TL;DR: Risk of ruin is the chance an account ever falls to a level you can't trade from, and position size moves it more than any other input. On one 86-trade E-mini record, 1% risk per trade gave a 23.98% chance of ever losing half the account. Short tests hide that risk, because ruin can take hundreds of trades to show up.
Risk of ruin answers a blunt question. If you keep trading exactly this way, what is the chance the account gets cut to a level where you have to stop? The answer depends on your win rate, your average win against your average loss, how much you risk per trade and where you draw the line called ruin. Of those four, only the risk per trade is something you set fresh every morning, and that is why the number belongs in a trading plan instead of a textbook.
This guide works through the definition, the formula, its better replacements and a full worked study on 20 years of E-mini S&P 500 daily bars, then turns the result into a risk per trade you can actually use.
What Risk of Ruin Means

Risk of ruin is the probability that an account is ever reduced to a level you have defined as the end. The classic gambling version is simple: stake everything on one coin toss and the chance of being wiped out is 50%. Spread the same money over many small bets and the risk per bet falls, but every extra bet is another chance for a bad sequence, so ruin risk builds up with play.
For a trader, ruin rarely means a zero balance. It means one of three things:
- A self-chosen floor, such as losing half the starting equity, past which you would stop and rebuild.
- An account rule, such as a maximum loss or trailing drawdown on a funded or evaluation account, which closes the account long before zero.
- A practical floor, the balance below which your smallest position size is too big for your stop.
That choice changes the answer by a large factor. The same strategy at the same size can have a tiny chance of ever reaching zero and a large chance of ever losing a quarter of the account. A risk of ruin figure without a stated ruin level is a number without a unit.
The inputs that drive risk of ruin are always the same:
- Win rate (p), the share of trades that make money.
- Payoff ratio (b), the average win divided by the average loss.
- Risk per trade (f), the share of the account lost when a trade hits its stop.
- Ruin threshold (T), the drawdown that counts as ruin.
Treat risk of ruin as a property of a strategy at a given size, not of the strategy alone. The same entries and exits can be nearly safe at one size and almost certain to fail at another.
Related lesson: Maximum drawdown
The Risk of Ruin Formula
The risk of ruin formula most traders meet first is a two-outcome shortcut. Every trade either wins b units of risk or loses one unit, and the account is ruined after a net loss equal to the threshold.
Risk of ruin = (q ÷ (p × b)) raised to the power (T ÷ f)
Here q is the loss rate (1 minus p), p × b is the expected win per unit risked, T is the ruin threshold as a share of the account and f is the risk per trade as a share of the account. The exponent T ÷ f is the number of full losses that separates you from ruin.
A hand-worked risk of ruin example
Take illustrative inputs: a 50% win rate, a 1.5 payoff ratio, 5% risk per trade, and ruin defined as losing half the account.
- Loss rate q = 0.5, and p × b = 0.5 × 1.5 = 0.75.
- q ÷ (p × b) = 0.5 ÷ 0.75 = 0.6667.
- The exponent is T ÷ f = 0.50 ÷ 0.05 = 10.
- 0.6667 raised to the power 10 = 1.73%.
Two warnings come with that 1.73%. First, when q ÷ (p × b) is 1 or more, the strategy has no edge and the formula returns 100% at any size. A coin flip with an even payoff is exactly that case. Second, the shortcut assumes the dollar risk never changes and ignores how far a win carries the account away from the floor.
Two better versions of the ruin equation
The two-outcome risk of ruin shortcut has two more precise replacements, each tied to a clear sizing rule. Both are shown on the same illustrative trader: a 50% win rate, a 1.5 payoff ratio, 5% risk per trade and ruin at a 50% drawdown, where the shortcut returns 1.73%.
- Fixed dollar risk. The bet stays the same dollar size as the account shrinks. The ruin equation for a random walk solves for the decay rate that balances the chance and size of wins against losses, and gives 3.73% for that trader.
- Fixed fraction of equity. The bet is a fixed share of whatever the account is worth, so it shrinks after losses and grows after wins. Run on log returns, the same equation gives 2.04% for that trader.
