The idea
Risk of ruin is the probability that you hit an account level you cannot come back from, before your positive expectancy has had time to express itself.
Note what it is not. It is not a measure of whether your method works — you can have a genuinely profitable method and still be very likely to be destroyed by it. Those two facts live together comfortably, and the combination is the single most common way skilled traders fail.
The reason is that expectancy is an average over many trades, and you only get those many trades if you survive the sequence. Ruin is about the path, not the destination.
Streaks are not rare
Most people badly underestimate how ordinary a long losing run is. Here is the chance of encountering a streak of a given length somewhere within a hundred trades.
So the planning question is not "what if I lose eight in a row". It is "when I lose eight in a row, what will be left". At 1% risk the answer is about 92% of your capital. At 8% the answer is about 51%, and the recovery needed is close to 100%.
How ruin scales with size
The relationship is not linear. Doubling your risk per trade does far more than double your chance of a serious drawdown.
Position sizing is not about how much you make when you are right. It is about whether you are still in the game when the average finally shows up.
Edge does not save you
The most counter-intuitive part of the subject: a larger edge does not compensate for oversizing, because the two act on different things.
Edge improves the long-run average. Size determines whether a specific short sequence bankrupts you. A method with terrific expectancy, traded at 15% per trade, is overwhelmingly likely to be destroyed by an ordinary streak long before the terrific expectancy shows up anywhere.
This is also why "I will size up because I am confident in this one" is so dangerous. It is a bet that the next trade is not part of the streak — and streaks are made entirely of trades somebody felt confident about.
Increasing size after a winning run. It feels like pressing an advantage, and it is the opposite: you are raising your exposure at the exact moment the sample has flattered you, and the mean reversion that follows now costs more. The disciplined version is to let the risk percentage stay constant, which sizes you up automatically as the account grows and down as it shrinks — without any decision being made in a good mood.
Read it yourself
- Your hit rate is 45%. What is the chance you meet a seven-trade losing streak in your next hundred trades?
- You risk 6% per trade. What does that streak do to your capital?
- Your method has excellent expectancy. Does that reduce the answer to question 2?
Show answers
1. About 75% — it is the likely case, not the unlucky one.
2. Seven consecutive 6% losses compound to 0.947 ≈ 0.65. You are down about 35%, needing roughly 54% to recover, having done nothing wrong.
3. No. Expectancy operates on the average of many trades; the streak operates on seven specific ones. Your edge is the reason the account grows between streaks — it has no effect whatsoever on what a streak does when it arrives. Only size controls that, and size is the one variable entirely within your control.
Terms used
- Risk of ruin
- Probability of reaching an unrecoverable account level before the edge plays out.
- Losing streak
- Consecutive losses. Frequent, expected, and the thing sizing exists to survive.
- Path dependency
- The order of outcomes mattering, not just the average of them.
- Fixed fractional
- Risking a constant percentage of current capital, which scales automatically both ways.