Where Prop Firm Revenue Comes From

A prop firm sells attempts at an evaluation. Fee income from those attempts is the revenue line, and profit share paid out to traders is the largest cost against it. In the only audited accounts anyone in the category has filed — FTMO’s FY2024 filings under Czech law — $322.8 million of revenue carried $176.3 million of payouts and $48.1 million of operating expenses, leaving roughly $98 million before tax.

The structure is unusual in one specific respect. Revenue is collected in full at the moment of purchase, while the cost that offsets it arrives later, in unpredictable size, and from a small minority of the people who paid. A firm knows its revenue on the day of sale and does not know its cost of goods for months.

Almost none of it is disclosed. Exactly one operator in the category files audited accounts, and only because its parent is subject to Czech filing obligations. Exactly one independent dataset describes the funnel: FPFX Technologies, a software provider to prop firms, covering more than 300,000 accounts, 100,000 traders and 10 firms, published by Finance Magnates in September 2024. One firm, Topstep, discloses some funnel statistics voluntarily.

Everything else in public circulation is either a figure a firm chose to put on its own homepage or an estimate produced by a vendor. That is why this page is built around one company’s filings and one independent dataset rather than around an industry average: an industry average would be an average of marketing claims.

For the same reason, we publish no figure for the size of the industry. The totals in circulation disagree with each other by more than an order of magnitude and trace back to vendor and content-marketing material rather than to filings, regulator returns or exchange data; there is no method by which a reader could check any of them. What can be observed instead are two anchors. FTMO reported $322.8 million of revenue in FY2024, up 53% year on year. Separately, Finance Magnates tracks crypto payouts across the top ten firms — a tracking set that excludes both FTMO and The5ers — and recorded roughly $325 million paid across those firms during 2025, with Q1 2026 volume of $115.1 million, flat against Q4 2025, across 61,682 individual payout events at an average of $1,865.

Those two anchors describe payout volume and one firm’s revenue. They do not add up to a market size, and we do not treat them as if they did.

Firm prices and terms below were checked on 31 July 2026 and change frequently; the FY2024 accounts are the most recent filed figures available at that date.

---

What does a prop firm actually sell?

It sells an evaluation. The buyer pays a fee, receives a simulated account with a profit target and a set of loss limits, and — if the objectives are met — is given a second simulated account on which a share of notional profit is paid out in cash. The fee is charged whether or not the objectives are met, and it is the primary revenue line at every firm in the category.

How often does the same buyer pay again?

The FPFX dataset is the only independent measurement of repeat behaviour in the category. Across its sample it records roughly three challenges per account, average spend of $800 per account, and 2.2 firms per trader.

The last of those three is the most consequential and the least discussed. Traders do not migrate between firms; they hold accounts at more than two firms at the same time. There is no observable loyalty behaviour in the data, which means the revenue line at any one firm is a share of a trader’s parallel spend rather than the whole of it.

Firms’ own product design reflects the expectation of repeat purchase in several places. FTMO refunds the evaluation fee in full on the first payout and offers unlimited free trials, with no paid reset product on sale. Funding Pips prices a reset at 15% off, available only within seven days of the breach. Topstep charges a $149 activation fee for each funded account earned, and prices reactivation of a lost account between $599 and $829 within a 30-day window. Apex applies a lifetime payout cap of $18,000 on a $100K sim-funded account, after which the account closes and a new evaluation must be purchased.

Each of those is a mechanical description of a published term, and each has the same effect on the revenue line: the unit of sale is an attempt, not a customer.

Is the list price the price anyone pays?

Frequently not. On 31 July 2026, several firms displayed a discounted price rather than the list price on their own product pages, and had done so continuously enough that the discount is best read as the operating price.

Funding Pips showed $555 / $529 / $399 across plan tiers for a $100K account and $99 versus $66 at the entry tier. E8 Markets listed $398 for its $100K plan with a $299 promotional price. The5ers ran a $149 promotional price. Apex’s 4.0 pricing sat in a roughly $20–$80 promotional band against list prices of $277 and $396. Blueberry Funded’s published $100K prices of $275 / $300 / $325 are described on the firm’s own materials as already carrying a 30% discount. Topstep’s landing page presented its $49 monthly plan as a reduction from $85.

A list price in this category is not a useful input to any calculation. The average realised ticket is the number that matters, and the only route to it from published data is indirect: $800 of average spend across roughly three challenges per account gives about $265 per attempt. That is arithmetic on the FPFX figures, not a price anyone advertises.

---

What is the largest cost?

Trader payouts. In FTMO’s FY2024 accounts, $176.3 million of $322.8 million of revenue went out to traders — a payout ratio of 54.6%, which is the only payout ratio in the industry derived from audited filings and roughly double the 25–40% that third-party revenue estimates assume.