All three numbers describe one trader. The shortcut is fine for a first look, but here it understates the fixed dollar answer by more than half, and the gap grows when wins are large relative to losses.
Worked Ruin Example on 20 Years of E-mini Futures
A worked ruin study needs a record with a known rule, so every figure here comes from one fully specified rule on E-mini S&P 500 (ES) daily bars. The data is the front-month continuous ES contract from 18 September 2006 to 15 September 2026, 5,031 complete sessions.
The rule, stated in full so anyone can rebuild it:
- Signal. A daily close beyond the highest high or lowest low of the prior 50 sessions.
- Entry. The next session's open, with one tick (0.25 point) of slippage against the position.
- Risk. 2.0 times the 14-session Wilder average true range at the signal bar, which also sets 1R.
- Stop. Checked first each session. A session that opens through the stop fills at the open, not at the stop.
- Exit. The next open after the first close through the opposite 20-session extreme, with one tick of slippage, or after 250 sessions.
- One position at a time.
The rule took 86 trades, from 5 December 2006 to 4 August 2026. It won 33 and lost 53, a 38.37% win rate. The average win was 1.4214R and the average loss 0.8497R, a payoff ratio of 1.6728. The ruin figures below use the unrounded inputs, 33/86 for the win rate and 1.672836 for the payoff ratio, because at these sizes the fourth decimal moves the answers.
Its expectancy was 0.0218R per trade, with a 95% interval from -0.2766R to +0.3202R. In plain terms, 86 trades can't tell this rule's edge apart from zero. That is fine for a ruin study, which is about sequence and size rather than proving an edge. A thin, uncertain edge is also what most real trading records look like.
Four risk of ruin answers to one question
The question: what is the chance this account ever loses half its starting value? It is answered four ways at five risk sizes. The first three columns are eventual probabilities with no time limit. The last column resamples the 86 real trades with replacement, 4,000 paths of 1,000 trades each, with compounding.
| Risk per trade | Two-outcome shortcut | Fixed dollar equation | Fixed fraction equation | Resample within 1,000 trades |
|---|---|---|---|---|
| 0.5% | 1.70% | 4.72% | 2.89% | 0.00% |
| 1% | 13.04% | 21.73% | 23.98% | 5.97% |
| 2% | 36.12% | 46.61% | 69.10% | 38.07% |
| 5% | 66.54% | 73.69% | 100% | 85.50% |
| 10% | 81.57% | 85.84% | 100% | 98.90% |
At 1% risk the shortcut says 13.04% and the fixed fraction equation says 23.98%, 1.84 times as much. At 5% the gap becomes a difference in kind. The shortcut still suggests one in three accounts survives, while the fixed fraction answer is certain ruin, because at that size the account shrinks on average even though the rule makes money per trade.
A separate Monte Carlo run checked the two equations: 20,000 paths of up to 40,000 trades gave 22.13% for fixed dollar and 23.78% for fixed fraction at 1% risk, close to the 21.73% and 23.98% the equations predict.
How Risk per Trade Changes Ruin
Risk per trade is the input that moves risk of ruin most, and the four-model table on the E-mini record shows it plainly. Doubling risk from 1% to 2% took the fixed fraction answer from 23.98% to 69.10%. Halving it to 0.5% took it to 2.89%. Win rate and payoff are hard to change. Size changes with one keystroke.
Risk per trade in contract dollars
Percentages only become decisions when they meet a contract. On the E-mini record, the median stop distance was $2,885.59 per ES contract, with a range from $1,168.48 to $12,502.60 as volatility changed. ES is worth $50 per index point and MES (the Micro E-mini) $5 per point, so the same stop costs $288.56 on one MES.
- At 1% risk, one ES contract on that median stop needs an account of about $288,559.
- At 1% risk, one MES contract needs about $28,856.
- On a smaller account, the honest choices are micros, a tighter stop with a different rule, or skipping the trade.
