Prop Firm Drawdown Rules, Explained: Trailing vs. Static, and How to Survive Both
A plain breakdown of trailing, end-of-day, and static drawdown rules — the real math behind how much room each one gives you, and how to size trades to survive an evaluation.
Most people who fail a prop firm evaluation don't fail because they picked bad trades. They fail because they never translated the drawdown rule into a number — a contract count, a stop distance — before they started trading. The rule sat in the fine print while the account traded on vibes.
That's backward. The drawdown rule is the actual constraint on the account. The profit target is just the finish line; the drawdown rule is the track you have to stay on to get there.
Three rules, three different shapes
Prop firms don't all define "max drawdown" the same way, and the difference isn't cosmetic — it changes how much real room you have at any given moment.
Static drawdown is the simplest. The floor is set once, from your starting balance, and it never moves. A $100,000 account with a 5% static drawdown has a hard floor at $95,000, full stop. Once you're up, that floor stays exactly where it started. This is the most forgiving version, because profit you've made doesn't shrink your future room.
End-of-day trailing drawdown recalculates the floor once per day, based on your highest end-of-day balance so far. If your $100,000 account closes a day at $102,000, the new floor becomes $102,000 minus your drawdown percentage — the floor only ratchets up at the close, not intraday. You can have a rough day within the session and still be fine, as long as you close above the new floor.
Intraday (real-time) trailing drawdown is the strictest. The floor trails your highest account equity, marked live, including open unrealized profit, at every moment of the session. Hit a new equity high at 10:14am and the floor moves up immediately — even if you give that profit back before the close. This version punishes not protecting an intraday high almost as much as it punishes an actual losing trade.
The same "5% drawdown" label can describe three very different amounts of real breathing room, depending on which of these three a firm is running. Read the rule before you read the profit target.
How much room a trailing rule actually gives you
This is the part that catches people. On a static rule, 5% of a $100,000 account is $5,000 of room, and it stays $5,000 no matter what you do. On a trailing rule, that $5,000 is only ever the room you have right now — and it shrinks every time you make a new high.
Say you're on a $100,000 account with a 5% intraday trailing drawdown. You start the day flat, floor at $95,000, giving you the full $5,000 cushion. You catch a good trade and your equity touches $98,000 intraday. The floor immediately trails up to $93,100. Your cushion from current equity is still $4,900 — barely changed — but your cushion from your starting balance has effectively shrunk to $1,900, because $3,100 of what used to be room is now locked in as a floor you can't fall below.
This is the mechanic people miss: on a trailing rule, profit doesn't just pad your account, it also seals off the floor behind you. That's good news if you protect gains — you're locking in progress toward passing. It's bad news if you don't respect it, because giving back an intraday high counts against a floor that already moved, not the one you started the day with.
Practically: on any trailing rule, the moment you're up meaningfully intraday, your effective remaining risk budget is smaller than the account's original drawdown percentage suggests. Trade the account you actually have room in, not the one on the label.
Sizing so you can't gamble your way out
The failure mode on a drawdown rule is always the same shape: a losing stretch shrinks the cushion, and the trader responds by sizing up to make it back faster — which is exactly backward, since the cushion is now smaller and a bigger position burns through what's left faster. The desk version of this problem is solved by never letting position size be a discretionary decision in the moment. It should be arithmetic, decided in advance.
Start from the rule, not from a "feel" for the market. Take your current drawdown cushion — the actual dollar distance from current equity to the current floor, not the original percentage — and decide what fraction of it you're willing to risk on any single trade. A reasonable ceiling is 5–10% of remaining cushion per trade. On a $5,000 cushion, that's $250–$500 of risk per trade, not per day.
From there the math is mechanical. If your stop distance on a setup is 20 points and each point is worth $2 per contract, that's $40 of risk per contract. A $400 risk budget for the trade supports 10 contracts — not "however many felt right." If the cushion has shrunk to $1,900 after a trailing floor moves up, the same 10% rule now caps you at $190 of risk, which is under 5 contracts on the same setup. The position size is a direct function of the cushion, recalculated every time the cushion changes, not a number you picked once at the start of the evaluation.
This is also why evaluations are rarely failed by one bad trade. They're failed by a trader who kept trading the size that matched the account's starting cushion after a losing stretch had already eaten most of it. The rule didn't get harder. The trader just stopped adjusting to the rule that was already there.
One thing to do before your next evaluation session
Before you place a trade today, write down three numbers: your current drawdown floor, your current cushion in dollars, and the position size that a fixed percentage of that cushion actually supports at your stop distance. Recalculate all three after every session, not just at the start of the evaluation. The rule that ends an evaluation is never a surprise on the statement — it's arithmetic that was knowable in advance, and most accounts that fail it simply never did the math.
I built Fourdesk's Risk Manager desk to track exactly this — the running cushion against a drawdown rule, recalculated after every trade instead of estimated from memory.