Most people size positions by feel, adjusted by conviction. On a personal account that is merely expensive. On a funded account it is fatal, because the account has a hard floor and a strategy that survives on average can still find the floor on a run of losses.
The fix is to derive size from the rules once, before the first trade, and then not to revisit it during a session.
The arithmetic
Start from the smaller of your two limits — the daily loss — and work back.
- Take the daily loss limit in currency. On a $50,000 account with a 4% daily limit, that is $2,000.
- Decide how many consecutive losses a normal bad day contains for you. If you take three trades and can lose all three, that is three.
- Divide, then halve. $2,000 across three losers is $667; halving it gives roughly $330 of risk per trade, and the half is what leaves you a day rather than a breach when the fourth trade also loses.
- Convert that risk figure into a position size using your actual stop distance, per instrument. This is the number you trade — not a lot size you remember from another account.
Then check it against the drawdown
The daily limit governs a session; the maximum drawdown governs the account. Run the same size through a losing streak — six, eight, ten trades — and see whether the account survives it. If ten losses at your calculated size breaches the overall floor, the size is wrong regardless of what the daily arithmetic said.
Ten consecutive losses is not a pessimistic assumption. On a strategy that wins 45% of the time it is an ordinary occurrence over a few hundred trades, and an evaluation is easily a few hundred trades.
Ten losses in a row is not a worst case. On a 45% win rate it is a Tuesday, and your size has to survive it.
Why conviction sizing fails here
Scaling up on high-conviction setups is defensible in principle and disastrous in this context, for a reason that has nothing to do with whether your conviction is accurate.
Conviction correlates with recent results. You feel most certain after a run of winners and most tentative after losses — which means variable sizing systematically puts your largest position on immediately before a losing streak begins, and your smallest immediately before a recovery. The sizing amplifies the swing in exactly the wrong direction.
Flat sizing gives up some theoretical edge and buys the thing an evaluation actually requires, which is surviving to the end of the sample.
Leverage is a ceiling, not a target
Plans publish a maximum leverage, and it is routinely mistaken for a recommendation. It is a limit on what the platform will let you do, set so it does not bind on any reasonable position. If your sizing arithmetic ever brings you close to it, the arithmetic has gone wrong somewhere upstream.