Home / Risk Management / Risk of Ruin
Risk metricAdvanced

Risk of Ruin

Risk of ruin is the probability that an account will fall to a defined ruin threshold before growing, given a strategy's edge, win rate and the fraction of capital risked per trade.

Quick answer: Risk of ruin is the probability that an account will fall to a defined ruin threshold before growing, given a strategy's edge, win rate and the fraction of capital risked per trade.

In simple words

Risk of ruin is the chance that a string of losses drives your account below a level at which you stop trading, whether that is zero or a personal cut-off. It depends on three things: how large your edge is, how you win (win rate and payoff), and how much you risk per trade. The uncomfortable lesson is that even a profitable strategy can have a meaningful chance of ruin if the bet size is too large.

Purpose

It exists to answer the only question that ultimately matters before compounding can help you: what is the probability I do not survive to see the edge pay off?

Professional explanation

The gambler's ruin foundation

Risk of ruin descends from the classical gambler's ruin problem. Consider bets that win one unit with probability p and lose one unit with probability q = 1 − p, starting with a capital of N units against an infinitely rich opponent. If p is greater than 0.5 the probability of eventual ruin is (q/p) raised to the power N; if p is 0.5 or less, ruin is essentially certain over enough bets. The intuition generalises to trading: a positive edge makes ruin less than certain, and more units of capital (smaller risk per bet relative to capital) drive the ruin probability down exponentially.

Why the exponent is the key lever

In the (q/p)^N form, N is the number of betting units your capital represents, which is inversely related to the fraction risked per trade. Halving the risk fraction roughly doubles N, and because N is an exponent, that transformation slashes ruin probability dramatically rather than linearly. This is the mathematical heart of why small position sizing is so powerful: it is not that each loss is smaller, it is that the probability of a fatal streak falls exponentially as the risk fraction shrinks. Conversely, oversizing does not just increase drawdown, it can push ruin from negligible to near-certain.

Beyond even money: unequal payoffs

Real strategies rarely win and lose equal amounts, so the simple even-money formula is only a starting point. With asymmetric payoffs the relevant quantity is the edge per bet and the variability of outcomes, and closed-form solutions become messy. A widely used approximation frames ruin in terms of the ratio of edge to standard deviation of returns, but for realistic strategies with fat tails and serial correlation, the honest approach is Monte Carlo simulation: resample the strategy's own trade distribution thousands of times and count the fraction of paths that hit the ruin threshold. This captures streaks, skew and path dependence the formulas assume away.

Defining ruin realistically

Ruin is rarely a literal zero. For most traders the practical ruin threshold is the drawdown at which they would stop, an allocator would pull funds, or margin rules force liquidation, often a 30 to 50 percent loss rather than 100 percent. Setting the threshold higher (closer to the starting capital) makes ruin more likely for the same strategy, which is why the risk-of-ruin question must always be paired with an explicit, honest definition of the point of no return. Using zero when your real stopping point is a 40 percent drawdown badly understates the true probability.

Edge, win rate and the fraction interact

The three drivers are not independent knobs. A higher win rate lowers ruin only if the payoff structure holds; a strategy with a 90 percent win rate but rare catastrophic losses can have a higher ruin probability than a 45 percent win-rate trend system, because the tail dominates. Edge (positive expectancy) is necessary but not sufficient: with a large enough risk fraction, even a genuine edge carries non-trivial ruin. The safe region is a positive edge combined with a small risk fraction and a payoff distribution without ruinous tails, and the only reliable way to locate it for a specific strategy is simulation.

From risk of ruin to system controls

Because analytical formulas rest on assumptions markets violate, production systems translate risk-of-ruin thinking into hard controls rather than trusting a single number. Small per-trade fractions keep the exponent large, portfolio heat caps prevent correlated aggregation that inflates effective bet size, and daily-loss circuit breakers and kill switches enforce the ruin threshold mechanically so that no streak or bug can breach it. Risk of ruin is thus both a diagnostic (simulate it) and a design principle (engineer so that the modelled probability stays negligible).