This is where many ruin problems start, and it rarely feels like a decision at the time. A trader who wants the full ES contract on a $50,000 account is risking 5.77% per trade on that median stop, well past the 3.07% size where this record's fixed fraction ruin reaches 100%. The temptation is to tighten the stop until the dollars fit, which changes the rule and its win rate without anyone testing either.
Fixed dollar sizing vs fixed fraction sizing
The sizing method changes how ruin behaves over time.
- Fixed dollar risk keeps the bet the same while the account falls, so each loss takes a bigger bite of what remains. It also keeps the bet the same after wins, so the floor gets further away in units of risk as the account grows.
- Fixed fraction risk cuts the bet after losses, which slows the fall, but it also raises the bet after wins, so a growing account carries the same percentage danger at every balance.
Neither is free. Fixed fraction sizing makes a total wipeout very unlikely, but a percentage drawdown floor, such as half the account, is just as reachable at any balance. In practice futures traders size in whole contracts, so real sizing sits between the two and steps up or down a contract at a time.
Related lesson: Fixed fractional position sizing
Monte Carlo Ruin Over 100 to 1,000 Trades

A Monte Carlo ruin test shuffles or resamples your own trade results into thousands of possible futures and counts how many of them hit the ruin level. It needs no formula, uses your real win and loss sizes, and can stop at any horizon, which is why most risk of ruin calculators built for traders run one.
The E-mini record was resampled 4,000 times with compounding and read at four horizons. The last column is the eventual answer for the same 86 trade results with no time limit.
| Risk per trade | Ruined by 100 trades | Ruined by 250 trades | Ruined by 500 trades | Ruined by 1,000 trades | Eventual, same trades |
|---|---|---|---|---|---|
| 0.5% | 0.00% | 0.00% | 0.00% | 0.00% | 8.73% |
| 1% | 0.00% | 0.07% | 1.05% | 5.97% | 41.54% |
| 2% | 0.53% | 9.07% | 23.12% | 38.07% | 90.62% |
| 5% | 32.00% | 58.33% | 74.10% | 85.50% | 100% |
| 10% | 72.65% | 89.58% | 95.88% | 98.90% | 100% |
Read the 1% row slowly, because it is the size most traders call safe. Inside 100 trades, not one of 4,000 paths lost half the account. By 1,000 trades only 5.97% had. Yet the eventual figure for the same trades is 41.54%. At 2% risk, 1,000 trades catch 42% of the paths that will eventually be ruined, and the rest are simply waiting their turn.
Here is where traders get caught: short backtests and short evaluation periods. A strategy can carry a large ruin probability and look completely safe for months, because the sequence that sinks it hasn't been dealt yet. When you run or read a Monte Carlo result, the first thing to ask is how many trades it simulated.
Losing Streaks That Reach the Ruin Level
The losing streak is where risk of ruin stops being abstract. A common shortcut says that at 1% risk it takes 50 straight losses to halve an account. That is 1 divided by the risk fraction, and it assumes every loss is taken from the starting balance.
Compounding changes the count. Fifty losses of 1% each leave 60.50% of the account, a 39.50% drawdown. It takes 69 losses to reach a true 50% drawdown, because each loss is taken from a smaller balance than the one before.
| Risk per trade | Losses by the shortcut | Losses with compounding | Account left after shortcut count | Chance of that streak in 250 trades |
|---|---|---|---|---|
| 0.5% | 100 | 139 | 60.58% | Effectively zero |
| 1% | 50 | 69 | 60.50% | Effectively zero |
| 2% | 25 | 35 | 60.35% | 0.0004% |
| 5% | 10 | 14 | 59.87% | 9.97% |
| 10% | 5 | 7 | 59.05% | 97.10% |
The streak chances use the 61.63% loss rate from this guide's E-mini breakout study and assume independent trades. The shortcut understates the required streak by a factor of 1.38 to 1.40 at every size, which is a consistent bias, not rounding.
The streak table also shows why low-risk ruin comes from grinding rather than streaks. At 1% risk, no realistic losing run halves the account, yet the eventual ruin figure is 23.98%. The damage comes from long stretches where losses slightly outnumber what wins give back. At 10% risk the reverse holds: seven straight losses are enough, and the E-mini record itself contains a run of exactly 7.
