Learn how to calculate prop firm Phase 1 vs Phase 2 risk of ruin without fake precision. Understand failure-boundary probability, losing streaks, expectancy, fixed vs fractional risk, daily and maximum drawdown, correlation, slippage, Monte Carlo thinking and worked examples.
Akash Mane is the Founder and CEO of Prop Firm Bridge, where he leads the company’s vision, platform growth, and long term strategic direction. He oversees operations across research, marketing, content systems, SEO, and product positioning while driving the platform’s mission of becoming a trusted authority in the prop firm industry. At Prop Firm Bridge, Akash plays a direct role in shaping educational frameworks, comparison systems, and trader focused resources designed to help users make informed decisions with transparency and confidence. His work focuses on building scalable organic growth systems, improving platform authority, and strengthening trust through accurate, structured, and search optimized content. In addition to leadership responsibilities, he actively manages growth strategy, social media marketing, search visibility, and brand development to expand the platform’s reach across global trading audiences.
Manoj Gholap is responsible for content accuracy, compliance, and factual integrity at Prop Firm Bridge. He acts as the final verification layer for all published content, ensuring that prop firm reviews, rules, and comparisons are clear, accurate, and aligned with transparency standards. Manoj plays a key role in maintaining trust and credibility across the platform.
Risk of ruin is one of the most useful trading ideas and one of the easiest to misuse. In ordinary portfolio language, “ruin” can mean losing so much capital that the strategy can no longer continue. In a prop firm evaluation, the more practical definition is different: ruin means hitting an account failure boundary before completing the phase. That boundary might be a daily-loss limit, maximum drawdown, trailing floor or another hard rule defined by the exact evaluation.
This distinction matters because a trader can still have most of the headline account balance visible and yet be mathematically very close to failure. A $100,000 evaluation does not give a trader $100,000 of usable risk capital. The relevant capital is the distance between the current account state and the applicable failure boundary. Position size, losing streaks, open exposure and drawdown mechanics determine how quickly that distance can disappear.
The second important caveat is about probability. There is no honest universal percentage saying a Phase 1 account has a 22% risk of ruin and a Phase 2 account has 14%. Those numbers depend on assumptions about win probability, average win, average loss, trade independence, risk per trade, correlation, execution cost, drawdown rules and the number of opportunities required to reach the target. A small Phase 1 sample is usually not enough to estimate those inputs with high precision.
Quick answer: In a prop firm evaluation, estimate risk of ruin as the chance that your trading path hits a hard failure boundary before the profit objective and other completion conditions are satisfied. Start with usable drawdown, one-R money risk, win probability range, average win/loss in R, losing-streak behavior, total simultaneous exposure and real execution cost. Then stress-test several sequences rather than trusting one exact probability. Phase 2 can have lower practical ruin risk when the target is closer and risk stays controlled, but it can also have equal or higher ruin risk if the trader increases size, opens more correlated exposure or misunderstands different drawdown rules. The phase label alone does not determine the probability.
Written by Akash Mane, Founder and CEO of Prop Firm Bridge. This guide explains risk-of-ruin math in simple language while keeping the assumptions and limitations visible.
Fact checked by Manoj Gholap. Drawdown structures and phase conditions vary by program. All examples below are hypothetical educational calculations and must be rebuilt from the exact current account rules.
If you need a refresher on account-floor mechanics before using these calculations, read the Phase 1 vs. Phase 2 Drawdown Calculations guide. For position-size mechanics, use the Phase 1 to Phase 2 Position Sizing Adjustments guide.
The phrase becomes useful only when “ruin” is defined in the same way the account defines failure. A trader is not calculating personal bankruptcy. They are estimating the chance that the evaluation becomes invalid before completion.
Imagine an account with a headline balance of $100,000 and a hard maximum-loss boundary that leaves only a few thousand dollars of usable room. The trader can lose a small percentage of the headline balance and still fail the account. That means the relevant denominator for risk is not the number displayed in the marketing headline; it is the account’s real distance to the failure floor.
This is why two traders using “0.5% risk” can face very different danger. One may have a large static drawdown cushion. Another may have a much smaller trailing floor. The same $500 planned loss consumes a different fraction of the usable risk budget.
Risk-of-ruin thinking begins by translating every hard account boundary into money and asking how many ordinary losing units can fit before that boundary is reached.
An evaluation does not continue forever. There is usually a profit target or another completion condition. The trading path is therefore moving between two broad outcomes: reach the completion conditions first, or hit a failure condition first.
