Your real trades against each firm's real rules — trailing drawdown, consistency, daily limits. Funded probability, cause of failure, and the modeled survival band, before the fee leaves your card. Everything runs in this browser tab; nothing leaves your device.
PropSurvival · Prop Firm Survival Analysis
Privacy note: No account. No cloud. Your strategy data stays on your device.
Model version: PropSurvival Engine v1.2 · build v8.1 · Local-only parameter ledger
Current model parameters used in the PDF output. This table is generated automatically from the values selected on screen.
Strategy Durability & Prop Challenge Report
Generated locally from the current assumption set, the Monte Carlo outcome distribution, and risk-optimization diagnostics. No account. No cloud. Strategy data never leaves this device.
| Strategy | My Strategy |
|---|---|
| Scenario | Custom |
| Report date | — |
| Model | PropSurvival Engine v1.2 · build v8.1 |
| Simulation run | Not yet run |
Confidential. This report is built from user-supplied assumptions and simulated outcomes. It is a risk simulation, not investment advice, and it is intended for the strategy owner only.
The composite score weighs expectancy, profit factor, break-even buffer, losing-streak risk, drawdown and recovery load, cost drag, and stress robustness into a single 0–100 reading.
—
—
The simulation replays the strategy day by day across thousands of independent paths. The distribution below — not any single path — is the honest picture of what these assumptions produce.
| Percentile | Final capital | Reading |
|---|---|---|
| Run the simulation to populate the distribution. | ||
Drawdown and ruin diagnostics appear here after the first run.
Return alone does not qualify a strategy; the depth of drawdowns and the time spent below the last peak determine whether the strategy is psychologically and financially carryable.
| Run the simulation to populate risk diagnostics. |
Run the Prop Survival test to populate this section.
The risk optimizer has not been run.
Each row re-scores the strategy with one assumption deliberately broken. Small deltas indicate a robust edge; large negative deltas mark the assumptions this strategy depends on most.
| Stress scenario | Verdict | Score Δ |
|---|---|---|
| Stress results populate after the first analysis. | ||
Every figure in this report derives from the parameters below. They are user-supplied, stored only on this device, and — with the recorded seed — fully reproducible.
| Parameter | Value |
|---|
Outcomes are generated by a day-by-day Monte Carlo engine. Each path draws individual trades from the stated win-rate and R-multiple assumptions, applies per-trade cost, optional volatility scaling, single-day shock events, and the daily loss brake, and sizes risk under the selected fixed or compounding model. Percentiles are read across all completed paths; a fixed seed makes any run exactly reproducible.
Unless the crisis regime is enabled, trades are drawn independently, so real-world autocorrelation and regime persistence may be understated. All inputs are user-supplied assumptions; results inherit their quality. Low path counts widen sampling error — percentile estimates stabilize as the path count increases.
This report is produced by a risk simulation engine and does not constitute investment advice, a recommendation, or an offer to trade any instrument. Simulated performance has inherent limitations: it is built on assumptions, it does not reflect actual trading, and no representation is made that any account will achieve results similar to those shown. Results depend on the quality of the assumptions and of any imported trade data.
All processing is local. No account is required, no data is transmitted to any server, and strategy records remain on this device unless exported by the user. The strategy owner retains sole responsibility for any trading decision.
These are the core assumptions that drive the strategy's headline outcome. Advanced Risk Settings do not replace them; they are a stress and risk-control layer added on top of this base set.
Use these to compare different market and execution conditions with a single tap.
This section extends the core strategy settings. Set your capital, trade frequency, and R assumptions first; then fine-tune additional risk layers here if needed.
R is the unit of loss initially risked on a trade. Target R is calculated automatically from your entry, stop, and target prices.
Annex to the Finding — the machine behind reduce to what, exactly. Both tools optimise the bad case rather than the average one: the P5 Target Solver holds the Desk's parameters fixed and returns the smallest change to one of them that still reaches a modeled 5th-percentile capital target; the Strategy Surface Explorer scans risk × average win R over the same session. The target is a capital multiple, not this firm's rule set — a solve here is modeled arithmetic about the strategy, not a funded outcome for the account.