Drawdown vs Ruin and the Recovery Math
Drawdown and ruin are related but not the same. A drawdown is any fall from a peak, and it is temporary if the account recovers. Ruin is a drawdown deep enough that you stop, by choice or by rule, so there is no recovery from it on that account.
Two probabilities follow from that split.
- Risk of ruin is measured from a fixed level, usually the starting balance. As the account grows, the floor gets further away, so the probability falls with every winning stretch.
- Risk of drawdown is measured from the running peak. The floor rises with the account, so the probability never improves with growth, and it is always higher than risk of ruin for the same depth.
The recovery arithmetic explains why the ruin level deserves care. A loss needs a larger percentage gain to repair it than the loss itself.
| Drawdown | Gain needed to get back to the peak |
|---|---|
| 10% | 11.11% |
| 25% | 33.33% |
| 50% | 100% |
A 50% drawdown needs the account to double just to get back to even. That is why a ruin threshold of half the account is already severe, and why many traders set their personal ruin level well before it.
Related lesson: Recovering from a drawdown
Ruin, Expectancy and the Kelly Fraction
Risk of ruin connects to the other performance metrics through one idea: positive expectancy is necessary but not enough. Expectancy says a rule makes money per trade on average. Risk of ruin says whether the account survives long enough to collect it.
- Expectancy and edge. With zero or negative expectancy, ruin is certain at any size if you trade long enough. This guide's 86-trade E-mini breakout study averaged 0.0218R per trade, positive but thin, which is why its ruin numbers are high.
- Maximum drawdown. A record's worst historical drawdown is one sample of the drawdown distribution. Risk of ruin asks how often drawdowns deeper than your limit occur across all the sequences the same trades could produce.
- The Kelly fraction. The Kelly criterion, described by John Kelly at Bell Labs in 1956, is the bet size that maximises long-run growth. For a two-outcome record it is p minus q divided by b.
On the two-outcome version of the E-mini record, Kelly is 33/86 minus 53/86 ÷ 1.672836, which is 1.5317% of equity per trade. So 1% risk uses 65.29% of the growth-optimal size, and 2% uses 130.57% of it.
Above Kelly, extra size buys more volatility and slower growth, not faster growth. Growth only turns negative at about twice Kelly. For this record that boundary sits at 3.07%, and past it the fixed fraction risk of ruin is 100% at any threshold. That is why, in the same study, 5% and 10% risk per trade give certain ruin under fixed fraction sizing even though the rule makes money per trade.
Many traders use half Kelly or less, because the inputs are estimates and an overestimated edge pushes full Kelly straight into the zone above it. A Kelly fraction computed from 86 trades with an expectancy interval spanning zero is a ceiling to stay well under, not a target.
Related lesson: Kelly criterion for trading
Using a Risk of Ruin Calculator
A risk of ruin calculator takes the four inputs (win rate, payoff ratio, risk per trade, ruin threshold) and returns a probability. Before trusting the output, check three things the calculator should tell you:
- Which model it uses: the two-outcome shortcut, a fixed dollar or fixed fraction equation, or a Monte Carlo simulation.
- What horizon it covers: eventual ruin with no time limit, or ruin within a stated number of trades.
- Which sizing rule it assumes: fixed dollars or a fixed percentage of equity.
If a calculator states none of those, its number can be off by a factor of two. In this guide's E-mini breakout study, the same 1% risk returned 13.04% from the shortcut and 23.98% from the fixed fraction equation.