This makes prop firm risk of ruin similar to a “first passage” problem. Which boundary does the account touch first? A strategy with positive expectancy can still hit the lower boundary first if risk per trade is large or if losses arrive in an unfavorable order. A strategy with modest expectancy can survive long enough to reach the upper boundary when position sizing is conservative.
The order of outcomes matters. That is why average return alone cannot describe evaluation survival.
Suppose a strategy historically wins often enough and has large enough average winners to produce positive expectancy. That is good, but positive expectancy describes the long-run average tendency under the model. It does not prevent short losing sequences.
If the evaluation has a narrow drawdown budget, one ordinary streak can end the account before long-run expectancy has time to appear. A profitable strategy can therefore have an unacceptable evaluation risk when position size is too large relative to the failure boundary.
This is one reason traders should separate “Is the strategy profitable?” from “Is this risk size compatible with this account?” Those are related but different questions.
A highly oversized trade can cross the failure boundary immediately. More commonly, ruin occurs through accumulation: several valid losses, correlated positions, slippage, repeated re-entries or a bad session where the trader keeps trading after the personal stop should have ended the day.
The model should therefore examine single-trade risk and cumulative path risk. A trade that looks safe by itself can become dangerous when another position is already open or when the account has already lost part of its drawdown cushion.
Risk of ruin is a property of the whole account path, not only the next ticket.
A trader rarely knows the true win probability, future payoff distribution or exact correlation of the next trades. Therefore the best use of ruin math is not to produce a number with two decimal places and believe it is scientific certainty.
The better use is sensitivity analysis: what happens if win rate is lower than expected? What happens if slippage is worse? What happens after six losses instead of three? How does ruin risk change when R drops from $500 to $250? What if three correlated positions stop together?
If small changes in assumptions make the account unsafe, the plan is fragile. If the account remains survivable across a realistic range, the plan is more robust.
The point of the calculation is not to make traders afraid to trade. Every valid strategy accepts uncertainty and losses. The purpose is to choose an account risk level where those losses remain ordinary rather than catastrophic.
Once the risk budget is built, a single valid stop should become easier to accept. The trader already knows the account can survive a reasonable sequence. That can reduce revenge trading and target panic.
A good ruin model therefore improves both mathematics and behavior.
Akash's research lens: I define prop firm ruin as “failure boundary touched before completion.” That makes the problem measurable and keeps the headline account size out of the risk calculation.
Book insight: Against the Gods by Peter L. Bernstein is useful because the history of risk management is largely the history of turning uncertainty into measurable choices. Risk-of-ruin thinking does exactly that. Page: varies by edition.
The phase label changes objectives and sometimes rules, but it does not directly tell the trader the probability of failure. The probability emerges from the interaction between rules, target distance and trading behavior.
If Phase 2 requires less net profit than Phase 1 and all other conditions are equal, the account may need fewer favorable outcomes to complete. Fewer required trades can reduce the time exposed to losing sequences. In a simplified model, that can lower the chance of hitting the failure boundary before the target.
However, “all other conditions equal” is a strong assumption. The trader can change risk after Phase 1. Market regime can change. Minimum trading days can force additional activity. Drawdown can be calculated differently. The actual account may therefore behave differently from the simplified expectation.
Do not convert a smaller target into the statement “Phase 2 has lower ruin risk.” It is one factor, not the whole answer.
A trader who increases R after passing Phase 1 can make each loss consume a larger share of usable drawdown. The number of trades needed to reach the target may fall, but the number of losses the account can survive falls too.
For example, if a personal drawdown budget can tolerate twelve $250 losses but only six $500 losses, doubling risk roughly halves the simplified survival depth. The actual probability is more complex, but the direction is clear.
Target distance should never be used as justification for larger position size.
Suppose the profit objective is achieved quickly but the account still needs qualifying days. The trader may have to continue taking some form of rule-compliant activity depending on the exact program. That means the path does not end at the profit target alone.
The effective success boundary becomes “profit target plus all remaining completion conditions.” Every extra required session can add some exposure. A model that stops as soon as the target is touched can underestimate practical failure risk.
This is why minimum days, profitable days and consistency rules must be included separately where they exist.
A static maximum loss behaves differently from a trailing floor. A floor that moves with profits can reduce future room even when the account is above the initial balance. End-of-day and intraday calculations can also create different path behavior.
Two phases with identical targets can therefore have different risk geometry when their boundaries differ. Conversely, two phases with different targets can have similar ruin risk if the same drawdown and conservative risk structure apply.
Always map the actual account floor before comparing phases.
Phase 1 might be completed during a strong trend regime where the strategy’s average winner was large. Phase 2 might begin in a range where false breakouts increase. The win probability and payoff distribution can change.
