How we rate funded and prop trading programmes
Updated: . Kept by The Funded Desk.
One methodology runs every ranking on this site. It was fixed on 23 September 2026, before a single programme was read against it, and it has not moved since. Each page decides which programmes appear and which criterion sets their order; the weights below decide the overall score, and a programme carries the same score everywhere here. Scores run from 0 to 20, rounded to 1 decimal place, under methodology version 1.1.
What the five criteria price
| Criterion | Weight | What it measures | How it becomes a score |
|---|---|---|---|
| Cost of access | 34 | The bill for reaching $100,000 of trading nominal: the firm's list price for it, and any margin the trader has to lock up to hold that much. | A point on this line costs $80 in fees, or $1,000 in margin the trader has to lock up to reach the same nominal. It opens at ten, both bills are deducted from it, and zero is the floor however deep the subtraction goes; a fee returned with the first payout is credited back at a point and a half. Why it is set that way: a fee is gone whatever happens, so it is charged at full weight; posted margin stays the trader's money and is charged at a tenth of the rate; where a firm sells no $100,000 tier, the nearest tier is scaled to $100,000 |
| Payout terms | 26 | The invoice running the other way: how much of the profit the firm hands back, how long the trader waits to ask, how small a request it will take, and what the money arrives as. | Nothing scores here until the split clears 70%; past that, every five points of split is worth one, and the count stops at six. Then three boxes to tick — two points where a payout can be requested at any time, one where the minimum is $100 or less, one where it settles in crypto — with the line capped at ten. |
| Evaluation model | 14 | What stands between the payment going out and the first payout coming back: how many phases, and whatever day counts and consistency clauses are bolted onto them. | no evaluation at all — 10; instant access whose loss limit trails profits — 8; one evaluation phase — 6; two phases — 4; three phases — 2; minus 1 for a minimum number of trading days, minus 1 for a consistency rule |
| Risk rules | 18 | The clauses that can write an account off — daily caps, trailing drawdowns, consistency rules — and whose money goes with it when one of them fires. | no daily loss limit — 3; loss limit fixed rather than trailing — 3; no consistency rule — 2; breaching a limit does not forfeit the account — 2; minus 2 where the trader's own capital is the thing at risk |
| Instruments | 8 | What the fee buys access to: a route out to a crypto order book or a simulation of one, and whether the firm prints how many markets come with it. | orders reach an exchange order book — 6; simulated or contract-for-difference — 3; market count published and 100 or more — 4; market count published and below 100 — 2 |
Each criterion is scored 0-10 on its own evidence, then weighted and rescaled to the 0-20 shown on the site. A weight is simply how far a criterion can move a total: cost of access by 6.0 points, payout terms by 5.0, the evaluation model by 4.0, risk rules by 3.0 and instruments by 2.0. Nothing else enters the arithmetic — not how long a firm has traded, not how it describes itself, not what its reviews say.
A score, worked through
Breakout, on the terms it published on 22 September 2026, is the clearest way to see the formula move. Its cheapest route to $100,000 of nominal costs $330 and the fee is not returned, so cost of access lands at 5.9 of 10: ten points less one for every $80 of fee. The split is 80%, payouts are on request from $50 in USDC, and that adds up to 6.0 on payout terms — two points for the split, two for paying on request, one for the low minimum, one for paying in crypto. One evaluation phase with no minimum days and no consistency rule gives 6.0 on the evaluation model. A daily loss limit costs it three points on risk rules, a fixed rather than trailing limit earns three back, and the account is forfeited on a breach, so 5.0. Crypto reaches no market and no market count is published, so instruments is 3.0.
Weighted at 30, 25, 20, 15 and 10 and rescaled to 0-20, that is 11.1. Every other programme is built the same way, and the parts are printed beside each score in the table above.
Every programme scored, part by part
Every programme that meets the scope rule below has a row here, 7 of them. Scores as of 22 September 2026, rebuilt from the terms each firm publishes with the formulas above. The five columns after the score are the raw 0-10 criterion scores, before the weights are applied.
| Programme | Score | Cost of access | Payout terms | Evaluation model | Risk rules | Instruments |
|---|---|---|---|---|---|---|
| Goat Funded Trader | 12.7 | 6.0 | 9.0 | 6.0 | 5.0 | 3.0 |
| Breakout | 11.1 | 5.9 | 6.0 | 6.0 | 5.0 | 3.0 |
| HyroTrader | 9.8 | 4.3 | 6.0 | 4.0 | 3.0 | 10.0 |
| FTMO | 9.3 | 5.2 | 4.0 | 5.0 | 5.0 | 3.0 |
| BrightFunded | 7.9 | 3.2 | 4.0 | 5.0 | 5.0 | 3.0 |
| Crypto Fund Trader | 5.8 | 1.8 | 2.0 | 5.0 | 3.0 | 7.0 |
| SabioTrade | 5.4 | 2.8 | 2.0 | 5.0 | 0.0 | 7.0 |
Read a row across rather than down the score column. HyroTrader takes 10.0 on instruments, the highest part score on the page, and still finishes third: reaching a real market is worth 2.0 points of the 20, while what the trader pays to get there is worth 6.0.
Why a fee counts more than ten times posted margin
The cost criterion divides a fee by 80 and posted margin by 1,000. In plain terms: $80 of entry fee costs one point out of ten, and so does $1,000 of margin. A dollar handed to a firm therefore weighs more than ten times a dollar the trader still owns. That is a judgement, not a measurement, and it is the single decision that does most to set the order of this site.
The case for it is that the two dollars have different futures. An evaluation fee is spent: it buys one attempt, and on most programmes here it is gone whether the attempt succeeds or fails. Posted margin is still the trader's balance, and can be withdrawn while it is not sitting in a losing position.
