How Much Capital an Automated Trading System Really Needs

Around €1,000 is the floor for running an automated forex system on your own account, and between €5,000 and €10,000 is where the position sizing stops distorting the strategy. Those numbers do not come from a marketing page. They come from the minimum lot size, which is the constraint that decides everything below about €10,000.

This page works through the arithmetic, shows what actually breaks at each account size, and covers the alternative: renting notional capital through a prop firm evaluation instead of posting your own.

• The binding constraint is the 0.01 lot minimum, not the software and not the broker.
• Below the floor you are not running the strategy, you are running a rounded-off approximation.
• Costs per round turn matter more at small sizes, because they are fixed while the position is not.
• A prop firm evaluation converts a capital requirement into a fee plus a rule set.
• Cent accounts move the decimal point, they do not solve the problem.

The arithmetic that sets the floor

On a standard forex account the smallest position is 0.01 lots, which on most major pairs is about 1 unit of account currency per pip. That single fact sets the minimum viable account size once you fix a risk-per-trade percentage and a stop distance.

The formula is short:

minimum account = (stop in pips × value per pip at 0.01 lots) / risk per trade

example: 30 pip stop, ~1 per pip, 1% risk
         (30 × 1) / 0.01 = 3,000

So a strategy with a 30 pip stop needs roughly 3,000 units of account currency before a single minimum-size position sits at 1 percent risk. Below that, taking the trade at all means taking more risk than the strategy specifies.

Gold deserves its own line because the contract size and typical stop distance are both larger, so an XAUUSD system needs meaningfully more capital than a major-pair system at the same risk percentage.

What breaks at each size

Under €1,000

Every trade is 0.01 lots because nothing else is available, so position sizing does not exist. Risk per trade swings with the stop distance instead of being held constant, which is the opposite of what the strategy is doing. A 15 pip stop and a 60 pip stop carry the same lot size and four times the difference in risk. The system's results will not resemble its test results, and the reason is arithmetic rather than the market.

€1,000 to €5,000

Sizing exists but is lumpy. The calculation says 0.037 lots, the broker takes 0.03 or 0.04, and that rounding is a 10 to 25 percent error on the intended risk. Tolerable, and worth understanding: on a small account you are systematically slightly off the intended size, in a direction that depends on rounding rather than on anything you chose.

€5,000 to €10,000

Rounding drops to a few percent of intended risk and stops mattering. This is where the strategy you tested and the strategy you are running converge.

Above €10,000

Sizing is no longer the constraint. What matters from here is whether the strategy has an edge after costs, which is a different question and the only one that was ever really interesting.

Costs bite harder on small accounts

The cost per round turn is roughly fixed: on a raw-spread account, in the region of 6 to 7 USD per lot plus the spread. On a 0.01 lot position that is a small absolute number, but measured against a position that small it is a large percentage of the expected gain.

The practical test: work out the average pip gain per trade the strategy needs, and compare it with the broker's real cost per round turn. If the strategy averages 4 pips of edge and the round trip costs 2.5, the edge is mostly gone before the account size is even discussed. Small accounts do not change that arithmetic; they just make it harder to notice, because the absolute numbers are small.

Add the fixed monthly costs if you host the system yourself. A VPS at 5 to 30 euro a month is trivial against a 50,000 account and a meaningful drag on a 1,000 one. That is one of the reasons we host the algo rather than asking clients to rent a VPS.

The alternative: rent the capital

A prop firm evaluation converts a capital requirement into a fee plus a rule set. Instead of posting 50,000 you pay a few hundred for an evaluation on a notional 50,000 and trade under daily loss, maximum drawdown and consistency rules.

It is a genuine option, with a genuine trade. You are not risking 50,000, you are risking the fee, and the notional size means the lot-minimum problem disappears entirely. In exchange, the rules can withdraw the account for behaviour that would be unremarkable on your own money, and the system has to be built to respect them. That is covered in which prop firms allow an EA.

Which is better depends on what you are short of. Short of capital, the evaluation is the answer. Short of time or patience for rules, your own account at a workable size is the answer.

What this means in practice

Our algo trades forex and gold on your own account at your own broker, and the practical minimum is around €1,000, for exactly the reasons above. We host it and set it up, so there is nothing to install and no VPS to rent. Pricing is €599 a year or €1,799 once, both including VAT.

The EA Hub has no account minimum, because you control the sizing yourself and can decide what compromise you are willing to make. If neither fits, the challenge passing service skips the capital question entirely.

The full picture of what running an automated system involves is in the automated forex trading guide.