Holding the active parameters from the Desk fixed, this finds the minimum parameter needed to reach the selected P5 multiple target; all others remain unchanged. The target is read as a multiple of starting capital, so the answer is a modeled floor for the strategy — this firm's daily loss limit, drawdown mode and profit target are not inputs to the solve.
The core surface is fixed: Risk per Trade × Average Win R. A third parameter can optionally be added.
Combinations that fail an enabled constraint are eliminated. When off, the constraint is not directly included in scoring.
Pick a system at the top and configure it in the wide area just below. Results appear below only in the selected analysis tab; the heavy Monte Carlo computation does not run while a slider is moving.
Negative: E_net ≤ 0R · Fragile: PF < 1.30 or Edge Buffer < 5pp or g ≤ 0 · Aggressive: positive edge but P(10 losses) ≥ 25% · Balanced: g > 0, PF ≥ 1.30, Edge Buffer ≥ 5pp · Institutional: PF ≥ 1.80, Edge Buffer ≥ 15pp, P(10 losses) < 15%.
Only the metrics and the active tab update when a slider changes. Monte Carlo paths are refreshed with the manual button, so the browser doesn't lock up while sliding.
Net expected value per trade (E_net) for the active systems, side by side. Bar height shows the size of the mathematical edge; color shows system identity.
Each system's median capital projected at its geometric growth rate (g): where balances end up as trade count grows. Green = above starting capital, red = erosion. This is the median trajectory; individual paths fan out on the Monte Carlo tab.
The decision table reads in one place; metrics are grouped into category blocks, and good/moderate/critical cells are highlighted consistently with the theme.
Shows how expected log growth (g) peaks and then declines as risk per trade rises, for all active systems; the selected system is the bold line, and the black marker is its current risk. Each curve has its own Kelly peak — beyond a point, more risk reduces growth.
The probability of seeing a consecutive losing streak within N trades. Low win-rate systems create a harsher psychological load even when profitable.
Shows the recovery burden of a loss. A 50% loss requires a 100% gain to get back to even.
Refreshed only on this tab, via the button. The P5/P50/P95 band shows how the same system math diverges across different trade orderings.
The net expectancy impact of win-rate and R combinations, computed with the selected system's current cost and loss assumptions.
The curve plots the minimum win rate needed to break even at each R. Active systems are placed as points: points above the curve have positive edge, points below are negative. This compares each system's edge margin on a single surface.
Shows how much the net edge erodes once trading costs are deducted from gross expectancy.
Before you put real money down: under the selected challenge rules, how many Monte Carlo paths does your strategy pass? Daily drawdown, total drawdown, minimum/maximum days, and trade flow are tested simultaneously — producing a GO / NO-GO / REDUCE RISK decision.
If the binding constraint above reads consistency rule: one outsized day is capped as a share of total profit so far, so no single trade can carry a pass — see the consistency-rule mechanic, worked through for the exact arithmetic.
Select a firm above — its rules are read from the firm's own documentation, and this shows which fields could not be reconciled against its own worked examples.
Inside a challenge already? Build your Case File — a forward-from-today assessment from your actual account state, computed on your device.
Firms change drawdown rules without announcements. Get a short email when a rule set you rely on changes — nothing else, ever. Verified stamps live on the changelog.
Runs your current statistics against every firm's real rule set at once. Ranks firms by pass probability, expected attempts, expected fee spend, and modeled first-year return on the challenge fee — so the next $150–600 goes to the firm the math favors.
Splits failed outcomes into daily drawdown, total drawdown, and timeout — showing which rule breaks the strategy.
For a prop trader, "I failed" is not a single answer. Whether the failure came from the daily limit, the total drawdown, or running out of time directly determines the right risk rule.
A band is modeled instead of a single optimum point, because in live trading slippage, daily trade count, and psychology never stay constant.
Pass probability alone is not enough. How many days it takes to reach the target makes both the timeout risk and the minimum-trading-day rule visible at once.
Firm limits (target, drawdowns, day windows) plus your strategy statistics. Fields marked synced are shared with the Strategy Lab — enter a value once and it is used everywhere.