Using Phase 1 statistics as if they were fixed can underestimate Phase 2 risk. That does not mean the strategy failed; it means the model inputs are conditional on market environment.
A robust calculation uses ranges rather than one optimistic number.
To compare Phase 1 and Phase 2, calculate each stage separately: target distance, failure floor, minimum days, current R, expected payoff range, opportunity frequency and account rules. Then run the same stress scenarios.
The result may show lower, similar or higher practical risk in Phase 2 depending on the account and behavior.
That is more useful than repeating the generic claim that one stage is always safer.
Akash's research lens: I never attach one ruin percentage to the words “Phase 1” or “Phase 2.” I calculate the specific account geometry and the specific risk plan.
Book insight: Thinking in Systems by Donella Meadows is useful because outcomes emerge from interacting parts. Target, drawdown, risk and behavior form one system. Page: varies by edition.
Any probability model is only as good as its inputs. Traders should use conservative ranges and understand where each number comes from.
A trader can finish Phase 1 with a 70% win rate across ten trades. That does not prove the true future win probability is 70%. Small samples move widely. The strategy’s broader historical or forward-tested record is usually more useful.
Use a range such as an optimistic, base and conservative estimate rather than one exact number. The conservative scenario is especially important for ruin analysis because it shows what happens when the strategy performs below expectation.
Phase 1 live data can adjust the model where execution conditions changed, but it should not replace a much larger sample without evidence.
One R is the planned loss if the stop is hit. Expressing average winner in R makes trades comparable across different position sizes. If average winner is 1.5R, the strategy earns one and a half planned risk units on an average winning trade under the sample.
Use realized outcomes after costs when possible. A strategy that theoretically targets 2R but realizes 1.4R because of partial exits and slippage should model the actual distribution.
Average winner is one of the strongest inputs because it determines how many losses a winner can offset.
Many traders assume average loss is exactly -1R. In reality, stops can slip, commissions can add cost and early manual exits can reduce some losses. Use the realized distribution.
If average full-stop loss is -1.07R after cost, the model should reflect that. Small differences matter across long sequences.
A ruin model built on idealized losses can look safer than the account really is.
The trader cannot control whether the next trade wins, but can choose how much account room the stop is allowed to consume. This makes R size the most practical variable in risk-of-ruin management.
Halving money risk does not halve every probability in a simple linear way, but it usually increases the number of ordinary losses the account can survive before reaching a fixed failure floor. That extra survival depth can be very powerful.
Position size should therefore be chosen from drawdown survival before target speed.
A simplified expectancy formula is: win probability multiplied by average win, minus loss probability multiplied by average loss. A positive result suggests a favorable long-run average under the assumptions.
But two strategies with the same expectancy can have very different ruin risk. One can win often with small payoffs; another can lose frequently and rely on rare large winners. Their losing streak distributions differ.
Ruin analysis needs the distribution, not only the average.
Create optimistic, base and stress assumptions. For example, the stress case can use a lower win probability, smaller average winner and slightly larger average loss. Then test whether the Phase 2 risk size still leaves enough survival room.
If the account remains robust only under the optimistic case, risk is too dependent on favorable assumptions.
Strong evaluation risk plans survive being somewhat wrong about the inputs.
Akash's research lens: I never feed one beautiful win-rate number into a ruin formula. I use ranges and ask whether the account survives when the strategy performs worse than expected.
Book insight: Fooled by Randomness by Nassim Nicholas Taleb is useful because small successful samples encourage overconfidence in estimated probabilities. Page: varies by edition.
Losing streaks are the practical bridge between strategy statistics and drawdown risk. Traders should understand them without pretending the future sequence is known.
If a strategy loses 40% of individual trades under a simplified independent model, the chance of four specific trades all losing is 0.4 × 0.4 × 0.4 × 0.4, which equals 2.56%. That number applies to one specific four-trade block under strong assumptions.
Across a long evaluation with many overlapping opportunities, the chance of seeing at least one four-loss streak can be much higher than 2.56%. This is why traders often underestimate streak frequency by calculating only one block.
The practical lesson is to stress-test streaks longer than the trader emotionally expects.
Trade outcomes can cluster because market regimes persist. A breakout strategy can experience several losses during a choppy week. Multiple trades can also share the same macro theme.
When losses are correlated, simple independent coin-flip math understates clustering risk. The strategy can have a good long-run win rate while experiencing periods where the conditional loss probability is much higher.
This is another reason to use stress scenarios rather than one elegant formula.
If the account’s personal total-loss budget equals eight R, a six-loss streak consumes most of the room before costs. If the same budget equals sixteen reduced-risk R, the account has far more flexibility.
