Monte Carlo Equity Simulation: Stress-Testing Your Trading Strategy

What Is Monte Carlo Simulation in Trading?

Monte Carlo simulation is a statistical technique that generates thousands of possible future equity paths for a trading strategy by randomly sampling from its historical outcome distribution. Instead of asking "what did happen?", it asks "what could happen?" across the full range of plausible sequences.

The mechanics: take your strategy's real win/loss results (or your assumed win rate and R:R), then run 1,000–10,000 simulated sequences of those outcomes. Each simulation represents a possible "parallel universe" of how your next 100 (or 500, or 1,000) trades could unfold, given that the underlying edge is the same but the order of wins and losses varies randomly. The result is not a single equity curve, but a distribution of equity curves — showing best, worst, and median cases simultaneously.

This is the right tool to use before committing real capital to any strategy, because no backtested equity curve tells you what would have happened with a different sequence of the same trades. The sequence matters, and Monte Carlo shows you the full range of what sequences can produce.

Why Backtesting Alone Is Insufficient

A backtest produces exactly one equity curve — the one that happened in historical order. This single curve has several problems as a forward-looking performance estimate:

• Sequence risk: If your 20 worst losses happened at the start of the backtest instead of being spread throughout, would your account have survived? The backtest cannot tell you — it only shows you the historical sequence.
• Selection bias: Backtests are typically run on strategies that were selected because they performed well historically. This inflates expectations compared to forward performance.
• Single path optimism: The historical sequence may have been unusually favorable for your strategy. Monte Carlo reveals how wide the range of plausible outcomes actually is.
• No ruin probability: A backtest says nothing about the probability of ruin. Monte Carlo quantifies it directly as the percentage of simulated paths that end below 50% (or whatever ruin threshold you set) of starting capital.

Monte Carlo does not replace backtesting — it augments it. Use the backtest to validate that your edge exists. Use Monte Carlo to understand the risk profile of trading that edge at your chosen position size.

Key Statistics from a Monte Carlo Simulation

A well-designed Monte Carlo output gives you several actionable statistics:

• Median equity (50th percentile): The "average" outcome — the level that 50% of simulations end above and 50% end below. This is a better central estimate than the single backtested equity curve.
• 5th percentile equity: The downside scenario — what the worst 5% of traders following this strategy and position size would experience. If this is still above zero, your strategy survives even in bad luck scenarios.
• 95th percentile equity: The upside scenario — what the best 5% of traders would experience. The gap between 5th and 95th percentile shows how much of the outcome is determined by luck versus edge.
• Median maximum drawdown: The worst peak-to-trough decline experienced in the typical simulation. This is usually significantly larger than the maximum drawdown in your single backtested equity curve.
• Ruin probability: The percentage of simulated paths where final equity falls below the ruin threshold (typically 50% of starting capital). This is the headline number for risk assessment.

How to Read the Equity Distribution Chart

The equity distribution chart shows a histogram of final equity values across all simulations after the specified number of trades. Reading it correctly:

• Green bars represent simulations ending above the starting capital — profitable outcomes.
• Red bars represent simulations ending below the starting capital — losing outcomes.
• A tight, narrow distribution centered well above starting capital indicates a robust, consistent strategy with high edge. Even bad luck is manageable.
• A wide, spread distribution indicates high variance. Even if the median is positive, the tails are extreme — some paths end very well, some end in ruin.
• A distribution skewed left (heavy left tail) is a warning sign — your downside scenarios are worse than your upside scenarios are good, suggesting position sizing is too aggressive for your edge.

The shape of the distribution tells you as much as the individual statistics. A clean, compact right-skewed distribution is the goal.

The Effect of Sample Size (Number of Trades)

The law of large numbers states that as sample size increases, the distribution of average outcomes converges toward the true expected value. In trading terms: more trades means more consistent results across simulations, reducing the influence of luck on your final equity.

With only 50 trades simulated, the distribution of final equity values is very wide — a few losses in a row can dominate the result. With 500 trades, the distribution tightens significantly. With 2,000 trades, if you have genuine edge, virtually all simulations are profitable.

This has a practical implication: low-frequency traders (5–10 trades per month) need to hold positions longer to allow their edge to express itself statistically. High-frequency traders see the law of large numbers work in their favor faster. If your sample size is small, your Monte Carlo confidence interval will be wide — that is not a bug, it is an honest reflection of statistical uncertainty.

Effect of Risk Per Trade: 1% vs 3%

Position size has a non-linear effect on the equity distribution. Consider a strategy with 55% win rate and 1.5R average win:

• At 1% risk per trade over 200 trades: tight distribution, ruin probability near zero, median drawdown around 8–12%, median equity approximately 1.5–2× starting capital.
• At 3% risk per trade over 200 trades: distribution spreads significantly. Median equity is higher (compounding works faster), but the 5th percentile drops sharply. Ruin probability may climb to 3–8%. Median max drawdown expands to 25–35%.

The tradeoff is not symmetric. The upside from higher risk scales linearly at best; the downside expands much faster due to the asymmetry of percentage losses and gains. At 3% risk, you need a larger gain percentage to recover from the same percentage drawdown.

Win Rate vs R:R: What the Simulation Reveals

Two strategies can have identical expectancy but very different Monte Carlo profiles:

• A 40% win rate with 3.0R reward has expectancy = (0.40 × 3.0) - (0.60 × 1.0) = 0.60R per trade. But its equity curve is choppy — long losing streaks are statistically normal, maximum drawdown is large, and many traders quit before the edge pays out.
• A 60% win rate with 1.0R reward has the same expectancy = 0.60R per trade. Its equity curve is smoother, drawdowns are shallower, and the 5th percentile outcome is much better — despite identical long-run expected returns.

Monte Carlo makes this concrete: run both strategies with the same position size and compare the ruin probability, median max drawdown, and 5th percentile equity. The smoother strategy almost always wins on risk metrics even when expectancy is the same.

Practical Uses Before Going Live

Monte Carlo simulation has three direct applications for traders:

• Pre-live validation: Before trading any strategy with real money, run Monte Carlo with your planned position size. If ruin probability exceeds 5%, reduce position size until it falls below 2%.
• Stress-testing losing streaks: Ask "what happens if I get 20 consecutive losses?" Set all simulations to start with a 20-loss streak. If median equity still recovers, your system is robust. If it does not, your position size is too aggressive for your win rate.
• Position sizing calibration: Run the simulator at 0.5%, 1%, 1.5%, 2%, and 3% risk levels. Find the highest risk level where ruin probability stays below 2%. That is your optimal position size — maximum growth consistent with survival.