What Is Risk of Ruin?
Risk of ruin is the probability that an account loses enough that recovery becomes impractical, and it is decided far more by position size than by win rate. A losing run that costs 20% at small size costs nearly 70% at large size, and the recovery required grows faster than the loss.
Most discussion of risk is about being wrong. This is about something else: how much of the account a run of ordinary losses removes, and whether what remains can still get back.
How it works
Risk of ruin is the probability that losses reach a level the account cannot come back from. It has three inputs: how often you lose, how much you lose when you do, and how many chances you get. Only one of those is under your direct control, and it is the middle one.
Size dominates. Improving a win rate is hard, slow and uncertain. Halving position size is a decision you can make in a second, and it moves this number far more than any amount of improvement to the trading itself.
And recovery is asymmetric. Losing 20% requires making 25% to get back. Losing 50% requires making 100%. Losing 70% requires making 233%. The hole gets steeper the deeper it goes, which is the mechanism that turns a bad stretch into a permanent one.
Why the three inputs are not equal. Win rate and average loss are properties of the strategy and they move slowly, if at all. Position size is a setting, it moves instantly, and it enters the calculation as an exponent rather than as a term — which is why halving it does not halve the danger, it collapses it.
And the number is only meaningful against a horizon. Ruin is not a fixed threshold; it is the point where the capital left can no longer place the trades the strategy requires. An account that cannot meet a minimum position size has been ruined at 40% down, while one trading fractional size has not been.
A worked example
Use the real run length from this site’s series. Direction runs average 2.01 bars, and the longest one-way run measured was 11 bars. Take that as the losing streak — not a worst case, a measured one.
At 2% risk per trade, eleven consecutive losses leaves the account at 0.98¹¹ = 80.1% of where it started. Down about 20%, needing 24.9% to recover. Unpleasant, entirely survivable, and the strategy gets to keep operating.
At 10% risk per trade, the same eleven losses leaves 31.4%. Down about 69%, and the recovery required is 219% — the account now has to more than triple to get back to where it began.
Same market. Same strategy. Same eleven losses. The only thing that changed was a number chosen before any of it happened, and it is the difference between a bad month and an account that is finished as a going concern.
The run itself was never unusual. Eleven bars one way is what this series actually did. A strategy sized so that a normal streak ends it was not unlucky; it was arithmetically doomed and waiting.
Why a good strategy does not protect you
Positive expectancy says the average trade makes money. It says nothing about the order. A run of losses at the start of a sequence and the same run at the end produce identical averages and completely different outcomes, because one of them happens to an account that still exists.
On this site’s series 95% of bars sat below a prior peak, the deepest drawdown ran 3.76%, and the longest wait for a new high was 73 bars — in a stretch that finished up 3.61% overall. Being under water is the normal condition, not the exception, and the sizing has to assume that.
The original data
Direction runs on this site’s shared series average 2.01 bars, with a longest of 11. Drawdown covered 95% of bars, the deepest was 3.76%, and the longest recovery took 73 bars.
Run those eleven against two position sizes and the whole argument is there: 2% risk leaves 80.1% of the account and needs 24.9% back; 10% risk leaves 31.4% and needs 219% back. The loss grew by a factor of about 3.4 and the recovery required grew by a factor of about 8.8.
And 0.0098 comes out of every round trip on top, on the winners and the losers alike, which pushes every one of those figures slightly further the wrong way.
When it fails
The characteristic failure is sizing from a run of good results. A strategy works for three months, confidence rises, size rises with it — and the size is now set by a period in which the long losing run happened not to occur. When it does occur, it arrives against a position size chosen by the calm stretch, and the account discovers its risk of ruin was never a number anyone calculated. Nothing about the strategy changed; the only variable that moved was the one that decides survival.
A second failure is sizing in contracts rather than in risk. The same number of units is a different percentage of the account in every volatility regime, so a fixed quantity is a moving risk.
A third is assuming the stop caps the loss. A gap fills below it, so the realised worst case is larger than the planned one — and every calculation above assumed the plan held.
A fourth is averaging down, which raises the size of a position precisely as the thesis weakens.
And a fifth is treating survival as pessimism. Sizing small is not a lack of confidence in the strategy — it is the condition under which the strategy gets enough trades for its edge to mean anything at all. An edge that needs a thousand trades cannot be collected by an account that lasts forty.
Related
Risk management covers the sizing rules that keep this number negligible. Expectancy covers whether there is an edge to survive for in the first place. And volatility covers why a fixed quantity is a moving risk.
This is the one piece of arithmetic I wish somebody had put in front of me before I placed a single trade. It is not about being right or wrong — a perfectly good strategy with the size set too high still ends the account, and the reason is that the recovery needed grows faster than the loss does. Nothing about trading better fixes it.
— Michael Whitman
This page is educational, not financial advice. Test every idea on your own charts before risking money.