Challenge rulebook · UpMe Ltd.

The rules, published
before the product

Every condition that can end an evaluation or a funded account, how each number is measured, and what happens when one of them runs out. Version 1.2, dated 2026-08-03 — before a single account exists to apply it to.

v1.2 · currentPublished 2026-08-03Status: not in forceBinds accounts opened under this version
Challenge corridor Rulebook 1.2 · 2026-08-03 · not in force
+10.00% Profit targetclosed profit only

No position: no account exists

0.00% Starting balancethe line every limit is measured from
-5.00% Daily floorequity at 00:00 UTC, trading pauses
-10.00% Maximum floorstatic, never moves

The cap that keeps a pass honest, and the limit that isn’t

≤40% Consistency capbest day & best trade, each ≤ 40% of total profit
none - no limit Time limitno phase has a deadline

Two-step route, phase one. Percentages, not amounts: account sizes are not set.

What this document is, and what it is not

This is the rulebook for an evaluation that opens at v1.2. No account operates under it yet, nobody has bought anything, and nothing in it is in force. It is published now for one reason. A set of rules written before there is a single trader to apply them to cannot have been reverse-engineered from a dispute.

Version 1.2, dated 2026-08-03. The version number is not decoration. When the product launches, every account will be bound to the rulebook version it was opened under, for the life of that account, and a change will only ever bind accounts opened after it. The mechanism is described in how a rule change works and the principle is restated on the transparency page.

Figures below are expressed against a starting balance rather than in currency. The published account sizes and fees are linked from the pricing page.

The number has moved five times already, every time before any account exists: on 2026-08-02 v0.10 added rule 10, on account multiplication and v0.11 revised its review clause — a published pause, a 72-hour cap, and a mandatory decision at the cap — on 2026-08-03 v1.0 added the best-day and best-trade consistency caps (11a and 11b) and raised the minimum trading days to 4, v1.1 added seven anti-gaming rules (12 through 18) and clarified the automation boundary in rule 9, and on 2026-08-03 v1.2 raised the phase-one profit target from 8% to 10% — a round target, harder to reach. The only time changes like these can be made without a dispute standing behind them is now.

What the account actually is

The evaluation and the funded stage that follows it are simulated accounts priced from live market data. They place no orders in any market. A payout from a funded account is a performance fee calculated on a simulated result — it is not investment income, and the account is not an investment account.

We are stating this at the top rather than in a terms page, because the distinction decides what the product legally is. A separate, real brokerage account is on the roadmap and requires authorisation we do not hold; it is described on the legal status page and nowhere on this one.

Every rule, and what breaching it costs

Ten numbered rules, with the consistency requirement (rule 6) spelled out in two caps — 11a and 11b — added at v1.0, and seven anti-gaming rules — 12 through 18 — added at v1.1. Each one gives the limit, how the number is measured, and what happens when it runs out. The measurement line is the one that decides arguments, and across the firms in our research it is the line most often published somewhere other than the rules page.

Rules 12 through 18 close the ways a simulated venue can be gamed rather than traded: the feed (12), the simulated fill (13), trade duration and account activity (14), event straddling (15), single-position exposure (16), conduct around the daily reset (17), and buying a pass on variance rather than earning it on skill (18). Some of them can only be enforced once the trading platform is wired; those carry that flag inline and openly, rather than being published as if already switched on.

1. Profit target

Limit
10% of the starting balance in phase one, then 5% in phase two.
One-step route: 10% in a single phase.
How it is measured
Closed profit only, measured against the starting balance. There is no minimum number of winning days, but from v1.0 the profit has to be spread: the best-day and best-trade caps in rules 11a and 11b mean no single day and no single trade may carry more than 40% of the profit at the moment the target is reached. A phase closes on consistent profit, not on one clean trade.
If it is breached
Nothing breaks. The account converts to the next stage.

2. Daily loss limit

Limit
5% of the balance at 00:00 UTC.
One-step route: 4%.
How it is measured
Measured against equity at 00:00 UTC, including open positions. It is a fixed figure for the day and does not trail an intraday high: a profitable morning does not raise the floor for the afternoon. It resets on the hour, not on a rolling 24-hour window.
If it is breached
Trading pauses until 00:00 UTC. The account survives and nothing is forfeited.

3. Maximum drawdown (evaluation)

Limit
10% of the starting balance, as a static floor.
One-step route: 8%.
How it is measured
Set once from the starting balance and never moved. It does not trail peak equity, so profit you have made is yours to give back without ending the evaluation.
If it is breached
The evaluation ends. One retry is offered at a reduced fee.

