Are Prop Firms Worth It in 2026? The Math Nobody Shows You

What do prop firm pass rates actually say?

Four of the nine futures firms PropRank tracks publish primary funnel statistics, and the published pass rates run from 16.8% to 36.22% per attempt — better than the folklore, worse than the ads. These are the only primary numbers in the industry, published largely for ad-substantiation compliance, as of August 18, 2026.

Watch the denominators. Take Profit Trader's older disclosure said 20.37% of registered users passed a test between January 1 and August 31, 2023. Its current number says 36.22% of Tests passed in 2025. Those measure different things in different years. "Registered users" includes everyone who signed up and barely traded; "tests" counts attempts. Neither is a lie. You just cannot compare them, and firms will always show you the flattering one.

Two more things the data says. First, no audited industry-wide pass rate exists anywhere. The "90-95% of traders fail" line you see everywhere is an estimate with no traceable origin study, and the FTMO 9-10% figure people cite is derived from public data and community surveys, not an official disclosure. Second, single-step futures evals genuinely pass at higher rates than the CFD-era folklore suggests. A 36% pass rate is not a 5% pass rate. But passing is not the point.

Does passing the eval mean you get paid?

No. Passing is the middle of the funnel, not the end, and the back half of the funnel is where the money disappears. Topstep's own 2025 funnel, the most transparent in the industry: 51.8% of participants eventually reached the Funded Level across their attempts. Of those funded traders, 33.3% received a payout. Multiply through: 0.518 × 0.333 ≈ 17.2% of participants got paid anything. And of Express Funded Account traders, 0.71% advanced to Live Funded Accounts in 2025.

Tradeify's disclosure corroborates the shape: 28.5% of its funded-level participants received a payout between August 2025 and July 2026, and 3.0% got called to live.

Roughly one participant in six received a payout, at the firm most willing to show you its numbers. That is the statistic the marketing does not show.

What is the expected value of a $50K eval?

Approximately breakeven: about +$12 on roughly $273 of expected spend, using Topstep's published 2025 funnel and 2026 effective prices, before valuing your time at anything above zero. Here is every step.

Step 1 — price the attempt. Topstep's 50K Combine sells at $49/month standing price (the strikethrough anchor above it, $85 to $165 depending on where you look, is doing perpetual-sale work; $49 is what people pay). Resets are $49 and the subscription rebills at $49/month, so each additional attempt costs about $49 either way.

Step 2 — estimate attempts. Topstep's two published numbers imply the attempt count. If each Combine passes at 16.8% and 51.8% of participants eventually get funded, then 1 − (1 − 0.168)^n = 0.518 solves to n ≈ 3.97. Call it four attempts. This assumes attempts are independent and your skill holds constant — both generous assumptions.

Step 3 — total the costs. Four attempts × $49 = $196 of eval spend. Add the $149 activation fee, paid only by the 51.8% who fund: $149 × 0.518 ≈ $77 expected. Expected total spend ≈ $273.

Step 4 — probability of ever being paid. 51.8% × 33.3% = 17.2%.

Step 5 — value the payout. Topstep does not publish an average payout, so we use the only independently reported average in our dataset: Lucid Trading's $1,652 average payout across 307,000+ payouts as of July 28, 2026, per fundedprogramfinder.com. It is a proxy from a different firm; flagged as such. Expected payout value: 0.172 × $1,652 ≈ $285.

Step 6 — the verdict. $285 − $273 = +$12.

Honest caveats, both directions. The model understates the upside because 33.3% "received a payout" includes traders who took many payouts; the entire positive tail of this trade is repeated withdrawals, and one average payout ignores it. It understates the downside because four attempts is months of screen time, and a 17.2% shot at $1,652 spread across four months is a negative hourly wage for the median entrant. Everyone in this business is selling you the right tail.

Run the same shape at Apex and the cost side shrinks: a 50K eval lists at $550 one-time but sells at $55 with the standing 90% code (SAVENOW), and Apex's discount dependence is the most extreme we track — the promo price is the price. No resets exist; a blown eval is a $55 repurchase, plus a ~$99 activation at funding (third-party figure, unverified). But Apex publishes no pass-rate funnel, so you cannot run this math on it, which is itself information. And its 6-payout hard cap, after which the account closes, amputates the only region where the EV goes positive. PropRank may earn a commission if you use a code on this page. That never changes a score.

A $50K account is really a $2,000 account

Every 50K evaluation PropRank tracks carries a $2,000 maximum drawdown, so the real account — the number that decides whether you survive — is $2,000 with $50,000 of buying power. That is the Real Account Size reframe, and it changes the worth-it question. You are not buying a $50,000 account for $49. You are renting a $2,000 risk budget with $50,000 of buying power attached, and the rental terms depend heavily on how the drawdown trails.

Your own $2,000 at a retail futures broker carries the same dollar risk with no rules, no profit split, and no expiry. What the eval fee actually buys is two things: the buying power, and a hard cap on your total cash loss at the sum of fees paid. For a disciplined trader, that cap is genuinely valuable. For an undisciplined one, it is a subscription to blowing up.

How big is the prop firm economy, really?