Do not interpret this as permission to keep trading mechanically until sixteen losses occur. Personal review thresholds should activate earlier. The count simply shows survival depth.
The trader wants normal historical streaks to fit comfortably inside the account rather than near its edge.
Consider ten trades with five wins and five losses. If losses arrive first, the account can hit a drawdown floor before later wins occur. If wins arrive first, the account may build a buffer.
The total numbers are identical, but the path is different. This is path dependence.
Risk-of-ruin modeling must therefore examine randomized sequences rather than only final expectancy.
A trader who increases size after losses changes the distribution. The second and third losses now cost more than the first. Simple fixed-R streak calculations no longer apply.
This is why recovery sizing is so dangerous. It creates convex downside exactly when the account is already closer to the failure boundary.
Keep risk state prewritten and avoid outcome-driven escalation.
If six valid losses are plausible, the trader should mentally rehearse the account after three, four and six losses. What happens to risk state? When does observation begin? What ends the session?
This preparation reduces surprise. A losing streak becomes an expected possibility rather than proof that the strategy suddenly stopped working.
Mathematical preparation can therefore reduce revenge trading.
Akash's research lens: I use losing-streak math to answer one question: can the account and the trader stay functional when outcomes arrive in the worst ordinary order?
Book insight: The New Trading for a Living by Alexander Elder is useful because position sizing and losing streak survival are central to keeping a trading system alive. Page: varies by edition.
Risk method changes how account exposure behaves as equity moves. Phase 1 and Phase 2 can use the same strategy with different risk wrappers.
Suppose the trader chooses $250 as one normal risk unit. Each valid setup is sized so the technical stop costs approximately $250 before execution differences. This makes losing streak calculations easy because six full losses are roughly $1,500 plus costs.
The advantage is simplicity. The disadvantage is that the same $250 becomes a larger percentage of usable drawdown after the account loses money and a smaller percentage after it gains.
A state-based reduction can solve that problem by lowering fixed-dollar R when the account reaches drawdown thresholds.
Risking a constant percentage of current equity automatically reduces money risk after losses and increases it after gains. In ordinary compounding portfolios, this can be useful.
In prop firm evaluations, the relevant denominator can be more complicated because the headline balance is not the true loss capacity. A fixed percentage of $100,000 can still be too large relative to a small trailing floor.
If percentage risk is used, compare the resulting dollar amount with the real drawdown budget.
A simple state system might use normal R while the account is healthy, reduced R after a personal drawdown threshold and zero live risk in observation or stop mode.
This keeps risk responsive to account survival without changing it after every tiny equity movement.
State-based sizing can be easier to execute psychologically because the trader knows in advance what happens after drawdown.
When the account is close to the profit objective, the value of additional aggressive exposure changes. A prewritten near-target state can reduce R or total simultaneous risk.
This can lower the probability that one normal loss removes a large portion of progress. The exact policy depends on the strategy and rules.
Risk changes at the account layer; technical stops and exits remain market-driven.
Increasing size after losses can make the account recover faster when a winner arrives, but it also makes continued losses increasingly expensive. In an account with a hard drawdown floor, this is especially dangerous.
Even a modest sequence of losses can create very large cumulative damage when size escalates. The strategy’s base win rate does not protect the account from the geometry of increasing exposure.
Recovery should come through future valid outcomes, not through larger bets.
A mathematically sophisticated sizing formula is useless if the trader cannot execute it quickly and correctly. The account needs a risk method that is conservative, auditable and easy to apply to every setup.
For many traders, one normal R, one reduced R and clear state transitions provide enough control without excessive complexity.
Simple risk systems are easier to verify when Phase 2 pressure rises.
Akash's research lens: I prefer risk methods that make the bad path easy to calculate. If I cannot explain the account after six losses, the sizing system is too opaque.
Book insight: The Psychology of Money by Morgan Housel emphasizes room for error. State-based risk is one practical way to build that room into an evaluation. Page: varies by edition.
Daily loss is often the nearest boundary because it can be breached within one session even when the total account drawdown remains healthy.
Use the exact current program formula. Some rules can include closed P&L, floating loss, commissions or day-start reference values differently. Do not use a generic five-percent example unless that is actually the account rule.
Translate the boundary into money. Then calculate current distance after any overnight or open position impact.
The trader should know the hard line before the first new order.
The hard line should be treated as a failure boundary, not a normal risk budget. Create a personal stop inside it, leaving room for slippage, calculation error and unexpected market movement.
If the personal daily stop is -2R, for example, the session ends after the planned threshold even if another A-grade setup appears. The exact number depends on the strategy.
This reduces the chance that one difficult session ends the entire evaluation.