Running the risk of ruin calculation backwards
The more useful way to use a calculator is in reverse. Pick the ruin probability you are willing to accept, then solve for the risk per trade that delivers it. On the E-mini record (33/86 win rate, 1.672836 payoff, ruin at a 50% drawdown), the fixed fraction equation gives this ladder.
| Target risk of ruin | Risk per trade, fixed fraction | Share of Kelly size | Risk per trade, shortcut | Shortcut overstates safe size by |
|---|---|---|---|---|
| 25% | 1.0200% | 66.59% | 1.4692% | 1.44 times |
| 10% | 0.7076% | 46.19% | 0.8846% | 1.25 times |
| 5% | 0.5745% | 37.50% | 0.6799% | 1.18 times |
| 1% | 0.3998% | 26.10% | 0.4423% | 1.11 times |
There is no universal acceptable risk of ruin. It depends on how bad ruin is for you and how sure you are of the inputs. A trader with a thin, uncertain edge like this one would want the lower rows, and on this record a 1% target means risking 0.3998% per trade, about a quarter of the Kelly size.
Tracking risk of ruin as your record grows
A ruin estimate is only as current as the trades behind it. Recalculate when one of these happens:
- Every 50 to 100 new trades, or monthly if you trade less.
- After any change to the rule, the market, the stop method or the account limit.
- When the live win rate or payoff drifts outside the range your backtest suggested.
- After a drawdown deeper than your backtest's worst, which is evidence the inputs were too optimistic.
Keep the inputs in your journal next to the result, so you can see whether the estimate is improving or decaying.
Related lesson: Trading expectancy
Assumptions and Limits of Ruin Models
Every risk of ruin model rests on assumptions, and each one moves the answer when it fails. The examples use this guide's E-mini breakout study (86 trades on ES daily bars, 2006 to 2026).
- The inputs are known. They are estimates. The E-mini record's expectancy interval runs from -0.2766R to +0.3202R, so its true ruin probability is a wide range, not the point figure a calculator prints.
- Every loss is the same size. Real losses vary. On the E-mini record, 33 stops lost exactly 1R, one session gapped through the stop for -1.4255R, and 19 of 52 channel exits lost an average of -0.5584R.
- Trades are independent. Losses can cluster in regimes, which makes long streaks more common than an independent model predicts.
- One position at a time. Several correlated positions behave like one bigger position, so per-trade risk understates true exposure.
- The rules stay fixed. Account rules, margins and your own behaviour can change during a drawdown, usually for the worse.
Feeding the E-mini record's real loss and win sizes into the fixed fraction equation, instead of the two-outcome average, shows how much this matters. Ruin at a 50% drawdown rises from 2.89% to 8.73% at 0.5% risk, from 23.98% to 41.54% at 1%, and from 69.10% to 90.62% at 2%. Averages in dollars hide the damage that uneven results do to compounding.
Data has limits too. The study uses a continuous front-month series without roll adjustments, which can add small artificial jumps at quarterly rolls. Treat the measured figures as an illustration of method, not a forecast.
Common Mistakes in Ruin Estimates
The same few errors make most ruin estimates look safer than they are.
- No horizon. A figure without a trade count mixes eventual ruin with ruin inside a backtest window. On the E-mini record at 1% risk those are 41.54% and 5.97% for the same trades.
- The wrong sizing model. Using the fixed dollar shortcut for an account that risks a percentage, or the reverse, can halve or double the answer.
- Counting streaks with 1 ÷ f. The shortcut understates the losses needed by about 1.4 times, which leads traders to believe streak risk is their only risk.
- Trusting a small sample. Twenty or thirty trades can't pin down a win rate. A few lucky winners can turn a losing rule into a calculator showing near-zero ruin.
- Ignoring the real ruin level. Defaulting to a zero balance when an account rule, or your own nerve, ends things at a 10% or 20% drawdown.
- Treating the number as precise. A risk of ruin estimate is a planning range. Two decimal places on a calculator don't mean two decimal places of certainty.
How to Reduce Risk of Ruin

The levers that reduce risk of ruin differ a lot in how fast they work, so they are listed from fastest to slowest.
- Cut the risk per trade. It acts on the next trade. In this guide's E-mini study, moving from 2% to 1% cut fixed fraction ruin from 69.10% to 23.98%.
- Lower the ruin threshold's damage, not just its odds. Define a personal stop level well before the real floor, so a bad run ends in a pause and a review, not a closed account.
- Improve the payoff or the win rate. Both reduce ruin, but they take many trades to prove, and a change that looks good in 30 trades often fades.
