Inside the Prop Firms Top score
We score 90 prop firms from 2.0 to 5.0 with seven weighted parts: countries accepted (30%), cost (20%), rules and drawdown (15%), payouts and split (15%), trader reviews (10%), ways to pay and get paid (5%) and brand popularity (5%). Today the scores run from 3.0 to 4.5, and the latest data run was on October 5, 2026. One formula and one dataset apply to every firm, the Editor's Choice included.
The formula
Each part gives the firm a mark from 0 to 100. We take the weighted average and convert it: score = 2 + 3 × weighted average. A firm with full marks everywhere would get 5.0, and one with zero everywhere 2.0.
| Part | Weight | What it measures |
|---|---|---|
| Countries accepted | 30% | How many countries the firm turns away |
| Cost | 20% | Cheapest all-in entry price and price per $1K of max loss |
| Rules and drawdown | 15% | Drawdown model, consistency rule, PT:DD and minimum days |
| Payouts and split | 15% | Fastest payout cycle and highest base split |
| Trader reviews | 10% | Public ratings, pulled towards neutral when reviews are few |
| Ways to pay and get paid | 5% | Kinds of deposit and payout methods |
| Brand popularity | 5% | Monthly searches for the brand name |
Countries accepted (30%)
Every restricted country takes 1/120 off a mark of 100, so a firm that refuses 120 countries or more scores zero. A ban on the whole EU counts as 27 countries. A ban on part of a country, such as one Canadian province, counts as half. A firm that publishes no restricted list gets a neutral 50. This part weighs the most because a firm that will not accept you is no use, however cheap it is.
Cost (20%)
Cost has two equal halves, and both compare the firm only with firms in the same market (forex, futures or crypto). A firm active in several markets gets the average of its markets.
- Entry price: the cheapest way in, counting the list price plus any fee charged after you pass. Prices in other currencies are converted to US dollars at the ECB reference rate.
- Price per $1K of max loss: the fee divided by the money you may lose before the account closes, as the median over all of the firm's accounts.
Why not price per $10K of account size? Because the nominal size overstates what you buy when the loss limit is tiny. Take two $100K accounts. One costs $99 with a 1% max loss, the other $549 with a 10% max loss. Per $10K of size the first looks more than five times cheaper ($9.90 against $54.90). Yet the first gives you $1,000 of room and the second $10,000, so per $1K of max loss the first costs $99 and the second $54.90. We still print price per $10K in our tables, because many traders shop that way.
Rules and drawdown (15%)
Every account the firm sells gets four marks. The firm's rules mark is the average over all its accounts.
| Rule | Share | How it is marked |
|---|---|---|
| Drawdown model | 35% | Static 100, end-of-day trailing 65, balance trailing 50, equity trailing 35, intraday trailing 30. Trailing of an unstated kind 45, unknown 50. |
| Consistency rule | 30% | No rule 100, a rule of any size 45, not stated 60. |
| PT:DD | 20% | Profit goal of all phases added together ÷ max drawdown, worked out by us. Full marks at 0.6 or lower, zero at 1.6 or higher, a straight line in between. Not stated 50. |
| Minimum trading days | 15% | 0 or 1 day 100, 2 or 3 days 80, 4 or 5 days 60, more 40, not stated 60. |
We compute PT:DD ourselves because the aggregated data mixed two opposite ratios. With a typical two-step at 8% plus 5% against a 9% cap, the ratio is about 1.5, so this mark mostly rewards tests that ask for less than usual. Our guide to passing shows the ratio across the market.
Payouts and split (15%)
Half is speed: a payout you can take within a day (daily or on request) scores 100, within 7 days 85, within 14 days 65, anything slower or unstated 45. Half is the highest base split: 90% or more scores 100, 80% or more 80, lower or unstated 55. A higher split sold as a paid add-on does not count.
Trader reviews (10%)
We blend the firm's Trustpilot rating with the review score in the aggregated comparison data, weighted by how many reviews stand behind each. A Trustpilot rating with no review count on record counts as 50 reviews. On our scale, 3.0 stars or less is 0 and 5.0 stars is 100.
Few reviews prove little, so we pull the mark towards a neutral 50. The pull fades as reviews add up: about 10 reviews keep a third of the distance from neutral, 100 keep two thirds and roughly 1,000 keep all of it. A firm with no rating at all gets 45. When Trustpilot warns that it removed fake reviews from a profile, the firm's distance from neutral counts at half the normal weight.
