Risk of Ruin Simulator
Simulate independent wins and losses with current-equity sizing. Count how often paths reach a fixed loss threshold below starting capital within your chosen number of trades. This is a conditional model estimate, not a forecast.
More paths reduce sampling noise under the same assumptions; they do not validate the model. At most 10 million trade steps per run.
Fixed initial-capital boundary, not a trailing peak drawdown. Paths stop trading at the first breach.
What Does Ruin Mean Here?
Ruin means the first balance at or below a fixed percentage loss from initial capital. With 10,000 and a 30% threshold, the boundary is 7,000 even if the account previously rose. Paths stop at the first breach and keep that balance on the chart. The result counts breaches at any trade, not just a final loss.
Risking less than 100% of current equity cannot make a positive balance exactly zero in a finite number of this model's trades: each loss multiplies it by (1 − risk). A meaningful positive loss boundary avoids interpreting a mechanically zero bankruptcy frequency as safety.
Reproduce the Position-Sizing Math
For current balance B, risk fraction f, and payoff b: a win gives B × (1 + f × b); a loss gives B × (1 − f). At 2% risk and 1.5 reward/risk, 10,000 becomes 10,300 after a win or 9,800 after a loss. Eighteen consecutive 2% losses cross the 30% initial-capital boundary; seventeen do not. These are arithmetic examples, not recommendations.
- Set a win probability and two net payoff outcomes. Include costs in the assumptions; the simulator does not add commissions or slippage separately.
- Choose a positive starting balance, a loss threshold, whole trade count, and whole simulation count. A run is limited to ten million trade steps for browser responsiveness.
- Run the simulation and compare the fraction of paths that breach. The chart shows only the first 15 sampled paths, not a confidence envelope or the full distribution.
- Repeat with changed probabilities, payoffs, and sizing. Each run draws new random numbers, so results can differ even without changing inputs.
Sampling Error Is Not Model Error
For N independent simulated paths and an estimated hit fraction p, a rough standard error is √[p(1−p)/N]. This approximation is unreliable near zero or one. If zero of 5,000 paths breach, the exact one-sided 95% binomial upper bound is 1 − 0.05^(1/5000), about 0.060%. That bounds sampling uncertainty under the assumed model, not real trading risk.
Trades here are independent with a constant win probability and exactly two outcomes. Correlation, gaps beyond the assumed loss, variable payoffs, margin rules, fees, and changing market regimes are absent. Increasing simulation count does not repair those assumptions.
The formula calculator uses fixed cash risk and an unlimited-horizon diffusion approximation. This simulator uses current-equity risk and a finite horizon, so their percentages need not agree.
Simulation calculations run locally and do not submit these numeric inputs through a simulation API. This statement does not cover ordinary website analytics or linked products. No sign-up or payment is needed for this simulator.
Statistical reference: NIST binomial distribution and exact binomial confidence limits. These justify conditional sampling calculations, not the trading model.
Frequently asked questions
What probability is estimated?
The fraction of simulated paths that first reach a fixed loss threshold below starting capital within the selected trade horizon. Paths stop trading at that breach. It is not exact-zero bankruptcy or peak-to-trough maximum drawdown.
Why use a loss threshold instead of zero?
With risk below 100% of current equity and bounded losses, a finite sequence of trades leaves a mathematically positive balance. Exact zero would therefore give a misleadingly reassuring result. The chosen loss threshold defines a practical stopping boundary for this model.
How many paths are enough?
There is no universal sufficient count. More independent paths reduce Monte Carlo sampling variation, but rare events may remain unobserved. Zero observed breaches does not establish zero probability, and model uncertainty remains regardless of path count.
Does the chart show every simulated path?
No. It displays the first 15 paths from the run. The percentage uses all requested simulations. Re-running draws new random outcomes; the chart is not a percentile band.
Is this a forecast of my account?
No. Independent trades, constant win probability, fixed relative payoffs, and current-equity sizing are assumptions. The tool does not import your trades or model correlations, gaps, margin liquidation, or changing probabilities.
Related Risk Management Tools
Risk of Ruin Calculator
Estimate the probability of reaching a fixed loss threshold using an infinite-horizon diffusion approximation with independent trades and fixed cash risk.
Maximum Drawdown Calculator
Determine the maximum observed loss from a peak to a trough of a portfolio, before a new peak is attained.
Position Size Calculator
Calculate position size from account risk, stop distance, estimated fees and quantity increments. Compare your risk budget with planned loss after rounding.
Win Rate & Profit Factor Calculator
Calculate win rate, profit factor, expectancy, and breakeven point to evaluate your trading strategy profitability.
Compare Assumptions with Observed Trades
Explore Pineify research and strategy tools, then compare historical and executed trades with the assumptions used here. A simulation cannot make live outcomes match a model.