Running several EAs inside one funded challenge is not a diversification play by default — it is a budgeting problem. You have a single daily drawdown limit, and every strategy is spending from the same account. The question is not "how much risk per EA" but "how do I divide one shared budget so the whole book survives its worst day."
Start from the budget, not the strategy
The account has one daily drawdown limit and one total limit. That daily number is your budget for the day, and it belongs to the portfolio, not to any single EA. The mistake is to size each strategy in isolation — "1% feels reasonable for this one" — and then run four of them, discovering only in a bad session that four reasonable numbers add up to an unreasonable day. Size the book first, then hand each EA a slice.
Your risk budget is the daily drawdown limit, minus a safety margin, divided among strategies by how much they can lose at the same time — not divided equally by default.
A worked split
Take a $100,000 account with a 5% daily drawdown, so a $5,000 daily budget. Never spend the whole thing: leave a margin for slippage, spread widening and the odd extra signal, and treat roughly 4% ($4,000) as usable. If you run four EAs and split equally, each gets $1,000 of worst-case daily loss — that is 1% per strategy. So far this is just arithmetic.
The arithmetic breaks the moment the four EAs are not independent. If they trade the same session or the same currency and tend to lose together, four simultaneous $1,000 losses is a $4,000 day — the full budget, gone at once, on a day that felt ordinary. Equal splitting only protects you when the strategies rarely lose on the same day.
Correlation decides the real slice
Independent strategies almost never all lose at once, so their combined worst case grows roughly with the square root of their number: four independent EAs at $1,000 each have a realistic bad-day loss closer to $2,000 than $4,000. Fully correlated strategies are effectively one big position — their worst case is the full $4,000. Real EAs sit between these poles, and where they sit determines how much you can safely give each one:
- Genuinely independent (different instruments, sessions and logic): you can size each slice up, because the aggregate stays well under budget.
- Partly correlated (same asset class, overlapping hours): keep meaningful headroom; the combined loss is larger than the square-root rule suggests.
- Highly correlated (same pair, same signal family): treat them as one strategy for budgeting and size as if a single loss hits all of them.
Common mistakes
- Equal split by headcount. Dividing the budget by the number of EAs assumes they are independent. If they are not, you have quietly sized to a worst case far above your limit.
- Counting diversification you do not have. Two EAs on EURUSD and GBPUSD are not two bets; in a dollar move they are close to one.
- Ignoring concurrency. An EA that can hold three positions at once is spending three slices, not one. Budget by maximum simultaneous exposure, not by the nominal per-trade risk.
- Spending the full budget. Sizing to exactly 5% leaves nothing for slippage or an extra signal, so a normal day tips over the edge.
A budgeting checklist
- Set the daily budget to the drawdown limit minus a margin (aim to use around 80%).
- Group EAs that tend to lose together and treat each group as one line item.
- Estimate how correlated the losses really are, honestly, from their history.
- Divide the budget so the combined worst case, at that correlation, stays inside it.
- Check maximum simultaneous positions per EA, not just per-trade risk.
- Re-check after adding any strategy — one new EA changes every other slice.
Sizing it without guessing
Turning "these EAs are somewhat correlated" into a concrete per-strategy number is what the Per-Strategy Risk Allocator below is for. Enter your account balance, the daily drawdown budget, how many strategies you run and a correlation estimate, and it returns the risk each strategy can take, the combined worst case at that correlation, and how much headroom is left against the budget — so you can see directly how much lower correlation lets you safely size each EA up, and how fast correlated strategies eat the whole limit. This is the split; the drawdown rules it must respect, and the danger of assuming equal fixed risk is safe, are covered in the companion guides.
Takeaway
Budget the book, not the bot. One shared daily limit, sliced by real correlation with a margin kept back, is what keeps a multi-EA challenge alive on its worst day.
