Trailing drawdown explained: end-of-day vs intraday, by plan and phase

Intraday trailing puts the line a published distance below your highest live account value, open profit included, and recomputes it in real time. In the worked example below, a $51,000 live peak on a $2,000 distance moves the line to $49,000 mid-trade, before anything settles, while the position is still open. End-of-day trailing puts the line the same distance below your highest settled daily close, so it moves at most once per day and an intraday spike that gives itself back never moves it. Static puts the line at the starting balance minus the distance and leaves it there.

Two definitions vary by plan and both change the arithmetic: the basis (some plans compute the line from balance (settled results only), others from equity (open positions included)) and the family, because at least one covered firm sells an intraday model and an end-of-day model as separate products chosen at checkout. The distance itself varies less than the type: seven of the eight covered firms publish exactly $2,000 of maximum drawdown on their standard 50K plan and the eighth publishes $1,750, which is why the example uses $2,000.

The expensive misreading of "end-of-day" is holding a deep dip through the session because the line "only updates at the close." Six of the eight firms in NANO's covered set publish the sentence that kills it: the line moves at the daily close, but a breach is tested in real time against open equity, touch the line mid-session and the account is liquidated, no close required. The other two publish no timing sentence at all; treat that silence as live enforcement.

Put that against the example numbers: with the line at $48,000, equity of $49,200 is $1,200 clear, but an intraday dip to $47,900 ends a live-enforced account even though the line itself would have stayed put until settlement. The decision this changes is hold depth: your real stop is the line, not the close, so the deepest dip you can carry is the current distance ($1,200 here) minus whatever you need to survive the exit.

Published trailing rules end in one of three ways: the line trails for the life of the account, it locks once it reaches the starting balance, or it locks $100 above the starting balance. Fourteen of the sixteen standard-plan phase cells in the covered set (eight firms, evaluation and funded) publish one of the two lock levels; one evaluation publishes that its line never stops, and one evaluation's answer depends on which platform vendor runs the account, so it stays unresolved. Where a plan publishes nothing, NANO models the trail as never stopping, the conservative direction, and states no lock the firm has not published.

The lock is what turns profit into room. While the line is still trailing, every new settled high drags it up behind you, so part of any buffer is on loan and the account can die above its own starting balance. Once the line locks, settled profit above the lock is permanent: on the example account a lock at $50,000 lets the account survive any give-back down to its starting balance, while a lock at $50,100 ends it $100 sooner. Small, but it is the number that decides how much of a funded buffer is genuinely extractable and when a withdrawal is safe to take.

Copying the evaluation's drawdown settings into your funded model because the account label looks the same is how traders meet intraday trailing for the first time with payout money on the line. On covered plans the type per phase is a recorded fact: some run end-of-day in both phases, some switch to intraday once funded, and inside one firm different plan families answer differently. The snapshot below records both phases for each firm's standard plan.

If your plan's funded row reads intraday, set the funded size from the live-peak arithmetic, the $200-of-room case in the example, where the end-of-day reading of the same session shows $1,200. If either row is not separately published, leave the model unresolved and confirm with the firm before trading that phase; an assumed rule is the one input the simulator cannot correct.