- Cut correlated exposure. Two positions in closely related markets carry close to double the risk of one.
- Recalculate on schedule. A ruin estimate built on stale inputs is a guess.
A five-minute risk of ruin check on your own record
- Write down your ruin level as a percentage drawdown. If an account rule sets it, use that.
- From at least 50 trades, compute win rate, average win, average loss and payoff ratio.
- Compute the Kelly fraction, p minus q divided by b. Treat it as a ceiling, not a target.
- Divide your actual risk per trade by that ceiling. Over 100% means more risk for less growth, and near 200% growth stops altogether.
- Count the losses that reach your ruin level with compounding: the log of (1 minus the threshold) divided by the log of (1 minus the risk fraction), rounded up.
- If the result is a probability you couldn't sit through, cut the risk per trade first. Size is the one input that changes tonight.
Related lesson: Maximum risk per trade
Ruin Levels on Prop and Funded Accounts With Loss Limits
On an evaluation or funded account, ruin is set by the account's loss limit, and that limit usually sits much closer to the starting balance than half the account. That changes the arithmetic sharply.
Take an illustrative account with a 6% loss limit and a trader risking 1% of the starting balance per trade. Six straight full losses reach the limit. At the 61.63% loss rate from this guide's E-mini breakout study, a run of six or more losses shows up in 99.77% of 250-trade stretches, and a run of 10 or more in 53.22%. A risk size that looks conservative against a 50% ruin level is aggressive against a 6% one.
Trailing limits add another wrinkle. A limit that trails the account's high-water mark is measured from the peak, so it is a risk of drawdown problem, not a risk of ruin problem. Growing the account doesn't move you away from the floor, because the floor follows you up. Size such accounts by dividing the dollar room to the limit by the dollar risk per trade, and keep that count of losses comfortably above the streaks your record produces.
On our accounts, the ruin level is the end-of-day trailing drawdown, enforced in real time against net liquidation value, and touching it fails the account. On a 50K Growth funded account that drawdown is $2,000. On our Sim Funded accounts, the floor locks at $50,100 once the end-of-day balance reaches $52,100, so until the lock the account is in a drawdown problem, and after it a fixed-floor ruin problem. Evaluation drawdowns never lock. Our funded accounts are simulated.
Risk of Ruin FAQ
What is risk of ruin in trading?
Risk of ruin is the probability that a trading account is ever cut to a level where you must stop, such as zero, half the account or an account's loss limit. It depends on win rate, payoff ratio, risk per trade and the ruin level chosen. Risk per trade usually moves it the most.
Is 1% risk per trade safe?
Not automatically, because 1% can still carry a high risk of ruin when the edge is thin. In this guide's E-mini breakout study, 1% risk gave a 23.98% eventual chance of losing half the account under fixed fraction sizing. That is safer than 2% risk at 69.10%, but far from zero.
What is a good risk of ruin?
A good risk of ruin is one you could sit through, measured against a ruin level you would actually stop at. There is no single standard, so go lower when your edge is thin or your sample is small. Then solve for the risk per trade that hits your target.
Can risk of ruin be zero?
Not while you are trading with real risk. Lower size makes it small, never zero, and estimation error in your inputs means even a small figure is uncertain. With zero or negative expectancy, ruin becomes certain over enough trades at any size.
Does risk of ruin change as the account grows?
Under fixed dollar risk it falls as the account grows, because the floor gets further away in units of risk. Under fixed fraction risk, the chance of a given percentage drawdown stays the same at every balance. A floor that trails the peak never improves with growth.
What is the difference between risk of ruin and risk of drawdown?
Risk of ruin is measured from a fixed level, usually the starting balance. Risk of drawdown is measured from the running peak, so it is always higher for the same depth and does not fall as the account grows.
Trading futures and other leveraged products carries substantial risk, and losses can exceed what you planned for when gaps or slippage carry a trade past its stop. The E-mini figures here measure one rule on one historical sample and do not predict future results, evaluation and funded accounts are simulated, and nothing here is investment advice.

