Portfolio Strategy · Guide

How Many EAs Should You Run on One Account?

Q: How many Expert Advisors can I run on one account?

Technically as many as your margin allows, but the useful number is far smaller. For most retail accounts, 3 to 6 genuinely uncorrelated EAs capture nearly all of the diversification benefit. Beyond that you get diminishing returns unless each additional EA is close to uncorrelated with the ones you already run.

Q: Does running more EAs reduce my risk?

Only if the EAs are uncorrelated. Adding an EA that is highly correlated with your existing ones simply increases your effective position size in the same trade, which raises risk rather than diversifying it. Diversification comes from low or negative correlation, not from the number of EAs.

Q: How do I know when I have added too many EAs?

When the next EA stops improving the portfolio efficient frontier — that is, when adding it no longer increases return for a given drawdown, or pushes drawdown past your budget. Simulating every combination on one shared balance shows exactly where that point is.

There is no universal number. The right number of Expert Advisors is the point where adding the next one stops improving your risk-adjusted return — and that depends on correlation, your drawdown budget, and margin, not on a round figure.

June 14, 2026Portfolio Strategy7 min read

It is the most common question among traders building an EA stack, and the honest answer frustrates people: there is no fixed number. Three uncorrelated EAs can be a better portfolio than ten correlated ones. The useful way to think about it is marginal: keep adding EAs only while each new one improves your return for the drawdown you are willing to take. The moment it doesn't, you're done — and that point arrives sooner than most traders expect.

By the FXOptimize research team. Last updated 2026-06-14.

The myth

Why “more EAs = more diversification” is wrong

Diversification does not come from the number of EAs. It comes from how differently they behave. Two profitable EAs that are 0.8 correlated are, for practical purposes, the same strategy run at double size — they win together and, more importantly, they lose together. Stacking five of those is not a diversified portfolio; it is one trade with five times the heat.

Real diversification comes from EAs whose returns are near-zero or negatively correlated. One such pairing can do more for your drawdown profile than five correlated additions — even if one of the uncorrelated EAs looks “worse” on its own backtest. The count is a side effect; correlation is the cause.

The ceiling

The three constraints that cap your number

1. Correlation saturation. Each new EA only helps if it is uncorrelated with the basket you already run. Once your candidates start overlapping with existing EAs, additional ones add risk and cost without adding diversification.
2. Your drawdown budget. Because drawdowns stack when EAs lose together, every EA you add raises the portfolio's worst-case excursion. If you're trading a prop-firm challenge with a hard daily or total loss limit, that budget — not your ambition — sets the ceiling.
3. Margin and execution. More EAs means more simultaneous open positions, more margin consumed, and more exposure to a single volatile session. The account has to survive the moment when several EAs are all in drawdown at once.

These interact. You cannot reason about them one EA at a time — they only show up in the combined equity curve, which is why the number can only be found by simulating the portfolio, not the parts.

The method

How to find your number

The reliable way to settle the question is to stop guessing and let the combinations answer it:

Backtest each candidate EA and export its MT4/MT5 report.
Simulate every combination on one shared balance — with dynamic lot scaling and chronological replay, so the interactions are real.
Plot the results on the efficient frontier (return vs. drawdown). Watch where adding the next EA stops pushing the frontier outward. That inflection point is your number.

Free tool

FXOptimize finds that point for you. Upload your EA backtests and it simulates every combination on one shared account, then shows the Pareto-optimal mixes with a correlation heatmap — so you can see exactly which EAs diversify and where adding more stops helping. Free to start, runs in your browser, and your files never leave your machine.

Find your number →

Rule of thumb

A practical starting point

If you want a number to anchor on before you run the analysis: for most retail accounts, 3 to 6 genuinely uncorrelated EAs capture the large majority of the available diversification benefit. Beyond six, returns diminish quickly unless each new EA is close to uncorrelated with everything already in the book. Fewer, well-chosen EAs almost always beats more, loosely-chosen ones — and the correlation matrix tells you which is which.

FAQ

Frequently asked questions

How many Expert Advisors can I run on one account?
As many as margin allows, but the useful number is far smaller. For most accounts, 3–6 genuinely uncorrelated EAs capture nearly all of the diversification benefit; beyond that you get diminishing returns unless each addition is close to uncorrelated.

Does running more EAs reduce my risk?
Only if they are uncorrelated. Adding a highly correlated EA increases your effective size in the same trade — that raises risk rather than diversifying it. Diversification comes from low or negative correlation, not from the count.

How do I know when I have added too many EAs?
When the next EA stops improving the efficient frontier — no longer adding return for a given drawdown, or pushing drawdown past your budget. Simulating every combination on one shared balance shows exactly where that point is.

Keep reading