Formula

RoR = (q ÷ p)ᴺ for p > 0.5, even-money bets

p = win probability, q = 1 − p = loss probability, N = capital measured in bet units (inversely proportional to risk-per-trade fraction). For p ≤ 0.5, ruin is effectively certain over enough bets. Unequal payoffs and fat tails need Monte Carlo, not this closed form.

Practical example

Illustrative example (Indian market)

Suppose an even-money strategy wins 55 percent of the time (p = 0.55, q = 0.45) and you risk 2 percent per trade, so your capital is about N = 50 betting units. The gambler's-ruin estimate of hitting zero is (0.45/0.55)^50 = (0.818)^50, which is astronomically small, effectively zero. Now risk 20 percent per trade, so N = 5 units: (0.818)^5 ≈ 0.37, a 37 percent chance of ruin despite the identical 55 percent edge. On a ₹5,00,000 account this is the difference between risking ₹10,000 and ₹1,00,000 per trade, and it turns a safe edge into a coin-flip on survival. In practice you would also raise the ruin threshold above zero and run Monte Carlo on the real payoff distribution, which usually shows meaningfully higher ruin than the even-money formula.

An NSE options-selling strategy may show a 90 percent win rate, tempting large size, but its losses cluster on gap days; a Monte Carlo on its actual trade log, including a few expiry-day tail losses, often reveals a risk of ruin far above what the high win rate suggests, which is why survivors size such strategies very small.

Advantages

  • Forces attention on survival probability, not just average return
  • Reveals the exponential payoff of reducing the risk fraction
  • Links win rate, payoff and bet size into one decision variable
  • Monte Carlo estimation captures streaks and tails formulas miss

Limitations

  • Closed-form formulas assume even money, fixed odds and independent bets, which markets violate
  • It is highly sensitive to the estimated edge, which is itself uncertain
  • Fat tails, serial correlation and regime change can make realised ruin far exceed the model
  • Defining the ruin threshold is subjective and materially changes the answer
  • A comforting low number can create false confidence if inputs are wrong

Why it matters in practice

  • Risk of ruin is the ultimate constraint: no edge matters if you do not survive to realise it
  • It quantifies why oversizing is fatal even with a real edge

Common mistakes

  • Assuming a positive edge alone makes ruin negligible regardless of bet size
  • Using the even-money formula on a strategy with very unequal payoffs
  • Setting the ruin threshold at zero when the real stopping point is a 40 percent drawdown
  • Trusting a high win rate while ignoring rare catastrophic losses that dominate the tail
  • Estimating edge from an overfit backtest and feeding that inflated number into the formula
  • Treating a single computed number as certainty rather than simulating the distribution

Professional usage

Quantitative desks estimate risk of ruin by Monte Carlo on the strategy's own return distribution, deliberately using conservative edge estimates and realistic, non-zero ruin thresholds, then size so the modelled probability is vanishingly small under stress. They treat the analytical formulas as intuition pumps rather than answers, stress-test with fat-tailed and correlated resampling, and enforce the ruin threshold with hard daily-loss limits and kill switches so that neither a streak nor a software bug can carry the account past the point of no return.

Key takeaways

  • Risk of ruin is the probability of hitting a stopping-point loss before the edge compounds
  • It falls exponentially as the risk fraction shrinks, because capital-in-units is the exponent
  • Even a real edge can carry high ruin if the bet size is too large
  • Use Monte Carlo on the real payoff distribution and a realistic, non-zero ruin threshold