The share of the account lost on a full stop-out. The single most decisive input in the verdict. Dragging this UP makes the verdict worse.
Winning trades ÷ all trades. Use your last 100+ trades, not your best month.
1.8R = winners average 1.8× the amount you risk per trade.
1R = a full stop-loss. Above 1 means you let losers run past the stop — and past 1R this makes the verdict worse, not better.
Your average daily trade count. Drives both speed to target and rule-breach exposure.
The time window to reach the target. Longer windows lower timeout risk.
Every simulated challenge starts with this account size. Firm presets set it to the account you would buy.
The profit the firm requires before it funds you, as a percent of the starting balance.
The account-level loss limit. Trailing firms move this line up with your equity peak — the engine models that.
Count me ruined at this loss of starting capital. Leave unset and the ruin figure asks instead of guessing.
not set — ruin figure will ask
The single-day loss limit. When a firm sets none, this reads 100 and the slider switches off — there is no daily number to choose.
Days the firm requires you to trade before a pass counts.
Commission + slippage per trade, as a fraction of your risk.
Stop opening trades after N losses in a day. 0 disables the brake.
No new trades open once intraday loss reaches this percent. 0 disables the lock.
Firm rules — written by the preset above from the verified corpus. Edit to model a hypothetical rule set.
How many simulated challenge attempts your verdict is computed on. Fast (1,000) previews in seconds, Standard (5,000) is the default depth, Deep (25,000) narrows the standard error on the verdict: it pins down the tail percentiles (p1/p5, risk of ruin) exactly where challenges are lost, and re-runs stay stable to a decimal.
Real losing trades cluster (regime, tilt) — an independent model understates ruin. Independent treats each trade separately; Normal and Clustered raise the chance a loss follows a loss, deepening drawdowns and lowering the pass rate. Modelled as a two-state Markov chain; your win rate is unchanged.
The run's own seed, path count and inputs are recorded in the colophon at the foot of the finding — the desk sets the depth, the document reports what was run.
Simulated exactly as written — your discipline is the model's assumption.
Every risk level from 0.10% to 5.00%, re-run against this account's drawdown rule. The question is not how fast the account grows — it is the level at which the drawdown constraint begins to bind.
The curve is drawn once the scan has run. It plots the modelled probability of a 20% drawdown, and of failure, at each risk level — against the tolerance set below.
| Band | Risk range | Admits a level when |
|---|---|---|
| Conservative | — | — |
| Balanced | — | — |
| Aggressive | — | — |
A band is the set of scanned levels that satisfy its filter — not a recommendation, and not a ranking. Where no level satisfies a filter, the row reads No level holds rather than returning a nearest match.
Scan provenance appears once the band has been computed.
Paths simulated per risk level. Fast sketches the band; Deep pins it down. Depth changes the precision of the band, never the model.
The goal isn't storage — it's finding out whether your real trading matches what you assumed. Save a strategy to compare it against a later one, or import your actual trade history to see if your live edge holds up against the model you built it on. No backend, no account, no cloud. Strategy records are kept in localStorage on this device; JSON and CSV files are read entirely inside the browser.
New to import? Download sample data file gives you a known-good .json in the exact shape the importer accepts — a working example when a file is rejected as “not a PropSurvival data file.”
| Strategy | Health | Verdict | PF | Expectancy | Risk | Primary Risk |
|---|---|---|---|---|---|---|
| Use the Compare button to compare saved strategies. | ||||||
Supported minimum formats: date,symbol,pnl · date,symbol,r_multiple · date,symbol,pnl,capital · date,symbol,entry,exit,size,pnl.
Upload a CSV; the system will analyze sample quality, invalid rows, and field mapping.
This section explains every core concept in the simulation in plain trader language — no statistics degree assumed. The goal isn't just to define terms; it's to show, with examples, how each parameter changes your equity curve, your challenge odds, your psychology, and how you read the report.
Enter the combined impact of commissions, spread, slippage, and fill differences in R terms. This cost is not a fixed percentage of capital; it is deducted from the 1R amount risked on each trade.
This percentage is not position size — it's the amount lost if your stop is hit. It's the main lever that simultaneously drives growth speed, drawdown depth, and critical capital risk.