Suppose three positions are open with combined stop risk that exceeds the remaining daily room. Even if current floating P&L looks acceptable, a simultaneous move can cross the boundary.
Calculate worst-planned equity: current account state minus planned losses at every open stop. If that level violates the daily rule or personal buffer, the portfolio is too large.
Daily ruin risk is therefore a portfolio question.
Several trades can lose together during one macro move. Independent-trade assumptions become weak exactly when markets become highly correlated.
Use theme-level exposure caps. If three positions express the same dollar view, treat them as one larger risk cluster.
This can dramatically reduce the chance that one event creates a daily breach.
The longer a trader stays active, the more chances they have to encounter losses, fatigue and lower-quality setups. A strategy may legitimately trade multiple sessions, but target frustration should not create extra hours.
Use a tested time window and personal attempt limit or total daily R cap.
Daily ruin risk is affected by how long the trader keeps giving the market opportunities to take risk.
Near the target, the trader can try several small trades to collect the final amount. The tickets look harmless, but cumulative daily loss can grow quickly if several fail.
A near-target state should usually have lower total risk tolerance than a healthy early-stage account, depending on the strategy.
The objective is to protect completion probability, not maximize daily return.
Akash's research lens: I model daily ruin before total ruin because one bad session can end an otherwise healthy evaluation. Worst-planned equity is the key number.
Book insight: The Checklist Manifesto by Atul Gawande is useful because preventing catastrophic error often depends on simple pre-action checks. Daily-loss room belongs on that checklist. Page: varies by edition.
The maximum-loss rule defines the account’s long-term survival floor, but different drawdown types create different paths to failure.
With a static floor, the hard minimum generally remains fixed relative to the initial account rules. The trader can calculate the dollar distance from current equity to that floor.
This makes R survival depth relatively easy to understand. If the personal budget has $3,000 of room and one normal R is $250, the simplified depth is twelve R before costs.
The trader should still activate review states well before the hard floor.
A trailing floor can rise as the account reaches new highs under the program’s formula. That means profits may not create as much permanent buffer as a trader expects.
A trader can be above the starting balance and still have only limited distance to the current floor. Using the original starting drawdown to calculate risk can therefore be dangerously wrong.
Track the actual current floor after every relevant update.
Some models update floors based on end-of-day values; others can respond intraday or through different reference rules. The timing affects whether floating profit temporarily increases the floor and how much room remains during open trades.
This is why generic drawdown advice can be misleading. The exact formula matters more than the label “trailing.”
Risk-of-ruin calculations should use the account’s operational rule, not a simplified social-media definition.
In a moving-floor account, the sequence of wins and losses can change the boundary itself. A large early winner may move the floor higher, then a later loss has less room before failure than the trader would have had under a static floor.
This makes simple fixed-boundary formulas less accurate. Scenario analysis becomes more important.
Track the floor through each hypothetical trade sequence rather than assuming it stays constant.
Near completion, the account can sit between a close upper target and a relatively close lower failure line. The trader’s R determines how much of that corridor one trade can consume.
A smaller risk unit can increase the number of attempts the account can survive while still allowing valid winners to finish the target.
This is the mathematical reason finish-line risk reduction can make sense for some strategies.
Whenever a trailing floor updates, a large winner changes balance or a drawdown state changes, recompute current usable room. Do not carry yesterday’s R capacity blindly into today.
For a deeper treatment of these mechanics, revisit the drawdown calculations guide.
Risk of ruin is dynamic when the failure boundary moves.
Akash's research lens: In trailing accounts, I never calculate risk from the original floor after the account has moved. The current floor is the only floor that matters.
Book insight: Thinking in Systems by Donella Meadows is useful because feedback changes system boundaries over time. Trailing drawdown is a direct example of a moving feedback rule. Page: varies by edition.
Single-trade probability models often underestimate real account risk because traders hold more than one position or repeat the same idea.
A trader can open EUR/USD, GBP/USD and gold positions that all depend heavily on the same dollar move. Three tickets do not create three independent outcomes.
If the dollar strengthens sharply, all three can lose together. The combined stop risk should therefore be treated as one correlated theme.
Use theme-level caps below the total account exposure limit.
Suppose one trade risks 1R and the personal daily stop is 3R. Three fully correlated positions opened at the same time can theoretically consume the entire daily budget in one market event.
A trader who thinks “I only risk 1R per trade” misses the account-level truth.
Every new position should be checked against total worst-planned loss.
A breakout fails, resets and triggers again. The trader can take three attempts and call each a separate trade. Economically, they may all be one thesis.
Track idea-level cumulative risk. If the maximum price for one idea is 2R, three 1R attempts are not allowed even if each setup looks individually valid.