4. Trailing drawdown (funded)

Limit
8% of the starting balance, trailing, locking at the starting balance.
How it is measured
Trails your highest closed equity until it reaches the starting balance, then locks there permanently and never moves again. It does not trail unrealised profit, and it does not reset when a payout is taken.
If it is breached
The funded account closes. Any payout already accrued is still paid.

5. Minimum trading days

Limit
4 days.
How it is measured
A day counts when at least one position is opened or closed. Days do not have to be consecutive, there is no upper bound, and there is no minimum volume attached to a day. v1.0 raised this floor from three days to 4.
If it is breached
The target cannot be claimed yet. Nothing else changes and nothing is lost.

6. Consistency requirement

Limit
Enforced from v1.0. Two caps, each 40% of profit at target — the best trading day and the best single trade, set out in full in rules 11a and 11b below.
How it is measured
A pass has to reflect repeatable skill, not one lucky day or one lucky trade. Two caps enforce that, both read at the single moment the profit target is reached: no one day and no one trade may carry more than 40% of the total profit. Earlier versions ran with no consistency cap and said so; v1.0 reverses that, in public and dated in the changelog, rather than leaving it as a silent omission. The exact measurement and what happens at the cap are in rules 11a and 11b.
If it is breached
The profit target cannot be claimed while either cap is exceeded. The evaluation does not end and nothing is forfeited — the trader keeps going until the profit is spread within both caps.

7. Time limit

Limit
None.
How it is measured
No phase has a deadline. An account is archived only after four consecutive quarters with no trade, and archiving is reversible on request.
If it is breached
Not applicable. There is no clock to breach.

8. Copying another funded account

Limit
Not permitted.
How it is measured
Mirroring another funded account, at upme or elsewhere, is checked on execution timestamps across accounts linked to one identity. It is a timing test, not an opinion about strategy: two accounts holding the same view is not a breach, two accounts filling within the same tick repeatedly is.
If it is breached
The payout is refused under this clause, the reason is published, and the evaluation fee is returned.

9. News trading, weekend holding, EAs

Limit
All permitted — with one boundary on what an EA may be built to do.
How it is measured
No blackout window around scheduled releases, no forced flat before the weekend, no restriction on automated execution. Automation is permitted: an EA that automates how you trade is fine. The boundary, clarified at v1.1, is on what an EA is built to exploit — an EA whose function is to work the feed (rule 12), the simulated fill (rule 13) or account hyperactivity (rule 14b) is not automation of a strategy, it is a prohibited practice run by a robot, and it is refused as the practice, not as automation. If any of this changes it changes for accounts opened after the change, never for accounts already open.
If it is breached
Not applicable to automation itself. An EA built to run a practice forbidden under rules 12, 13 or 14b is treated as that practice.

10. One trader, counted once

Limit
Up to 3 concurrent evaluation accounts per person.
How it is measured
The cap binds to a verified identity at payout, where KYC happens, and to a correlated-behaviour cluster before that. Three things breach it: (a) opposite or offsetting positions across accounts — the trader’s own, or coordinated with someone else’s — where the combined exposure nets to roughly zero while one side of the pair harvests a pass or a payout; (b) transferring, buying or selling an account, or having a third party trade an evaluation; (c) the same trades duplicated across a cluster of accounts to multiply payouts. Detection is a trade-correlation review — timing, instrument, direction, size — that runs continuously through the evaluation rather than being saved for the payout gate, plus payment-method linkage at purchase and identity linkage at KYC. What counts as a violation and what it costs are published; the detection thresholds are not, and we are saying so here rather than leaving it unsaid.
If it is breached
The evaluation is void, and the void is recorded in the public denial ledger on the payouts page citing this rule. A violation found after a payout is published the same way.

11a. Consistency — best trading day

Limit
No single trading day may account for more than 40% of the total profit at the moment the profit target is reached.
How it is measured
Checked once, at the instant the target is met, not day by day during the evaluation. The net closed profit of the single largest day is divided by the cumulative profit; that share must be at or below 40%. It never caps a day in isolation and never ends a session — it only tests the largest day’s share of the finished result, so a pass reflects more than one good session.
If it is breached
The target is not claimable on that result. The evaluation continues and nothing is forfeited; the trader trades on until no single day carries more than 40% of the total.

11b. Consistency — best single trade

Limit
No single trade may account for more than 40% of the total profit at target.
How it is measured
The same instant and the same arithmetic at the trade level: the profit of the single largest closed trade, divided by the cumulative profit, must be at or below 40%. One outsized trade cannot carry a pass on its own.
If it is breached
The target is not claimable yet. The evaluation continues and nothing is lost; the trader trades on until no single trade exceeds 40% of the total.