1R is the planned loss unit accepted before entering a trade. All wins, losses, costs, and shocks are measured on this common scale, so different price levels and stop distances can be compared in the same language.
Break-even means a trade's gross R result closes at roughly 0R. But if cost R isn't zero, that trade isn't truly neutral for you; it produces a small negative result equal to the cost.
This metric is the share of paths whose final capital closes above starting capital. It is not a percentile; it shows the fraction of all simulated paths that end profitable.
Target capital counts as achieved if it is touched at least once during the simulation. MOIC logic shows the ratio of final capital to starting capital; 2.0x means the capital doubled.
Critical Risk here doesn't have to mean the account goes to zero. Touching the user-defined critical capital threshold counts as a serious risk-management warning.
Drawdown measures how far capital has fallen from its peak. Time under water shows the longest stretch spent without making a new peak. One describes the depth of the pain, the other its duration.
This parameter is not daily trade count, market noise, or position size. It sets how widely each trade's realized R outcome scatters around your average win R and average loss R inputs.
This setting stops new trades from opening once the intraday loss threshold is reached. It is not a loss guarantee or a hard cap; a large single loss or a shock can overshoot the threshold.
Shock parameters represent abnormal market conditions. They act as a daily hit separate from the normal trade distribution and primarily affect P5, max drawdown, and critical capital touches.
Risk mode determines which capital base the 1R amount is computed from. This choice changes the strategy's growth curve and its defensive behavior in bad periods.
Sensitivity analysis shows which assumption moves the result the most when changed. Use it to understand which dials the model responds to most — not as a measure of forecast precision.
The seed is the starting key of Monte Carlo randomness. With a fixed seed, the same parameters reproduce the same paths — enabling consistent comparisons for reports, audits, and client presentations.
Expected R shows a trade's average R expectancy. The theoretical value uses only win rate, break-even, average win/loss, and cost. The distribution-adjusted value also accounts for how R volatility widens the win and loss distributions.
Percentiles are not a single forecast — they are different points on the probability distribution. P50 is the median path, P5 the worst-case floor, and P95 the strong but less likely upper band.
This section reveals the path quality behind the headline results and the sensitivity of the worst case. The goal isn't more charts — it's to clarify what drawdown cost a high final capital came with, and how widening the R distribution affects the downside floor.
Final capital, drawdown, losing streaks, and target/critical threshold data will all be read from the same core Monte Carlo paths.
R volatility impact and sensitivity results are translated into an executive note here.
Simulation output reads in three families: final capital range, path comfort, and edge quality — so you see how results relate to each other instead of isolated boxes.
This chart doesn't follow a single simulation path. Each day, all Monte Carlo paths are re-ranked and that day's P5/P50/P95 thresholds are drawn. The P5 line is therefore not 'the same bad path walking through time' — it's the daily distribution's bottom-5% boundary.
Even when the equity fan slopes upward, this chart reveals the peak-to-trough declines experienced along the way.
Core Mode determines the strategy's edge, growth speed, and probability of reaching the target. Risk per Trade states what percentage of capital a 1R loss represents; its nominal equivalent on starting capital is shown live on the control card. More risk can accelerate growth up to a point, but beyond the critical threshold, drawdown, ruin probability, and compounding-loss geometry can deteriorate quickly. Advanced Risk Settings control the impact of bad days and the width of the distribution. Because cost is measured in R, no incorrect commission is deducted from equity without knowing the stop distance.
Compares bad, middle, and good scenario outcomes in a table.
| Percentile | Starting | Final Capital | Net P/L | Total Return |
|---|---|---|---|---|
| Simulation has not run yet. | ||||
Every verified firm, the full survival verdict, unlimited Monte Carlo depth (accuracy is never gated), the complete Firm Fit & Fee ROI ranking, and your run history — all free, and everything stays in this browser.
Now: nothing — no newsletter, no drip sequence. At launch: one email telling you checkout is open. Your address is used for that message and nothing else.
Your trades and simulations stay in this browser. If you hold a license key for a future paid tier, the only thing sent anywhere is the key itself, for activation. Terms · Privacy
Model Commentary