is not remarkable — over 200 trades it is close to a coin flip that you get one. Ten in a row is uncommon but entirely ordinary; roughly one trader in fifty running that strategy will see it.

Now put a number on it. At 1% risk, ten straight losses costs about 10% of the account: unpleasant, survivable, and you take the eleventh trade exactly as written. At 5% risk the same streak costs 40%. You now need a 67% gain to get back to flat, and — far more importantly — almost nobody takes the eleventh trade the same way. They halve the size, or skip the setup, or start “waiting for confirmation,” and at that moment the tested strategy stops being the strategy that is running.

That is the real mechanism behind most blown accounts. Not a bad system. A good system, sized so that its normal variance was intolerable.

The two numbers that decide it

Edge per trade. Expectancy in R: win rate × reward:risk − loss rate. This is what pushes the balance up. At 45% and 1:2 it is +0.35R.

Risk per trade.How much of the account each of those R's is worth. This scales both the gains and the variance — but not equally. Double your risk and the expected growth roughly doubles; the depth of the drawdowns roughly doubles too, while the probability of hitting any given depth rises far faster than that. Ruin is not linear in position size, which is exactly why intuition fails here and a table does not.

The ladder above shows the same edge run at every position size from 0.25% to 10%. It is the single most useful output on the page, and the column to read first is the drawdown one.

Percentage risk versus fixed cash risk

These behave very differently in a drawdown and the calculator models both.

Percentage of current balance is self-correcting. A £10,000 account risking 2% stakes £200; after a bad run down to £6,000 it stakes £120. The position shrinks with the account, losses become geometric rather than linear, and the balance can never mathematically reach zero. It also means the recovery is slower, because you are climbing back with smaller trades than the ones that dug the hole.

Fixed cash risk — deciding once that you risk £200 a trade and never revising it — is where accounts actually die. That £200 is 2% at the start and 3.3% by the time the account is at £6,000. The risk per trade is silently rising at precisely the moment the account can least afford it, and unlike the percentage method there is nothing stopping it reaching zero.

What the ruin threshold should be

“Ruin” is not zero. Almost nobody trades an account to zero; they stop somewhere above it, and where they stop is a personal number worth being honest about before you need it.

A drawdown of 50% requires a 100% gain to undo. At 75% you need to quadruple what is left. Those are the arithmetic facts, and they are why professional risk limits sit so much tighter than retail intuition suggests. If you trade a funded account the threshold is chosen for you — usually 8–12% total drawdown — which is a far harsher constraint than most challenge-takers size for. The prop firm challenge calculator models that case directly, including the daily-loss limit that ends most attempts.

How to get inputs worth putting in the box

This tool amplifies whatever you feed it. A win rate remembered rather than measured, or measured on a chart you had already scrolled through, produces a confident answer to the wrong question.

The number you want is a win rate and an average R taken from a real sample of occurrences, marked on price you could not see ahead of. That means bar-by-bar replay, not scrolling back and finding the setups that worked — the second one reliably produces a win rate ten to twenty points above the truth, which here shows up as a comfortable ruin figure that is fiction.

And check the sample before you trust it. Thirty trades constrains almost nothing; the backtest sample size calculator tells you whether your win rate is a measurement or a rumour, and the mistakes that make backtest results worthless covers the ways a good-looking number gets manufactured by accident.

What this model assumes — and where it breaks

Every trade is independent, with a fixed win rate and a fixed reward:risk. Real trades are not independent: strategies work in some regimes and stall in others, so wins and losses cluster. All clustering does is make streaks longer than modelled, which means the true risk of ruin is higher than the figure above, never lower.

There are no costs. Spread, commission and swap come straight off the average winner, and a strategy at +0.05R gross is frequently negative net. There is no slippage, no gapping through a stop, and no weekend risk — all three of which convert a planned 1R loss into something larger at exactly the wrong moment.

And it assumes you keep taking the trades. That is the assumption most likely to fail. The value of the drawdown columns is not that they predict your equity curve; it is that they tell you, in advance, what you are signing up to sit through — so you can pick a size you will still be trading at the bottom of it.

Frequently asked questions