Glossary

Perp trading glossary

Risk of Ruin

RoR

Risk of ruin is the probability a positive-edge strategy still blows up from variance — too many losses in a row, sized too aggressively.

Definition

Risk of ruin is the probability an account reaches a chosen failure point — down 50%, down 90% — even though the strategy has positive expectancy. Ruin is almost never caused by losing your edge; it's caused by sizing that edge too big, so a normal losing streak compounds into a hole you can't climb out of.

The Kelly criterion gives the growth-optimal fraction to risk per trade for a known edge. Because you only ever estimate your edge from a noisy sample, the practical rule is to risk a fraction of Kelly — commonly a quarter.

Formula

Expectancy (R) = p·W − q·L Kelly fraction f* = (p·W − q·L) ÷ (W·L)

p = win rate, q = 1−p, W = average win in R, L = average loss in R. The trader-safe ceiling is ¼·f*, to absorb the error in your own estimates.

Worked example

Win rate 50%, average win 2R, average loss 1R: expectancy = 0.5·2 − 0.5·1 = +0.5R per trade. Full Kelly = 0.5 ÷ (2·1) = 25% of the account per trade — mathematically optimal but wild in practice. A quarter-Kelly ceiling is ~6.25%, and most traders sit well below even that.

Why it matters

  • Most blowups are oversizing, not a lost edge — a +EV strategy run at 3× the right size still ruins.
  • Losses compound worse than gains: a 50% drawdown needs a 100% gain to recover. Kelly is the math that respects that geometry.
  • 'Risk 1% per trade' is folklore — too aggressive for a weak edge, far too timid for a strong one. The right number falls out of your measured win rate and average R.
  • Endpoint ruin math understates path risk: real intra-run drawdown is typically 30–50% deeper than the end-point probability suggests.

Related