12. Feed integrity

Limit
Not permitted. Enforced once the trading platform is wired.
How it is measured
An evaluation is priced from live market data on a simulated venue. Strategies whose profit depends on the difference between that feed and the real market — acting on a faster external feed, on a delayed or mispriced tick, or on a price the platform shows in error — are not permitted. It is a source-of-profit test, not a speed limit: fast trading on the price everyone sees is fine; trading against a price only you can see because it is stale or wrong is not. The detection thresholds are not published; the ground for refusal is, and we are saying so here. This ground publishes now and is enforced once the trading platform is wired.
If it is breached
Profit from the affected trades does not stand. A deliberate, repeated pattern voids the evaluation and is recorded in the public denial ledger citing this rule.

13. Execution realism

Limit
Impossible fills are reversed. Enforced once the trading platform is wired.
How it is measured
Because fills are simulated, a result that could not have occurred against real liquidity does not stand — an impossible stop execution, a zero-slippage entry in a fast market, a stray fill through a gap, a queue-position exploit. This pairs with the gap-reversal clause below: the same principle applied to a fill rather than to a gapped stop. It publishes now and is enforced once the trading platform is wired.
If it is breached
The trade is reversed to the price the real market would have produced. The account is not ended unless the pattern is deliberate and repeated.

14. Trade duration & account activity

Limit
(a) Trades held under 60 seconds may together carry no more than 30% of total profit at target.
(b) No more than 2000 platform requests or 200 executed trades in a single UTC day. 14b’s counting is enforced once the trading platform is wired.
How it is measured
(a) The same share-cap arithmetic as rules 11a and 11b, read at the same instant the profit target is reached: the profit from trades held under 60 seconds, divided by the cumulative profit, must be at or below 30%. Like the other consistency caps it never ends a session and never caps a trade in isolation — it defers the claim rather than ending the account. (b) An account may not exceed 2000 platform requests or 200 executed trades in a single UTC day; the counting publishes now and is enforced once the trading platform is wired.
If it is breached
(a) The target is not claimable while short-hold trades carry more than 30% of the total; the evaluation continues and nothing is forfeited. (b) New orders pause until the next UTC day — nothing already closed is reversed.

15. Event straddling

Limit
Not permitted.
How it is measured
News trading stays permitted — rule 9 is unchanged. What is not permitted is placing opposing or bracketing orders around a scheduled high-impact release to capture the gap whichever way it breaks. One position carried through an event is trading; two opposite pendings straddling it to harvest the volatility is not. It needs no additional platform wiring — reading a pair of pendings around a scheduled release is enough.
If it is breached
The straddling orders are void and any profit from them is removed. A deliberate, repeated pattern voids the evaluation citing this rule.

16. Single-position exposure

Limit
No single position may put more than 2% of the starting balance at risk at entry. Enforced once the trading platform is wired.
How it is measured
Risk at entry is distance-to-stop × size, or the full notional where no stop is set — a trade with no stop is treated as risking its full notional. This is not a view on strategy; it is the ceiling that stops the daily-loss floor (rule 2) and the drawdown floors (rules 3 and 4) being brute-forced in a single click. It publishes now and is enforced once the trading platform is wired.
If it is breached
The order is refused at entry once enforcement is live. Nothing already closed is reversed.

17. Reset-window conduct

Limit
Not permitted.
How it is measured
The daily-loss floor is measured on equity at 00:00 UTC and does not trail — rule 2 is unchanged. Deliberately structuring trades around that reset is a breach: carrying maximal risk across the 00:00 UTC boundary to reset the floor beneath a losing position, or generating a disproportionate share of profit in a narrow window around it. It is reviewed under the same trade-correlation logic as rule 10 — timing, instrument, direction, size — and needs no additional platform wiring beyond that review.
If it is breached
Reviewed as a rule 10-style correlation pattern; a deliberate pattern voids the evaluation and is recorded in the public denial ledger citing this rule.

18. Passing on probability, not skill

Limit
Not permitted.
How it is measured
Opening evaluations in quick succession, or across the 3-account cap and its retries, in a pattern that deliberately sacrifices accounts to let variance carry one of them to a pass, is a breach of the same identity-cluster rule that rule 10 governs. A retry is for a trader who came close, not a lottery ticket bought in bulk. It extends rule 10 and is reviewed the same way, and it punishes the exploit, not an honest trader who simply tried again.
If it is breached
Treated as a rule 10 identity-cluster violation: the manufactured pass is refused and recorded in the public denial ledger citing this rule. The fee-return doctrine is unchanged — an honest retry keeps every refund right in rule 3 and the fee-back on first payout.

