◆Payouts and firm economics
Pass Rates and Payout Rates: Every Number That Exists
Only one independent dataset measures prop firm pass rates. Here is every published number, its sample, its publisher, and what it cannot show.
In the only independent dataset published on this question, 14% of prop firm evaluation accounts reached a funded account and 7% of buyers ever received a payout. That dataset — FPFX Technologies, 300,000-plus accounts across 10 firms, published in September 2024 — is the single independent measurement in existence. One firm discloses its own funnel voluntarily. Nothing else is audited.
No regulator collects pass rates. No exchange publishes them, no filing requirement produces them, and no firm in the category is subject to an audit that covers them. What exists is one dataset from a software vendor whose customers are prop firms, one firm’s voluntary disclosure, and one third-party analysis of failure causes that we could not trace to a published dataset.
The funnel and the revenue structure are the same arithmetic seen from two ends; the other end is where prop firm revenue comes from.
Terms used here that carry a firm-specific meaning — funded account, payout, Combine — are defined in the glossary.
What datasets actually exist?
| Dataset | Publisher and date | Sample and period | What it measures | What it does not measure |
|---|---|---|---|---|
| FPFX Technologies funnel data | Published by Finance Magnates, 18 September 2024. FPFX is a software provider to prop firms | 300,000+ accounts, 100,000 traders, 10 firms. The covered period is not stated | Share of accounts reaching a funded account; share of funded accounts receiving at least one payout; share of all buyers ever paid; average payout as a share of account nominal; average spend per account; challenges per account; firms per trader | Which firms are in the sample; the period the accounts span; variation between firms; results per person rather than per account; anything about payout requests that were refused |
| Topstep’s own funnel statistics | Topstep, figures circulated for 2025 | One firm, US futures only, subscription pricing | Share of Combines passed; share of people who reached funded at least once; share of funded traders who received a payout; share of Express Funded traders who reached a Live Funded Account | Not audited. Single firm, single product type. We were not able to locate the full set on a single Topstep-owned page; the figures reach the public through third-party comparison write-ups |
| hoc-trade failure-cause analysis | Reported by Invezz on 24 July 2026, citing a vendor review of six firms’ rulebooks | ~500,000 traders | The split of failure causes: breach of a loss limit versus failure to reach a profit target | Pass rates, payout rates, or anything about money. The underlying dataset is not published, and we could not verify it against a primary document |
| FTMO FY2024 filed accounts | Czech filings, reported by TradeInformer and Finance Magnates | One firm, one financial year | Money: $322.8M revenue against $176.3M paid to traders | People. FTMO publishes no pass rate and no payout rate of any kind |
Sources, in order: Finance Magnates on the FPFX dataset; Topstep’s own company page and payout policy; Invezz, 24 July 2026; TradeInformer on FTMO’s filings.
Only the first entry is independent. Topstep’s figures are self-published by a single firm, the hoc-trade split is a secondary report of an analysis nobody has published, and FTMO’s filed accounts measure money rather than people.
What do the funnel numbers say?
| Stage | FPFX, 10 firms, per account | Topstep, one firm |
|---|---|---|
| Reached a funded account | 14% | 16.8% of Combines initiated |
| Reached funded at least once, measured per person | not measured | 51.8% |
| Of funded traders, received at least one payout | 45% | 33.3% (28.3% in 2024) |
| Received a payout, as a share of everyone who paid | 7% | approximately 5.6% |
| Became sustainably paid over more than six months | 1–3% | not published |
| Reached a live, non-simulated funded account | not measured | 0.71% of Express Funded traders |
The FPFX row that gets quoted most is the last-but-two: 7% of buyers ever receive a payout, a figure whose relationship to firm revenue is worked through in the payout-ratio piece and is not repeated here.
The Topstep column is the more instructive one, for a different reason. It is the only place in the public record where the same population is measured two ways by the same party — and the two numbers differ by a factor of three.
Why is “the pass rate” the wrong question?
Because the denominator is never the same twice, and it is almost never stated.
