Two prop firms, the same $2,000 drawdown, opposite risk
Almost every backtest measures drawdown the way a brokerage statement does: on closed trades. Most prop firms don't. We pulled the current rulebooks for two of the biggest futures evaluators and found the same headline number — a $2,000 max drawdown on a 50K account — enforced two completely different ways. One of them can fail your account on profit you never booked.
The same limit, measured two ways
| Firm (50K) | Max drawdown | Checked on… | Open-trade dip below the line? |
|---|---|---|---|
| Firm A (combine) | $2,000 | every tick, incl. open P&L | fails you |
| Firm B (standard eval) | $2,000 | closed trades / end of day | not caught |
Firm A's max-loss line only ratchets up at the end of the day, but the breach is checked intraday, in real time, against your equity including open positions. Firm B's trailing line updates after a trade closes; an intraday dip in unrealized P&L during an open trade is simply not the thing that fails you. Same $2,000. Opposite intraday risk.
We learned this the expensive way
On a Firm-A practice account, two of our runs came within ~$87 and ~$95 of the liquidation line — while realized P&L never went near it. The account was almost killed by an open-trade trough that a closed-trade backtest would never have flagged. The signal was fine. The position would have recovered. The rule didn't care.
A backtest that scores drawdown on closed trades is silently assuming Firm B's rulebook. Run that same strategy under Firm A and "you survived" can quietly become "you were liquidated at 2:14pm on a position that closed green an hour later."
Why this is a backtest lie, not a footnote
Drawdown type isn't paperwork — it's part of whether the edge survives at all. An overnight mean-reversion trade that dips hard before it pays is perfectly fine under a realized/end-of-day rule and a death sentence under an intraday-unrealized one. The same code, the same fills, the same fortnight can pass one firm and bust the other. If your backtest doesn't model which drawdown rule applies, its survival number is fiction — and it's the optimistic kind of fiction.
It gets one layer deeper: at one firm the rule even changes by account family. Its standard evaluations are realized-trailing, but a funded variant trails on unrealized profit until a buffer is built. Two accounts at one firm, two drawdown engines.
What we do about it
Our risk layer encodes each firm's drawdown rule explicitly — intraday-unrealized vs realized vs end-of-day — from one source of truth, and the protective guard is tuned tighter on the firms that can bust you on open profit. Before any strategy is allowed near a funded account, it's checked against the actual rule it will trade under, not the brokerage-statement version a notebook assumes. The edge is only half the question. The other half is whose rulebook you're surviving.