This prevents one stubborn market opinion from consuming the account.
Markets that appear loosely related during normal conditions can move together during major macro events. This means historical average correlation can underestimate exactly the periods where failure risk is highest.
Use conservative caps during event-heavy or risk-off environments. Do not assume diversification remains constant.
Stress correlation deserves more attention than average correlation.
A momentum setup and a trend setup may look different but both depend on the same market direction. Strategy labels do not create diversification.
Identify the underlying driver: currency, index theme, commodity beta, rate sensitivity or another factor relevant to the instruments.
Portfolio risk should be grouped by what can make the positions lose together.
When simulating scenarios conceptually, include blocks where several correlated positions lose together. Also include re-entry sequences.
This creates a more realistic bad path than assuming every ticket is an independent coin flip.
A robust Phase 2 plan should survive clustered losses as well as isolated ones.
Akash's research lens: I count risk ideas, not just tickets. Correlation and re-entry are where “small” positions quietly become a large ruin event.
Book insight: Against the Gods by Peter L. Bernstein is useful because diversification only works when risks are genuinely different. Correlated trades need to be treated as concentration. Page: varies by edition.
Idealized formulas assume trades close exactly at planned prices. Real evaluation accounts operate with costs and imperfect execution.
A strategy that trades frequently can pay meaningful commission relative to one R. If average gross winner is 1.5R but costs reduce it to 1.35R, expectancy and target speed both change.
Include average commission in realized win and loss distributions. Do not add it only as an afterthought.
Small cost differences can matter when Phase 2 uses many trades to reach a small target.
Wider spread can reduce effective reward and increase how quickly the stop is reached depending on instrument and execution mechanics. Spread can also expand during rollover or major events.
Measure the spread during the actual trading session. Advertised minimum spread is not enough for risk analysis.
A marginal setup can become negative after realistic cost.
If a stop is planned at -1R but often fills at -1.05R or -1.10R during fast markets, a losing streak consumes drawdown faster than the ideal model predicts.
Use Phase 1 live execution data to estimate a conservative buffer. If data is limited, use a stress scenario.
Ruin calculations should prefer slightly pessimistic execution assumptions.
Stops are instructions, not guarantees of exact price. Sudden gaps or illiquid moves can fill beyond the planned level. This matters for overnight, weekend or event exposure where the account permits it.
Do not model every trade as capped perfectly at one R if the strategy routinely faces gap risk.
Use smaller size or avoid specific conditions where the potential excess loss would threaten the account.
Technical problems can delay exits or create duplicate orders. The probability may be low, but the consequence can be large in a tight evaluation.
Use protective stops, tested order routines and a contingency plan. Know how to contact official support if an execution dispute occurs.
Operational risk is part of ruin risk even though it is not generated by the strategy signal.
If average losing trade is -1.06R after all costs, stress-test six losses as roughly -6.36R rather than -6R. If event losses occasionally reach -1.3R, include those tail cases in scenario analysis.
This makes the model less elegant but more realistic.
Evaluation survival benefits from realism more than from clean formulas.
Akash's research lens: I model the account that actually trades, not the frictionless account in a spreadsheet. Costs and imperfect fills consume real drawdown.
Book insight: The Black Swan by Nassim Nicholas Taleb is useful because rare large deviations can matter more than average behavior. Gap and slippage risk deserve explicit room. Page: varies by edition.
Monte Carlo analysis sounds technical, but the central idea is simple: instead of testing one expected sequence, test many possible sequences drawn from the strategy assumptions.
A spreadsheet might say a strategy earns 0.2R per trade on average, so twenty-five trades should earn 5R. Real trades do not arrive as a smooth +0.2R line. They arrive as wins and losses in irregular order.
One sequence can hit the target quickly. Another with the same long-run expectancy can hit drawdown first.
Monte Carlo thinking focuses on the distribution of paths, not only the average endpoint.
A simple model can represent each trade as a win or loss drawn from assumed probabilities and payoff sizes. A richer model can sample from actual historical R outcomes.
Then the outcomes are shuffled or resampled many times. Each path tracks target progress and failure boundaries.
The percentage of paths that hit the lower boundary first becomes an estimated ruin probability under the model.
Because the true future win rate is unknown, run multiple scenarios: base win probability, lower win probability, worse average winner, larger slippage and higher correlation.
If ruin risk remains low across the stress scenarios, the risk plan is more robust. If it jumps sharply when win rate falls five percentage points, the account depends heavily on optimistic assumptions.
Robustness is more important than one precise estimate.
A realistic prop evaluation simulation should stop a path when the daily loss or maximum drawdown is breached. It should also account for trailing floors where possible, minimum days and any other completion condition that materially changes when the path ends.