- Per account. FPFX reports 14%. Its sample is 300,000-plus accounts held by 100,000 traders, so the average trader in the sample holds about three accounts. A 14% figure computed on accounts is not a 14% probability for a person.
- Per person. Topstep reports that 51.8% of its traders reached a funded account at least once, against 16.8% of Combines passed. Same firm, same customers, same year. The per-person figure is roughly three times the per-attempt figure, and both are correctly described as a “pass rate”.
- Per purchase. A purchase is not a fixed quantity. At FTMO or FundedNext an evaluation is a single fee; at Topstep and MyFundedFutures it is a monthly subscription that can be renewed several times within one attempt. Counting purchases therefore counts something different at each firm.
Same behaviour, three answers.
The unit problem shows up inside the measurements themselves. FPFX reports roughly three challenges per account. A separate measurement by Swiset covering about 10,000 traders, reported by Finance Magnates, gives 1.6 challenges per user with a maximum of 18. Both can be true: one counts per account, the other per user.
So two firms quoting pass rates cannot be compared unless both state their denominator, and in the material we reviewed none of them does. A firm publishing “our pass rate is 20%” is making a statement that cannot be checked against another firm’s 12% without knowing whether either number counts accounts, people, attempts or subscription payments. A pass rate without a stated denominator is not a number.
What causes the failures?
The only figure available is the failure-cause split, and it needs its provenance stated carefully. An analysis attributed to hoc-trade, covering approximately 500,000 traders, is reported to find that around 70% of evaluation failures are breaches of a loss limit rather than failures to reach a profit target. That figure reached the public through a review of six firms’ rulebooks reported by Invezz on 24 July 2026; the underlying dataset has not been published, and we could not verify it against a primary document. We publish it as a reported figure rather than as a measurement.
If it is even directionally right, it locates the decisive variable in the rules rather than in the market: the mechanic that ends most evaluations is the loss limit, and loss limits differ substantially between firms in how they are anchored and whether they move. That mechanic is covered in how trailing and static drawdown floors actually work.
What does nobody publish?
The absences are consistent across the whole category.
No firm publishes a per-cohort pass rate. Nothing is broken down by purchase month, product, account size or one-step versus two-step. Since rule sets change frequently — Topstep made eight rule changes in five months to April 2026, and Blueberry Funded operates two parallel rulebooks for accounts opened before and after 12 March 2026 — an undated, unsegmented pass rate would not be interpretable even if it were published.
No firm publishes a payout refusal rate. Not one operator publishes how many payout requests were declined, on what grounds, or as what share of requests made. Several firms do publish payout ledgers: Funding Pips maintains a live rewards page recording 67,914 rewards and $66.78 million paid in 2026, and Alpha Capital publishes a register of $62,973,647.94 across 30,352 transactions. Those record payments made. Neither they nor any other public source records requests refused.
No firm publishes a distribution of payout sizes. Only averages exist. FPFX gives an average payout of about 4% of account nominal across ten firms. FTMO publishes average payout by product, ranging from 5.61% to 6.80% of nominal. Finance Magnates’ tracking of crypto payouts across the top ten firms gives an average of $1,865 across 61,682 payout events in Q1 2026, down from $2,020 in Q4 2025. An average tells you nothing about whether the population behind it is uniform or is a handful of large payments and a long tail of small ones — and the FPFX finding that only 1–3% of buyers become sustainably paid suggests the second shape rather than the first.
No published pass rate is audited. The FY2024 FTMO accounts are the only audited numbers in the category, and they contain no funnel data at all.
Related reading: claimed against independently verified payout totals covers the same provenance problem on the money side of the funnel.
Written by the upme.com research desk. Every figure above is attributed to the party that published it, with its sample size and unit of measurement, because in this subject the unit is the finding. Figures are as of 31 July 2026. Corrections to the address on our sourcing page. Nothing here is investment advice.
Sources
Every factual claim above is drawn from one of the documents below. Where a document has been superseded since the date given, tell us and the piece is corrected with a dated line.
- Source
- Source
- Source
- Source
- Source
- Source
- Source