Prop firms for scalpers: the three tests behind “scalping allowed”

When a firm reviews fast trading it runs one of three published tests, and they measure different things. A minimum-hold test measures time: the trade must stay open past a stated cutoff. A concentration test measures distribution: how much of your profit (and in some clauses your trade count as well) came from holds inside a stated window. A simulated-fill test measures execution realism: rapid orders that only work because the sim fills you at prices a live order book would not.

Across the firms in the list below, the published windows run from five seconds to two minutes and the profit-share triggers from 20% to 50%. Passing one test clears nothing about the other two: a plan with no minimum hold can still flag a profit concentration, and a plan with no concentration rule can still terminate an account for fill abuse. The question to answer per plan is which test it runs, on which denominator, in which phase.

• Time test: hold duration against a stated cutoff.
• Concentration test: share of profit, sometimes trade count too, from holds inside the window.
• Execution test: rapid orders that depend on fills a live book would not give; reviewable even where no minimum hold exists.

Pull your platform's trade export for your last 20 sessions and bucket every closed trade by hold time: under 5 seconds, under 10, under 30, under 2 minutes. For each bucket, compute the share of total profit those trades produced. That one column is your side of every clause in the list below, and it exists before you spend anything.

The worked example below shows why the distance from the trigger is the decision number. At $460 of sub-five-second profit in an $800 session, the share is 57.5%. To sit under a 50% trigger, short-hold profit had to stay at or below $400. The session is $60 over, and moving those 7.5 points into longer holds means a different exit, a wider target, a changed method. A session does not drift back under a trigger by itself.

The decision rule this guide uses: treat any plan whose published trigger sits within 10 points of your measured share as restricted for you, whatever the marketing label says. The clauses publish thresholds; most do not publish the lookback window, so assume one flagged session can be reviewed on its own rather than diluted by clean ones.

Scalping fit is also a drawdown question, because the two rules tax the same equity shape. Under an intraday trailing model, the loss limit follows your live peak including open profit, so it can tighten in the middle of a trade. Under an end-of-day model it ratchets only when the daily close settles higher, so intraday spikes leave it alone. A static limit never moves at all. A scalper's curve (many small peaks of open profit that partly give back before the close) is exactly the shape that separates these models.

Run the example's $460 through that lens. Assume it arrived the ordinary scalper way, as a run of peaks that each gave part of itself back rather than one clean step up. Under an intraday trailing rule every one of those peaks pulled the loss line up before the giveback, so the same closed $800 leaves you less room than it would under an end-of-day rule that only ratchets on the settled close. The decision this changes: read the drawdown model for the exact plan and phase beside the scalping clause, because on some plans the model switches at funding.

The expensive version of this mistake: you read “all strategies welcome” on a pricing page, pass the evaluation scalping, and get flagged in the funded phase, where the same threshold now carries a heavier penalty. The published consequence ladder across the list runs from a documented warning, to advancement held at the end of an evaluation, to account termination with the short-hold profit forfeited. One covered firm publishes exactly that split at a single parameter: a warning at one profit share, a violation at a higher one, held advancement during the evaluation, termination plus forfeiture once funded.

So finish the read with three facts from the clause itself: which phase the language covers (evaluation and funded text have to be checked independently; what the denominator of the test is) profit, trade count, or both; and what happens at the threshold. A clause that warns at one share and violates at a higher one gives you an operating margin. A single line with termination behind it gives you none, and your 10-point buffer has to come from your own distribution instead.