Every Expert Advisor looks good in isolation. The backtest shows a smooth equity curve, a manageable drawdown, and an attractive profit factor. But the moment you run multiple EAs on the same account, reality diverges from the individual backtests — sometimes dramatically.
An EA portfolio simulator answers the question individual backtests can't: what actually happens when these EAs trade together on a shared account?
FXOptimize is a free, web-based EA portfolio simulator that tests every possible combination of your Expert Advisors using true shared-balance simulation. Upload your MT4 or MT5 backtest reports and get accurate portfolio performance in seconds.
EA portfolio simulation is the process of combining multiple Expert Advisor backtests into a single account simulation to see how they perform together. Unlike viewing individual backtests side by side, a proper portfolio simulation accounts for:
The goal is to find the combination of EAs that produces the best risk-adjusted returns — not just the highest profit, but the best balance between returns and drawdown.
Consider three EAs, each backtested individually:
Looking at these numbers, you might expect a portfolio of all three to have roughly 221% annual return with a max drawdown somewhere around 18% (the worst individual drawdown). After all, diversification should help, right?
In reality, the combined portfolio's max drawdown could be 35% — or worse — if the EAs' drawdowns overlap. And the combined return won't be 221% either, because the shared balance means each EA is effectively trading with a smaller portion of the account.
This is why you need a simulator, not a spreadsheet.
Drawdown stacking is the phenomenon where multiple EAs experience losses at the same time, and the combined loss is greater than any individual EA's max drawdown.
It happens more often than most traders expect. Market-moving events — NFP releases, central bank decisions, geopolitical shocks — tend to affect all currency pairs simultaneously. If your EAs trade correlated pairs or use similar entry logic, their drawdowns will cluster.
Here's the math that makes it dangerous: if EA Alpha draws down 15% and EA Beta draws down 12% at the same time on a shared account, the combined drawdown isn't 15% (the max individual). It's approximately 25% (compounded on the same balance). Add a third EA drawing down 10% in the same period, and you could be looking at 40%+.
A proper expert advisor portfolio simulator reveals these stacking events that are invisible when looking at individual backtests.
When you have 8 EAs, there are 255 possible combinations (2⁸ - 1). With 12 EAs, that's 4,095 combinations. Testing each one manually is impractical. This is where Pareto optimization comes in.
Pareto optimization finds the set of Pareto-optimal (or "non-dominated") portfolio combinations. A combination is Pareto-optimal if no other combination is better on all criteria simultaneously.
For example, Portfolio A with 80% return and 15% drawdown is Pareto-optimal if no other portfolio has both higher return AND lower drawdown. Portfolio B with 90% return and 20% drawdown is also Pareto-optimal — it's better on return but worse on drawdown.
FXOptimize automatically calculates the Pareto frontier across all possible EA combinations, showing you the optimal trade-off between return and risk. You choose where on that frontier you want to be based on your risk tolerance.
FXOptimize calculates 17 risk metrics for each portfolio combination. Here's what each one tells you:
Total profit minus total losses. The bottom line, but shouldn't be your only criterion.
Gross profit ÷ gross loss. Above 1.5 is solid. Below 1.2 is fragile. One of the most useful single metrics.
Largest peak-to-trough decline. The number that determines if you can sleep at night (and if your account survives).
Drawdown as a percentage of the peak balance. More useful than absolute drawdown for comparing different account sizes.
Net profit ÷ max drawdown. How many times the portfolio recovered from its worst loss. Higher is better — above 5 is excellent.
Risk-adjusted return using standard deviation. Above 1 is good. Above 2 is excellent. The classic portfolio metric.
Like Sharpe, but only penalizes downside volatility. More relevant for traders who don't mind upside variance.
Percentage of winning trades. Important context: a 40% win rate can be highly profitable if winners are large.
Average profit per trade. The mathematical edge expressed in dollars per trade.
Number of trades in the simulation. More trades = more statistical significance. Below 100 trades, be skeptical.
Ratio of average winning trade to average losing trade. Combined with win rate, tells you about the edge structure.
Longest losing streak. Critical for psychology — can you handle 12 losses in a row?
Compound annual growth rate. Normalizes returns across different backtest periods for fair comparison.
Annual return divided by max drawdown. One of the best single metrics for portfolio quality. Above 2 is strong.
Measures the depth and duration of drawdowns. Lower is better. More nuanced than max drawdown alone.
CAGR ÷ max drawdown over a defined period (typically 3 years). Focuses on recent risk-adjusted performance.
Two EAs can both be profitable individually yet be terrible together — if they're highly correlated. Correlation analysis reveals which EAs actually provide diversification and which are essentially duplicating the same trades.
FXOptimize calculates Pearson correlation on daily returns between every pair of EAs in your portfolio. The result is displayed as a color-coded heatmap:
The best portfolios combine EAs with low correlation to each other. This reduces drawdown stacking because when one EA is losing, the others are more likely to be winning (or at least not losing as much).
Your backtest represents one specific sequence of trades. But what if the trades happened in a different order? What if there were more losing streaks? Monte Carlo simulation answers these questions by:
A robust portfolio should show acceptable performance across most Monte Carlo simulations, not just the historical sequence. If Monte Carlo reveals that 10% of simulations produce 50%+ drawdowns, you know the portfolio is riskier than the backtest suggests.
Unlike tools that simply merge trade lists, FXOptimize replays all trades chronologically on a shared account balance. Every trade's lot size is recalculated based on the actual account state at that moment. This produces the same results you'd see on a live account running all those EAs simultaneously.
The simulation process:
All of this runs in your browser. No data is sent to any server. The simulation typically completes in seconds, even for large portfolios.
An EA portfolio simulator is essential if you:
Upload your MT4 or MT5 backtest reports and get true shared-balance portfolio simulation with correlation analysis, Pareto optimization, and Monte Carlo stress testing. No signup required.
Start Simulating Free →