The rest of the cost side is smaller, but it is where the differences between operators sit.

The profit comes from operating leverage. With payouts at 54.6% and operating expenses at 14.9%, roughly 30% of revenue survives to the pre-tax line — not from retaining trader profit, but from what it costs to run the business around it. The filed holding-level result for FY2024 was a net profit of CZK 1.3 billion, about $62.5 million, on revenue of CZK 6.84 billion.

No firm publishes its acquisition cost. What is published is the affiliate rate card: FTMO pays 8–20% of referred revenue on a tiered basis, FundedNext 8–15%, Funding Pips 5–10%. Cost-per-acquisition figures and “effective CAC per funded trader” numbers circulate widely in the sector, but the ones we found trace to vendor reports rather than to any firm’s disclosure, so we do not publish them. The one audited comparison point available is from the adjacent retail brokerage business: Trading 212 spent 18.5% of FY2025 revenue on marketing, at five million clients and with an established brand. That is a floor for a business with those advantages, not a benchmark for a new entrant.

Payment costs are structurally material and structurally invisible. Under Visa’s consolidated acquirer monitoring programme, the Excessive Merchant threshold falls to 1.50% from 1 April 2026, at 1,500 or more events per month, and the acquirer portfolio threshold is 0.50%, enforced from 1 January 2026, with a $8 fee to the merchant per dispute at the Excessive level. Mastercard’s excessive chargeback programme escalates fines from $1,000 in months two and three to $100,000 and $200,000 from month 19, and measures the ratio against the previous month’s volume — so a firm that stops marketing still carries the tail of disputes into a smaller denominator. There is no dedicated merchant category code for the purchase of a trading evaluation; in practice these sales are commonly processed under MCC 8299, schools and educational services, while Visa’s own merchant data standards manual requires a business line carrying high integrity risk to be coded to a category that reflects it. No firm in the category has published its dispute rate, and we found no audited disclosure of one.

Why does the model need scale?

Take the audited cost structure and the independent per-account spend figure and put them side by side. This combines two different sources — one firm’s FY2024 accounts and a ten-firm dataset from 2024 — so it describes the shape of the model rather than the results of any particular company.

Against $800 of lifetime spend per account, a 54.6% payout ratio removes about $437 and operating expenses at 14.9% of revenue remove about $119, leaving roughly $244. Affiliate revenue share at the published 5–20% takes $40–$160 of that, before any spend on paid channels, sponsorship or brand. What remains has to cover licensing, platform, risk systems, support and staff — all of which are close to fixed.

That arithmetic is why the businesses that survive in this category are large. It also explains the observed behaviour of the payout line, which is our reading of the same sourced numbers rather than a finding: because the FPFX dataset records that only 1–3% of buyers become sustainably paid over more than six months, the payout cost does not spread evenly across a customer base — it concentrates in a small tail of accounts that keep withdrawing. A former risk director at The5ers put the same point publicly as “Less Than 1% of Traders Can Bankrupt a Prop Firm.”

There is a documented instance of that mechanism operating. On 12 July 2026 Alpha Futures closed its flagship Premium plan, stating that the plan had produced more than $25 million in payouts over two months and describing the result as “significant operating losses.”

---

What does the model rest on?

1. The unit of sale is an attempt, not a customer. Revenue is evaluation fees, bought roughly three times per account, at $800 of lifetime spend and 2.2 firms per trader — about $265 per attempt on those figures. Resets, add-ons, upgrades and data are second order next to that, and the firm’s retained share of trader profit is not separately disclosed by anyone in the category.
2. Payouts are the dominant cost and the volatile one, at 54.6% of revenue in the only audited accounts filed, against operating expenses including all marketing at 14.9%.
3. No market-size figure survives checking. We publish payout-tracking volumes and one firm’s filed revenue instead.

Related reading in this cluster: what “up to 100% profit split” leaves out covers how much of a trader’s profit converts into cash. The clauses that let a payout be refused covers the conditions attached to that conversion. Self-reported payout totals against independently tracked ones covers why firms’ own totals cannot be used as revenue inputs. Which firms refund the evaluation fee, and on what condition covers the refund mechanics referenced above. Every published dataset on pass rates and payout rates covers the funnel figures and their provenance in full. Across clusters, the licensed brokerage arms that firms are now building covers the second revenue model appearing at the largest operators.

---

Written by the upme.com research desk. The FY2024 figures above come from filed accounts reported by trade press rather than from a filing we retrieved ourselves, and we say so where it matters. Prices and terms were checked on 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
• Source
• Source
• Source