This is more complex than a normal bankroll ruin formula, but it matches the actual evaluation problem.
The goal is not to build perfect software; it is to think in terms of path-specific rules.
If a model says 8% of simulated paths fail, that means 8% under the exact assumptions and sampled distribution—not that the trader has an objective 8% future failure probability.
Change the inputs and the estimate changes. Real markets can also produce outcomes outside the historical sample.
Treat the result as a decision aid and sensitivity tool.
Run the same model at 1R = $500, $250 and $125. Observe how the proportion of failing paths changes. Also examine median drawdown, worst decile drawdown and number of trades to target.
This can show the trade-off between speed and survival more clearly than a single formula.
Often the lesson is simple: slightly lower risk can dramatically improve survival while only moderately increasing expected time to completion.
Akash's research lens: Monte Carlo thinking reminds me that the account can experience many valid paths. I care less about the average path and more about whether bad-but-plausible paths survive.
Book insight: Fooled by Randomness by Nassim Nicholas Taleb is useful because alternative possible histories matter. Monte Carlo methods are one practical way to visualize those alternative histories. Page: varies by edition.
Phase 1 gives live information, but the sample should be blended with broader evidence rather than treated as the entire probability model.
The strongest Phase 1 data often concerns live spread, slippage, commission, actual stop loss and platform behavior. These variables are directly relevant to Phase 2.
If planned -1R stops realized at -1.06R on average, update the risk model. If one session consistently produced poor fills, include that cost.
Execution evidence is less likely to be overfit than a tiny win-rate sample because it describes the operating environment.
Compare Phase 1 win rate with the larger historical strategy record and current market regime. If the first stage was unusually strong, use a more conservative Phase 2 range.
If Phase 1 was unusually weak but the process remained valid, do not assume the lower win rate will persist either.
The correct model acknowledges uncertainty in both directions.
If the first stage’s worst streak was four losses, the second stage should be able to survive more than four if broader data says longer streaks are plausible.
A passing sample may simply have avoided the worst ordinary sequence.
Risk should be designed for plausible adversity, not for replaying the pass.
Did the trader increase size after wins? Take more trades after losses? Skip setups near the target? These behaviors change the effective strategy distribution.
Phase 2 should remove the behavioral errors before assuming the same statistical inputs.
The trader is part of the system being modeled.
If Phase 1 occurred in trend expansion and Phase 2 begins in range compression, use historical data from similar range conditions instead of copying the first-stage statistics.
This makes the probability assumptions conditional on current market state.
Regime-aware inputs are usually more honest than one all-time average.
Create a base, conservative and severe scenario. Record the estimated number of R losses the account can survive, the target distance in R and the likely trade count range.
The trader does not need a perfect “success probability.” The range can already show whether current risk is fragile.
If severe-but-plausible assumptions still leave substantial survival room, the plan is more resilient.
Akash's research lens: Phase 1 is strongest as an execution and behavior sample. I refuse to let ten or twenty trades become a fake precise forecast of Phase 2.
Book insight: Thinking, Fast and Slow by Daniel Kahneman is useful because small samples create strong stories. Probability estimates need deliberate resistance to that instinct. Page: varies by edition.
The final framework turns all the theory into a repeatable process that can be used before and during either evaluation stage.
Write the daily-loss formula, maximum-loss formula, trailing-floor rule, server reset and any other hard breach condition. Convert the current boundaries into money.
Do not use the headline account size as usable risk capital.
Risk begins with distance to failure.
Write the profit target plus minimum days, profitable days, consistency or other completion conditions where applicable.
The path is complete only when all required conditions are satisfied.
This prevents models from ending too early at the profit target alone.
Use ranges for win probability, average winner, average loss, execution cost and opportunity frequency.
Use broader historical data and current regime, with Phase 1 live data as an adjustment.
No single input set should be treated as truth.
Calculate the number of R units between current account state and the personal review line. Stress-test losing streaks.
Use the position-sizing guide to convert technical stops into units.
Risk should be small enough that plausible streaks fit comfortably.
Calculate maximum total open risk and personal daily stop. Include correlation and simultaneous stops.
Ask what happens if every open position reaches the stop before the next winner arrives.
This prevents one session from becoming the ruin event.
Track static or moving maximum-loss floors through wins and losses. Recalculate after material account changes.
For trailing rules, update the floor inside each scenario.
Path-dependent boundaries require dynamic risk awareness.
Use realized commission, spread, slippage and occasional gap assumptions.
Stress-test worse-than-average execution.
Do not assume every -1R stop is perfect.
