Risk of ruin

Quick answer: Risk of ruin is the probability that cumulative losses reduce capital below a threshold at which the account can no longer continue, before any long-run edge can express itself. It is a property of the path, not the average, so a positive-expectancy strategy can still ruin the trader.

In simple words

Risk of ruin is the chance that a bad run wipes you out before your good idea has time to work. Imagine an edge that pays off on average over a thousand trades. That average is no help if the first thirty trades take your account below the level where you can keep going. Ruin is about the order things happen in, not the long-run average. Two things push it up: betting a large slice of the account each time, and having occasional losses far bigger than your typical win. A high win rate feels reassuring, but if the rare loss is large enough, the account can still be emptied by one cluster of bad luck. Survival first, edge second.

A precise definition

Risk of ruin is the probability that a sequence of outcomes drives capital below a threshold from which the trader cannot continue, before the strategy's long-run edge has had the trades it needs to appear. The threshold is not necessarily zero. For a margined trader it is the level at which positions are force-closed; for a fund it may be a redemption trigger or a stop-out mandated by risk. The key word is before. Every edge is a statement about a large number of trades, and the law of large numbers only rewards those who are still trading when the large number arrives. Ruin is the event of being removed from the sample early. This reframes the whole subject: the question is not whether a strategy makes money on average, but whether its path stays above the floor long enough for the average to matter.

The classical formula

For a simple game of even-money bets — win one unit with probability p, lose one unit with probability one minus p — the probability of eventually going broke starting from u units of capital is ((1 minus p) divided by p), all raised to the power u. The structure is worth internalising. The base, (1 minus p) over p, is below one whenever p exceeds one-half, so more capital, a larger u, drives ruin down. But it does so as an exponent, which means the effect is geometric, not linear. Doubling your units does not halve your risk of ruin; it squares it. This single formula is the cleanest available demonstration that survival is bought with capital units relative to bet size, and that the relationship is far steeper than intuition suggests.

Why bet size dominates win rate

Because capital enters the formula as an exponent, bet size is the most powerful lever, and it acts superlinearly. Halving the units of capital — equivalently, doubling the bet as a fraction of the account — does not double ruin; it can nearly triple it, as the worked figures show. Raising the win rate helps, but it is the weaker lever, and it comes with a trap. If a higher win rate is bought by risking more per trade to win less per trade, the loss size scales up with the win rate and the two effects fight. A strategy is not made safe by winning often; it is made safe by keeping each bet small enough relative to capital that no plausible losing cluster reaches the floor. Bet small, and even a modest edge is durable. Bet large, and even a strong edge is fragile.

Short premium and negative skew

Selling options tends to win often and lose rarely but large — a negatively skewed distribution. The high hit-rate flatters the strategy and hides its true risk of ruin, because the classical even-money formula assumes wins and losses of equal size. When the occasional loss is several times the typical win, the effective units of capital per worst-case loss are far fewer than the account balance suggests, and ruin is correspondingly higher than the win rate implies. This is why a short-premium book that has printed a long string of green days can still be one gap away from a loss that consumes months of credit. Hull and standard derivatives texts make the same point through the payoff asymmetry: a strategy whose losers dwarf its winners must be sized as if the rare loss is the normal case, because for ruin it is.

Ruin is a path property, not an expectation

The most important and least intuitive fact is that positive expectancy does not prevent ruin. Expectancy is an average over outcomes; ruin is an event on a particular ordering of them. A strategy can have a genuinely positive edge and still carry a meaningful probability of hitting the floor first, purely because of the sequence in which losses arrive. This is why two traders running the identical positive-edge strategy can end the year in opposite places: not because one was better, but because one met the losing cluster while still capitalised and the other did not. Managing risk of ruin means managing the path — keeping bet size small, capital ample and the worst-case loss survivable — so that the edge, if it is real, is given the trades it needs to show up.

Common mistakes

• Believing a positive expectancy makes an account safe ignores that ruin is a property of the path, so a genuinely profitable strategy can still empty the account if the losing cluster arrives before the edge has had enough trades.
• Reading a high win rate as low risk of ruin overlooks loss size, because a strategy that wins small and often but loses large and rarely has a worse risk of ruin than its hit-rate suggests.
• Sizing bets large to make the account grow faster raises risk of ruin superlinearly, so doubling the bet as a fraction of capital can nearly triple the chance of being wiped out for the same edge.
• Applying the even-money formula to a short-premium strategy understates the danger, because that formula assumes equal win and loss sizes and short premium's losses are several times its wins.
• Confusing the account balance with units of survivable capital hides the true risk, since what matters is how many worst-case losses the account can absorb, not the rupee figure on the statement.
• Adding capital after a drawdown to keep the same large bet size resets the balance but not the risk-of-ruin arithmetic, so the account remains fragile to the next cluster of losses.

Professional usage

Institutions manage risk of ruin as survival probability rather than expected return. Funds set stop-out levels, maximum-drawdown limits and per-strategy risk budgets specifically so that no losing sequence can remove them from the game before their edge compounds. Quant desks run Monte Carlo simulations of the return path, not just the mean, to estimate the chance of breaching a floor over a horizon, and they size positions to hold that probability low. Redemption terms, gates and reserve capital all exist to lengthen survival. Ed Thorp's practical use of the Kelly framework was as much about avoiding ruin as maximising growth — he bet fractionally below optimal precisely to protect the path. Retail traders cannot access reserve capital or redemption gates, but the core discipline, keeping bet size small enough that no plausible run reaches the floor, is fully available and free.

Frequently asked questions

Voice search & related questions

Last reviewed 9 July 2026. Educational content only — not investment advice.