Risk of ruin: the math that decides whether you survive a losing streak

A profitable strategy can still ruin you. That sentence sounds like a contradiction, and it is the most important non-contradiction in trading. A strategy with a genuine positive edge will, over enough trades, make money — but only if the account survives long enough to get there. Risk of ruin is the probability that it does not: that a normal losing streak drains the account before the edge can express itself.

Losing streaks are not bad luck — they are arithmetic

People treat a run of losses as a sign something is broken. Usually nothing is. If a strategy wins 50% of the time, a streak of six losses in a row happens roughly once every sixty-odd trades — it is not rare, it is scheduled. Lower win rates make long streaks more common still: trend-following systems with a 40% win rate routinely string together eight or ten losers. The streak is not the strategy failing. The streak is the strategy behaving exactly as its win rate guarantees it sometimes will.

What drives risk of ruin

Risk of ruin rises and falls with three things. Win rate and risk-reward set how the streaks are shaped — a lower win rate means longer expected streaks. But the variable you control directly, and the one that dominates, is position size: how much of the account you risk per trade. Risk 2% per trade and a ten-loss streak costs roughly 18% — painful, survivable, recoverable. Risk 20% per trade and the same ten-loss streak — the same normal, scheduled streak — is total destruction. Same strategy, same edge, same market. Only the sizing changed, and the sizing decided survival.

Keeping risk of ruin near zero

1. Risk a small, fixed fraction per trade — 1-2% of the account is the well-worn range. Small per-trade risk is what makes any streak survivable.
2. Know your strategy's streak profile — use the backtest, and better, Monte Carlo, to see the longest losing runs it has plausibly produced. Size so the worst of those is uncomfortable, not fatal.
3. Size off the stress case, not the average — the drawdown you must survive is the bad one, not the typical one. Use the Monte Carlo tail.
4. Never increase size to 'win it back' — raising risk after losses, the martingale instinct, is the single fastest route to ruin.
5. Cap total exposure — per-trade sizing is not enough if ten correlated positions can all lose at once; limit aggregate risk too.

The mindset shift

Risk of ruin reframes the job. Your first task is not to maximise returns; it is to guarantee survival, because survival is the precondition for the edge ever paying you. Get position sizing right and risk of ruin drops close to zero — losing streaks become weather you wait out rather than events that end you. Get it wrong and the best strategy in the world cannot save an account that was sized to be killed by an ordinary run of bad luck.

Noon Barbari's Monte Carlo page shows the distribution of losing streaks and drawdowns across 1,000+ resampled runs, so you can size against the stress case rather than the average. The strategy designer enforces per-trade and aggregate risk limits directly, and every backtest reports the worst losing run it saw.

An edge tells you the destination. Risk of ruin tells you whether you live long enough to arrive. Size every trade so the answer is yes.