forex-scalping.com

Cent accounts, lot rounding, and why small tests lie

Testing a scalping idea on a cent account feels safe and sensible. The trouble is that cent and micro accounts often route differently. Some brokers fill them internally against their own book, while standard accounts get passed through to liquidity providers.

Results follow. A strategy that prints money on a cent account may struggle on a standard one because the fills are no longer internal. The spread on micro accounts is sometimes the same as standard but the slippage behaviour is not.

Rounding matters too. A position of 0.013 lots might be rounded to 0.01, which changes the effective risk per trade. Over hundreds of trades that skews the whole statistical picture, and no one notices because the platform reports whatever it executed.

Fair approach: use a micro account to test mechanics, not performance. Check that orders do what they should, that the stop placement works, that the code is stable. Then move to a standard account with the smallest size and judge the strategy there.

Did the results on a small account survive the move to a standard one, and how large was the gap?

My results didn't survive the move, and the gap was bigger than I'd guessed. On the cent account a breakout scalp on EURUSD was showing a profit factor around 1.4 over a couple of hundred trades. Moving to a standard account at the smallest size, it dropped to just under 1.1 over a similar number, which is barely worth the effort.

When I dug into it, the difference was almost all in the entries. On the cent account the fills were nearly always at the exact price I clicked, which in hindsight should have been a red flag. On the standard account there was a steady bit of slippage on every breakout entry, small each time but enough to matter on a short target.

I now use the small account exactly as you say, to check that orders, stops and code behave properly. I don't trust a single performance number from it anymore.