Group correlated positions and re-entries. Model occasional blocks where several trades lose together.
This makes the scenario more realistic than independent-ticket math.
Portfolio risk controls the account.
Use Monte Carlo software if you have a reliable tool, or conceptually create multiple shuffled sequences from historical R outcomes.
Track how often target or failure is reached first under different risk sizes.
Focus on sensitivity rather than one exact probability.
Repeat the framework using each phase’s actual target, rules and current market inputs.
Do not assume Phase 2 is safer because the target is smaller or riskier because funding is closer.
Let the account geometry answer.
If the stress scenarios show excessive failure probability, lower R, reduce simultaneous exposure or improve the account-rule fit.
Do not use the model to decide whether the next setup will win.
Risk models govern exposure, not direction.
Write the assumptions next to every estimated probability. Update them when regime, execution or behavior changes.
Never market a simulation output to yourself as certainty.
The best result of risk-of-ruin work is a more survivable account, not a beautiful percentage.
Akash's research lens: My complete ruin framework asks how many bad-but-plausible paths the account can survive. If the answer depends on perfect assumptions, risk is too large.
Book insight: Against the Gods by Peter L. Bernstein is useful because risk management is not about eliminating uncertainty; it is about making better decisions while uncertainty remains. Page: varies by edition.
It means the probability that the account hits a hard failure condition such as daily loss or maximum drawdown before all completion conditions are satisfied. It does not mean literal personal bankruptcy.
No. A smaller target can reduce the distance to completion, but risk per trade, drawdown rules, minimum days, market regime, correlation and trader behavior can offset that advantage.
You can estimate model-based ranges, but not know the exact future probability. The true win rate, payoff distribution, future market regime and trade correlation are uncertain. Treat simulations as decision aids, not certainty.
Risk per trade and total simultaneous exposure are among the strongest controllable variables. Lower exposure generally increases the number of ordinary losses the account can survive.
Use the strategy’s broader historical losing-streak data plus a safety margin rather than the worst streak from one successful Phase 1 sample. The account should survive a plausible unfavorable sequence comfortably.
No. A positive-expectancy strategy can still hit the drawdown boundary first when losses arrive early or risk per trade is too large.
Correlated positions can lose together, making several small tickets behave like one large trade. Independent-trade probability models can underestimate this clustered risk.
Either can work when it is tied to real drawdown survival. Many evaluation traders benefit from a simple state-based framework with normal and reduced R because the usable risk budget is smaller than the headline account balance.
It means testing many possible orders of wins and losses instead of assuming one smooth average path. It helps show how often a strategy might hit the target or failure boundary first under specific assumptions.
Use conservative input ranges, model hard account rules, include costs and correlation, test bad-but-plausible streaks, compare several risk sizes and choose exposure that remains survivable even when assumptions are somewhat wrong.
Final takeaway: Risk of ruin is not a magic formula that predicts whether Phase 1 or Phase 2 will pass. It is a framework for understanding how quickly ordinary uncertainty can consume the account’s real loss budget. The most important insight is simple: the headline account balance is not your risk capital. The distance to the failure boundary is. Once that distance is expressed in R, losing streaks, correlation, costs and phase targets become easier to compare. The trader cannot control the next outcome, but can control how much damage an unfavorable path is allowed to create.
Prop Firm Bridge’s Evaluation Mastery Center is designed to help traders turn those calculations into practical risk states so a statistically normal losing sequence does not become an avoidable evaluation failure.
It is the model-based probability that the account hits a hard failure condition such as daily loss or maximum drawdown before all completion conditions are satisfied.
No. A smaller target can reduce distance to completion, but risk per trade, drawdown rules, minimum days, correlation, market regime and behavior can change the result.
You can estimate ranges under stated assumptions, but the exact future probability is unknowable because future win rate, payoff, market regime and correlation are uncertain.
Risk per trade and total simultaneous exposure are among the strongest controllable variables because they determine how quickly drawdown room is consumed.
Use broader historical losing-streak data plus a safety margin rather than relying only on the streak observed in one Phase 1 sample.
No. A positive-expectancy strategy can still hit a hard drawdown boundary first when losses arrive early or risk is too large.
Correlated trades can lose together, making several small positions behave like one large account event and invalidating simple independence assumptions.
Either can work when tied to real drawdown survival. A simple state-based framework with normal and reduced R can be practical in evaluations.
It tests many possible orders of wins and losses instead of assuming one smooth average path, helping estimate how often target or failure boundaries are reached first under a model.
Use conservative input ranges, include account rules, costs and correlation, stress-test losing streaks and choose exposure that remains survivable when assumptions are somewhat wrong.