About $850 million of revenue in 2026, by the only bottom-up estimate available — the $12-20 billion figures you see quoted are unverified and measure something else. Track360's 2026 industry report sizes retail prop trading at ~$850M revenue in 2026, up 45% year over year (versus ~$450M in 2021), across 30+ major firms, ~2.1M funded traders, and ~12M annual challenge purchases. The competing claims — "$12B in 2025, possibly $20B in 2026," or Business Research Insights' $7.14B for a combined "forex and prop trading market" — use incompatible definitions and unverifiable methodology. The roughly 14x spread between the projections reflects different questions, not disagreement about the same one.

Two lines in the revenue mix deserve your attention. Challenge fees are 42% of industry revenue ($357M): the industry is structurally paid per attempt, not per success. And affiliate commissions are 15% ($127.5M), with affiliates driving 40-50% of new-trader acquisition. That is why nearly every "review" you read ends in a coupon code, and why PropRank discloses its own.

Who actually pays your payout?

On a sim-funded account, the firm pays you out of its own revenue. Your trades fill in a simulation; no market counterparty ever loses money to you, and your withdrawal comes from the pool of eval fees, resets, activations, and whatever the firm earns trading its own live capital. This is not a hidden scandal. It is the disclosed model, and the honest futures firms say so plainly. But it changes what "funded" means: you are not managing the firm's capital, you are in a revenue share on a simulation the firm underwrites.

Live capital is the exception, not the rule. Topstep moved 0.71% of Express Funded traders to Live in 2025. Tradeify called 3.0% of funded participants to live accounts. TradeDay automates the path: $10,000 of gross profit triggers an overnight review to Funded Live. The money in sim payouts is real — Topstep claims $1.4B+ paid all time, and Lucid's payouts were independently verified at $463M by MondoTraders. The market exposure usually is not.

Will regulation change the math?

Possibly within months. The CFTC opened a public consultation on futures-prop oversight, with comments closing November 30, 2026, per Track360's Q3 roundup. The core questions: whether challenge fees constitute commodity-pool participation interests, and whether futures-prop operators should register as CTAs. Either answer pulls firms into CFTC/NFA registration, capital requirements, recordkeeping, and audits. Separately, NFA Notice I-26-12 is reported effective December 1, 2026: mandatory affiliate disclosures, prohibition of "guaranteed funded account" language, and a required sim-vs-live distinction in promotional material.

Context matters: the CFTC's My Forex Funds case was dismissed with prejudice in May 2025, which chilled enforcement appetite, so expect deliberate movement rather than sudden shutdowns. But the direction is clear. Registration would raise compliance costs, favor large firms, and probably kill the messiest marketing. Consolidation is already underway — roughly 80-100 firms exited between 2024 and early 2026, with $50M+ in trader funds blocked industry-wide. A regulated industry likely means fewer firms and cleaner claims. It may not mean cheaper evals.

When is a prop account actually worth it?

When you can already prove the edge, and the fee buys leverage instead of hope. The checklist is short:

1. You have months of your own data. Positive expectancy in your own sim or small live account, documented, before you pay anyone. The eval should feel like a formality.
2. The fee is your full downside. Your maximum cash loss is the fees you pay, not the drawdown. Priced at $49-$90 effective, that is cheap, defined-risk access to $50K of buying power.
3. The structure is survivable. End-of-day drawdown at both stages, cheap resets, no lifetime payout caps. That is what the PropRank Survivability Score measures: Lucid Trading leads at 90/100 (EOD-lock everywhere, $0 activation, $90 resets — with the honest youth flag that it launched in 2025), with Topstep, TradeDay, and My Funded Futures tied at 86.
4. You size it like a $2,000 account. Because it is one.

When is it not worth it?

When the monthly fee is a lottery ticket against rules you have not read. The tells are consistent: you have no data on your own trading; you bought because it was 90% off; you cannot explain whether your drawdown trails intraday or end-of-day; and the firm caps the tail. Apex closes the account after six payouts. Elite Trader Funding caps lifetime sim payouts at $25,000 per trader — it scores 56/100 on Survivability, the weakest of the nine we track, with a 3.8 Trustpilot and 24% one-star reviews as of August 18, 2026. A capped tail stapled to a breakeven base case is negative expected value by construction.

The subscription treadmill is its own trap. At $79-$85/month effective, three months of not passing costs more than a one-time eval, and the rebill is designed to feel like sunk cost. If you are resetting for the fourth time on feel, the math above already told you how this ends.

The blunt recommendation

Do not buy an eval to find out whether you can trade. That is the most expensive possible way to learn, and the funnel data says it plainly: about one entrant in six sees a payout, and the average one roughly breaks even before valuing their time. Build three green months in your own sim first. It costs nothing.

If you have the data, then a prop account is one of the cheapest forms of defined-risk leverage available to a retail futures trader, and the move is boring: pay the effective price, never list. Pick an EOD-drawdown structure with cheap resets and an uncapped tail — Topstep is our best overall, Lucid is the survivability leader with the 2025-launch caveat, TradeDay if you want the automated sim-to-live path. Size it like the $2,000 account it really is. Take payouts early and often, because a sim payout is a promise from firm revenue, not a market claim — and two-thirds of funded traders never collect even one. And keep an eye on the calendar: the CFTC comment window closes November 30, 2026, and the rules may look different a year from now.