Ways to pay and get paid (5%)
Half covers kinds of deposit method: card, crypto, e-wallet, bank transfer and local options such as UPI or Pix. Four kinds or more score in full. The other half covers kinds of payout method: crypto, payout platforms such as Rise, bank transfer and e-wallets. Three kinds or more score in full.
Brand popularity (5%)
We use monthly Google searches for the firm's name across all countries, from Ahrefs keyword data (pulled September 30, 2026). Sub-brands carry their parent company's number. Each firm is placed as a percentile among the others, and a firm with no search data gets 25. Demand shows a firm is known and in use. It says nothing about its rules or payouts, which is why it weighs only 5%.
How country pages differ
A country page first removes every firm that lists that country as restricted. For the rest, the countries-accepted part is replaced by local fit: how easily someone there can pay and get paid. Deposits score 100 with a method that is local to that country, 70 with cards, 60 with e-wallets, 50 with crypto only and 30 with nothing we can match. Payouts score 100 with a local method, 80 with a payout platform or bank transfer, 70 with e-wallets and 60 with crypto only. The other six parts stay as they are, so a firm can place higher in one country than it does worldwide.
How topic lists are ordered
A topic list, such as scalping, payouts or static drawdown, answers one question, so its order starts from the data that question needs. Each list rates firms on its own measure: commission per lot for scalpers, the depth of the fixed floor for static drawdown, the allocation ceiling for scaling, years and review volume for track record. Firms are placed as percentiles on that measure and on the overall score, and the two are blended. The fine print under each full list names the measure and the weight the overall score keeps. Three lists sort on one figure alone: the cheapest firms by price per $1K of max loss, the $1 list by upfront price and the daily payout list by payout cycle. Firms with an open warning never make a list's top seven or its picks.
When a value is missing
On our pages a missing value stays empty, and we never fill it with an estimate. Inside the score a missing value gets the neutral mark listed above for that part, so a gap neither rewards nor punishes a firm much.
What never enters the score
Affiliate commissions, the Editor's Choice pick and our own site traffic have no input. Neither does search demand by country, which we use only to decide which pages to build.
The Editor's Choice
Hash Hedge is our Editor's Choice, a pick made for editorial and commercial reasons. It is pinned above the ranking on every page whose filter it meets, carries a label and links to our affiliate disclosure. Beside it you see two numbers: an editor's rating of 4.5/5, which is our opinion, and the calculated score of 4.0/5, which puts it at #45 of 90 in the full ranking. The formula gives it no boost, and only the calculated score sets rank order.
Public warnings
When we find a public warning, such as a regulator alert naming the company behind a firm or a review site freezing its rating, we flag it at the top of the firm's review and beside its name in our tables. The formula stays the same, but firms with an open warning are kept off our lists for beginners and for the longest-running firms.
Where the data comes from
- Aggregated public comparison data: prices, account rules, programs, platforms and restricted countries for 66 of the 90 firms. We use list prices only and ignore promo codes published elsewhere.
- Our own dossiers: for 24 firms we built the whole record from the firm's pricing pages, rules and terms (last checked October 5, 2026). These include the Editor's Choice and firms the aggregated data misses or gets wrong.
- Our corrections file: 54 fixes across 27 firms, each with the official page it comes from and the date we checked it (latest October 5, 2026). A correction overrides the aggregated value.
- Trustpilot public pages: ratings for 67 firms. For the 30 most searched firms we also record review counts, star spread and replies (taken September 29, 2026).
- ECB reference rates for prices set in other currencies (rate of September 29, 2026).
- Ahrefs keyword data for brand popularity, as above.
What we check by hand, and how often
In October 2026 we compared the records of all 31 firms that have a full review with their official pricing pages, rules and terms. Every difference we confirmed went into the corrections file. Own dossiers are rebuilt by hand from the firm's site. We refresh the aggregated data every two to four weeks, and each new correction or refresh triggers a full recalculation of every score. Each page shows the date its content last changed.
How errors get fixed
If a price, rule or country ban here no longer matches what the firm publishes, email us with a link to the firm's page. We check it against the official source, log the fix with its date and rebuild every page that uses the number. Our editorial policy covers the rest of the process.
How we earn
At present our only affiliate partner is Hash Hedge, the Editor's Choice. If you buy through its links, we may earn a commission. That money is never an input to the formula. Our affiliate disclosure has the details.