When the price gaps through your stop

Weekend and news gaps fill at the first available price. That is how the market works and we are not going to pretend otherwise. What we will not do is let a gap on its own end an evaluation.

Where a limit is crossed solely because of a gap — that is, where the position would not have breached the limit had a price existed between the two ticks — the breach is reviewed automatically and reversed. The trader does not have to notice it, argue for it, or open a ticket. The trigger thresholds are published rather than kept internal:

Major FX
gap > 12 pips
Indices
gap > 0.4%
Metals
gap > 0.5%
Review
automatic, same day

Gaps that only contribute to a breach are reviewed individually and the outcome is recorded either way. Reversal is not discretionary generosity; it is a rule with a number attached, which is why the number is here.

When our platform is the problem

Any breach recorded during a declared outage is void. Positions open during an outage are closed at the pre-outage price where that price is better for the trader. Every incident is posted with timestamps, whether or not anybody noticed it and whether or not it cost anyone anything.

How a rule change works

Rule sets in this industry move. No firm in our research publishes a change log, several 2026 amendments reached accounts that were already open, and the current page is normally the only published page. The mechanism against that is not a promise of good behaviour. It is a version number.

  1. Every account records the rulebook version it was opened under.
  2. That version governs the account for its entire life. A later version cannot reach back into it, including where the later version would be more favourable — consistency matters more than the direction of the change.
  3. Every published version stays online at its own address, so the document an account is bound to remains readable rather than being replaced.
  4. Changes are listed in the changelog below with a date and a one-line description of what moved and in whose favour.

If we get it wrong

A breach decision or a payout decision can be disputed. A named person answers within one business day with the tick data the decision was made on — not a summary of it, the data. Where a dispute is not resolved, the outcome goes to the public record on the payouts page along with everything else.

Until there is a product, the address for anything at all — including a factual error on this page — is hello@upme.com.

Changelog

Rulebook version history
VersionDateWhat changed
1.2 2026-08-03 Raised the phase-one profit target from 8% to 10% of the starting balance — a round target, harder to reach. Phase two stays 5% and the one-step route stays 10%. A restriction on traders, tightened before any account exists. Supersedes v1.1, archived at v1.1.
1.1 2026-08-03 Added feed-integrity (rule 12), execution-realism (rule 13), trade-duration and account-activity (rule 14), event-straddle (rule 15), single-position-exposure (rule 16), reset-window (rule 17) and probability-pass (rule 18) rules — the seven anti-gaming grounds that close the ways a simulated venue can be gamed rather than traded. Several enforce once the trading platform is wired and say so inline. Clarified the automation boundary in rule 9: an EA remains permitted unless its function is to exploit the feed, the simulated fill or account hyperactivity. All restrictions on traders, added before any account exists. Supersedes v1.0, archived at v1.0.
1.0 2026-08-03 Added the best-day and best-trade consistency caps (rule 11a and rule 11b): no single trading day and no single trade may account for more than 40% of the total profit at the moment the target is reached. Raised the minimum trading days from 3 to 4. Both are restrictions on traders — a pass now has to reflect consistent skill, not one lucky day or trade — added before any account exists. Replaces v0.11’s stated absence of a consistency cap; the superseded text is archived at v0.11.
0.11 2026-08-02 Revised the review clause of rule 10: a payout under an active review pauses the 24-hour clock in public, the review is capped at 72 hours, and at the cap a decision is mandatory — settle with the +2% penalty, or refuse citing the violation. Replaces the unconditional +2% self-penalty, which pressured review speed over review accuracy. In upme’s favour on the penalty, in the trader’s on visibility. The superseded text is archived at v0.10.
0.10 2026-08-02 Added rule 10, one trader, counted once (account multiplication), and the matching refusal ground on the payouts page. A restriction on traders; the 24-hour review clause inside it binds upme. Added before any account exists.
0.9 2026-07-31 First public version. Published before launch, before any account exists, and before there is anything to dispute.

The first changes are already in the table, made while there was still nobody they could apply to. Every later row has to look like these: dated, described, and stating in whose favour it runs.

Questions people ask about rules like these

The measurement details that decide arguments, and the terms the rulebook commits to — each answer taken from a rule above or a section of this document. Open any one; they work with scripting off.

Does unrealised profit and loss count against the daily loss limit?