The case against it is that margin can be lost, and a trader who breaches a limit loses far more of it than any fee would have cost. The methodology answers that in the risk-rules criterion, which subtracts two points where the trader's own capital is the thing at risk. That subtraction was not enough, which is the next section: every programme in the ranking pays a fee and posts no margin, so the ratio now settles the order among fees alone and the margin half of the rule scores nothing at all.
Why an unreadable price means listed, not scored
Cost of access carries 30 of the 100 weight points, more than any other criterion, and it needs one number: what the firm charges for $100,000 of trading nominal. Two programmes on this site do not publish it in a form we could record. Bitfunded prices its accounts behind a tier selector we could not read reliably, and FundedNext renders its pricing table in the browser, so no fee for a $100,000 account was read on the pages we opened.
We could estimate one, and the estimate would be worth up to 6.0 of the 20 points. So both are listed with their terms on the pages where they belong, left out of the ranking, and the reason is printed beside them. That cuts against us as often as for us, and we accept it: a programme can be cheaper than everything scored here and still have no row in the table. The omission says nothing about the firm except that we could not read its price.
Which programmes we score
The ranking covers programmes where the firm puts up the capital a trader is about to hold: proprietary firms that sell an evaluation, and firms that sell instant access. One of those is scored when it publishes a price for $100,000 of trading nominal, or a tier that scales to it. Two other kinds are described on these pages and carry no total. A firm whose price we could not read is one. The other is an exchange programme that raises a trader's nominal against margin they post themselves, where nobody lends anything: four of the five criteria would be scoring the absence of terms that were never on sale, so the terms are printed and the column is left empty.
One programme is described on almost every page here and carries no total: EVEDEX Funded Trading. It raises the nominal a trader can hold against margin they post themselves. Nobody lends them capital, no firm balance sheet stands behind the position, and only the real balance can be withdrawn.
It was scored to begin with, on the same five criteria, and the result is why it no longer is. Four of those five ask what a firm charges for its capital, what it takes back before paying, how many hurdles it puts in front of a payout and which of its rules can end an account. A programme that sells none of those things scores the absence of all of them, and the formula put it more than a third clear of the highest-scoring firm on the site — a gap that described our criteria rather than the product. The capital itself, which is the entire thing a proprietary firm sells, carries no weight in the formula at all and cannot be added to it without rebuilding every score on the site.
So the terms are printed and the column is left empty. That cuts against the programme on pages where it would have ranked well, and we accept it: a total nobody can compare is worse than no total. The same rule would apply to any exchange margin tier we added later.
Prices we converted or scaled
Four programmes do not publish their entry cost in the form the formula needs, so here is the arithmetic behind the number we used.
- FTMO: €439 for the $100,000 one-step challenge, recorded as $504 at the ECB reference rate of 1.1490 USD per euro for 21 September 2026.
- BrightFunded: €477 for the $100,000 evaluation, recorded as $548 at the same reference rate.
- SabioTrade: no $100,000 plan is sold, so the Plus plan at $289 for a $50,000 balance is scaled to $578 for $100,000 of nominal.
- EVEDEX Funded Trading: nothing is charged as a fee, and the 3,335 USDT of margin that opens a 100,000 nominal enters the formula as posted margin instead (EVEDEX funded trading documentation, read 18 September 2026).
Where a firm sells several routes, the one named in its table row is the one scored. Goat Funded Trader also sells instant funding; its rule set was not read, so the one-step challenge is what the score describes.
What we read and what we did not do
- Prices, targets, loss limits, profit splits, payout terms and risk rules: each firm's own pricing pages, rule pages and disclosures, read on 22 September 2026.
- EVEDEX Funded Trading: the exchange's own help pages and terms of use (EVEDEX funded trading documentation, read 18 September 2026).
- Nothing else. No affiliate dashboards, no review aggregators, and no claim a firm makes about itself that its own terms do not carry.
We bought no evaluation, passed none and requested no payout. That is the boundary of this method: published terms are a promise about future conduct, and a firm can write a fair rule set and still argue about a withdrawal. Where a marketing page and the terms disagree, the terms are what we record.
Ties, rounding and updates
Programmes whose final scores are equal after rounding share a place and are marked as tied. Inside a shared place the order is alphabetical and carries no meaning. The table on this page is ordered on the unrounded totals, so it can separate two programmes that a ranking page shows as tied.
A change of weights, formulas or scale gets a new version number and a new date here, and every ranking is rebuilt with it rather than mixed with the old one. Programme terms are read again twice a quarter, and out of turn whenever a firm moves a price, a target or a payout rule. A date on this site moves only when something behind it moved.
Why the seat is priced differently this time
Version 1.0 priced a seat at cost 30, payouts 25, evaluation 20, rules 15, instruments 10. This one prices it at cost 34, payouts 26, rules 18, evaluation 14, instruments 8.
The desk says on every page what it is for: the price of a seat, the firm's cut of the profit, and the clause that can void the account before payday. Two of those three led the old weights and the third did not — the rule that ends an account outranked by the shape of the evaluation that opens one. A trader who fails an evaluation pays again; a trader caught by a consistency clause or a hidden drawdown rule after passing loses the payout the seat was bought for. The rules now sit above the evaluation model, and instruments gave up two points because every firm here is priced on the same crypto perpetuals.
What did not happen: the weights were not tried against the table first. They were written off the line above, fixed, and the table recomputed once. The order came out exactly as it had at version 1.0 — GoatFundedTrader, Breakout, HyroTrader, FTMO, BrightFunded, CryptoFundTrader, SabioTrade — with totals within a few tenths. Had it come out differently it would have been published differently.