Frequently asked questions

What is risk of ruin?
Risk of ruin is the probability that an account falls to a defined ruin threshold, such as zero or a personal stop-trading level, before it grows. It depends on the strategy's edge, its win rate and payoff, and the fraction of capital risked per trade.
What is the risk of ruin formula?
For even-money bets with win probability p greater than 0.5, the classical gambler's-ruin estimate is (q/p) raised to the power N, where q = 1 − p and N is capital measured in bet units. For p at or below 0.5, ruin is effectively certain over enough bets.
How does position size affect risk of ruin?
Enormously. The number of bet units N is inversely related to the risk fraction and sits in the exponent, so halving the fraction roughly doubles N and slashes ruin probability. Oversizing can push ruin from negligible to near-certain even with a positive edge.
Can a profitable strategy still go broke?
Yes. A positive edge lowers ruin below certainty but does not eliminate it. With a large enough risk fraction, or a payoff distribution with catastrophic tails, even a genuinely profitable strategy can carry a serious probability of ruin.
Why is a high win rate not enough?
Because ruin depends on payoff as well as frequency. A 90 percent win-rate strategy with rare huge losses can have higher ruin than a 45 percent trend system, since the tail losses dominate. Win rate must be read alongside the loss distribution.
Why use Monte Carlo instead of the formula?
The closed-form formulas assume even money, fixed odds and independent bets, which real strategies violate. Monte Carlo resamples the strategy's actual trades to capture unequal payoffs, streaks, fat tails and path dependence, giving a far more honest ruin estimate.
How should I define the ruin threshold?
Realistically, as the drawdown at which you would actually stop trading or be forced to, often 30 to 50 percent rather than zero. A higher threshold raises the computed ruin, so using zero when you would really quit at 40 percent understates the danger.
What is gambler's ruin?
Gambler's ruin is the classical probability problem of a bettor with finite capital facing a richer opponent; it gives the (q/p)^N result and is the mathematical foundation from which trading risk of ruin is derived.
Does risk of ruin account for correlation?
The simple formulas do not; they assume independent bets. In reality correlated positions inflate the effective bet size, so risk of ruin must be paired with portfolio-heat controls and estimated by simulation that respects correlation.
Is risk of ruin the same as maximum drawdown?
No. Maximum drawdown is the worst observed peak-to-trough loss; risk of ruin is the probability of reaching a defined loss threshold. Drawdown is a realised outcome, ruin is a forward-looking probability, though the two are closely linked.
How sensitive is risk of ruin to the edge estimate?
Very. Because the edge sits inside an exponent, a modest overestimate of win rate or payoff can understate ruin by orders of magnitude. Since edges are hard to estimate and often overfit, conservative inputs are essential.
How do systems enforce a low risk of ruin?
By keeping per-trade fractions small, capping portfolio heat to stop correlated aggregation, and enforcing the ruin threshold mechanically with daily-loss circuit breakers and kill switches, so no streak or bug can carry the account past the stopping point.

Voice search & related questions

Natural-language questions people ask about Risk of Ruin.

What is risk of ruin?
It is the chance you lose enough to have to stop trading before your edge pays off. It depends on your edge, your win rate and how much you risk per trade.
Can I go broke with a winning strategy?
Yes, if you bet too big. A real edge lowers the odds of ruin but does not remove them when the position size is large.
How do I lower my risk of ruin?
Risk less per trade. Because capital in bet-units sits in the exponent, cutting the fraction in half cuts the ruin probability far more than half.
Is a ninety percent win rate safe?
Not by itself. If the rare losses are huge, they dominate. A high win rate with catastrophic tails can be riskier than a low win rate with small losses.
What counts as ruin for a real trader?
Usually not zero, but the drawdown where you would actually quit or get liquidated, often thirty to fifty percent. Measure ruin against that honest level.
How do I actually calculate my risk of ruin?
Simulate. Resample your own past trades thousands of times and count how often the account hits your stopping level. That beats any tidy formula.

Sources & references

    Last reviewed 11 July 2026. Educational content only — not investment advice. Markets and rules change; verify current conventions with SEBI, NSE/BSE and your broker.

    ← Previous
    Portfolio Heat
    Educational content only — not investment advice. Examples use illustrative numbers and simplified models. Algorithmic trading and derivatives involve substantial risk. See our Risk Disclosure and SEBI Disclaimer.