Yes. The daily loss limit is measured on equity, which includes open positions. It is measured against equity at 00:00 UTC and does not trail an intraday high, so a profitable morning does not raise the floor for the afternoon. It resets on the hour, not on a rolling 24-hour window.

Is the maximum drawdown trailing or static?

On the evaluation it is static: a floor set once from the starting balance that never moves, so profit you have made is yours to give back without ending the evaluation. On a funded account it trails your highest closed equity and then locks permanently once it reaches the starting balance.

How many trading days does the evaluation require?

4. A day counts when at least one position is opened or closed. The days do not have to be consecutive, there is no upper bound, and no minimum volume is attached to a day. v1.0 raised this floor from three days to 4. Until you have traded on 4 days the profit target simply cannot be claimed yet; nothing is lost and nothing else changes.

Is there a consistency rule?

Yes, from v1.0. A pass has to reflect repeatable skill, so two caps apply at the moment the profit target is reached: no single trading day may account for more than 40% of the total profit (rule 11a), and no single trade more than 40% (rule 11b). Earlier versions ran with no consistency cap and said so; v1.0 reverses that, dated in the changelog. Neither cap ever ends an evaluation — the target simply cannot be claimed until the profit is spread within both.

What trading practices are forbidden?

v1.1 adds seven anti-gaming rules (12–18) on top of the numbered limits. They forbid trading against the price feed rather than on it (rule 12), keeping a simulated fill that could not have happened against real liquidity (rule 13), and bracketing a scheduled news release with opposing pendings to harvest the gap either way (rule 15). Rule 14 caps how much profit very short trades may carry and how many trades or requests an account may fire in a UTC day; rule 16 caps single-position risk at 2% of the starting balance; rules 17 and 18 extend the daily-reset and account-multiplication logic to conduct around 00:00 UTC and to sacrificing accounts to buy a pass on variance. Some of these enforce only once the trading platform is wired, and each says so inline rather than pretending it is already switched on.

Can I run an EA or automate my strategy?

Yes. Automated execution is permitted and rule 9 is unchanged on that. The one boundary, clarified at v1.1, is on what an EA is built to do: an EA that automates how you trade is fine, but an EA whose function is to exploit the feed (rule 12), the simulated fill (rule 13) or account hyperactivity (rule 14b) is not automation of a strategy — it is a prohibited practice run by a robot, and it is refused as the practice, not as automation.

What is the profit split on a funded account?

80% of simulated profit is paid to the trader, from the first dollar. There is no split ladder that starts lower and improves with tenure, and no performance tier to climb. The number is written into the rulebook rather than left to a dashboard.

Is there a cap on how much I can be paid in a month?

Payouts from a funded account are capped at 2% of the simulated account size per calendar month, with full carryover. Nothing above the cap is forfeited or shaved: the amount queues, stays visible in the account, and pays out in following months. The cap is what keeps the 24-hour settlement and the flat 80% split solvent in the worst month, not just the average one. The one time carryover is lost is a rule 4 breach, because the queue lives with the account.

Can I get the evaluation fee back?

Two refund rights are written into the rulebook. The evaluation fee is refunded in full with your first payout, so a trader who reaches a payout has paid nothing net to be evaluated. Separately, an account that is purchased but never activated can be refunded within 14 days.

Can I trade more than one upme evaluation at the same time?

Yes, up to 3 concurrent evaluation accounts per person. The cap binds to a verified identity at payout and to correlated behaviour before that. What voids an evaluation under rule 10: opposite positions across accounts that net to roughly zero while one side harvests a pass or payout, an account transferred or traded by a third party, or the same trades duplicated across a cluster of accounts. A violation is recorded in the public denial ledger citing the rule.

Can upme change these rules after I open an account?

Not for an account that is already open. Every account is bound to the rulebook version it was opened under for its entire life, every version stays published at its own address, and a change binds only accounts opened after it.

How do I pay for a challenge?

Checkout is crypto-first: BTC, ETH, USDT, USDC and LTC are accepted, with more rails to follow. No card networks or app-store billing are shown anywhere on the site, because none are taken. A chargeback or payment dispute filed in bad faith voids the account and any pending reward; the crypto-first checkout makes that near-moot today, and the clause binds any future fiat rail.

Can I buy a challenge today, and are these rules in force?

A free account is open now: it holds your dated seat, your public forecast record and your place in the queue for the first cohort. The paid evaluation itself opens at rulebook v1.2. The rulebook is published and versioned in advance, and every account is bound for its whole life to the version it opens under, so the terms in it are on the record before the first one is sold.