0.01 or 0.1 lot steps: how rounding changes risk on small accounts

[FXS London Scalper]

On small accounts, rounding the lot size changes actual risk more than people think.

A worked example: €2,000 account, 1% risk (€20), EURUSD with a 7 pip stop. The exact size is about 0.31 lots (assuming roughly €0.92 per pip per 0.1 lot, depending on the rate).

• With 0.01 lot steps: 0.31 lots, risk close to 1%.
• With 0.1 lot steps: 0.3 lots, risk about 0.97%. Fine.
• With a 4 pip stop instead: exact size about 0.54 lots. With 0.1 steps, rounding to 0.5 gives 0.92%, rounding to 0.6 gives 1.1%.

On very small accounts, the minimum lot itself can exceed your intended risk. A €500 account with 1% risk (€5) and a 10 pip stop needs about 0.05 lots. If the broker's minimum is 0.1, you're forced to risk double.

Practical points:

• Always round down, never up.
• Check the volume step in the symbol specification before choosing an account type.
• For tight stops on small accounts, a broker with 0.01 steps matters.

How do you handle rounding in your size calculator?

[Amanhasnoname]

I ran into the minimum lot problem when I opened a small cent-style account to test a new setup with real fills. The account was around four hundred euros, and on GBPJPY with a 9 pip stop the honest size came out at something like 0.03 lots. The broker allowed 0.01 steps, so it was fine, but I had nearly picked their other account type where the minimum was 0.1, and that would have meant risking over three times what I intended on every single trade.

In my calculator I do exactly what you suggest and floor the result to the volume step, but I also print the actual risk after rounding next to the planned risk. Seeing 0.84% instead of 1% on screen is a good reminder that the rounding isn't free, and on the odd day it lands much lower I sometimes just skip the trade.

One extra check I'd add is the maximum volume per order. It sounds irrelevant on a small account, but some brokers also cap lots on certain exotic pairs, and an EA that doesn't know that can fail silently.