🛡️ Risk management

Funded7's Payout Was Held Under OREF "Rule 1" | Explaining the MIN(median×2.5, Q3+1.5×IQR) Consistency Review With a Real Review Email [September 2026]

Published: 9/10/2026Updated: 9/24/2026

※ This article is based on a Funded7 payout review email sent to this site's owner on September 9, 2026, and Funded7's official FAQ (Rule 1, Rule 2 & 3, Rule 4). Funded7 itself states that the calculated values will change on future reviews, so the numbers in this article are "a snapshot at this point in time." Details of the trading strategy (logic, lot size, time of day) have been withheld.

📌 Added September 24, 2026: This account was subsequently terminated on September 23, 2026. It received a final settlement of ¥200,000, which arrived in full the following day. The complete record from the ¥2M account's evaluation through its termination (12 payouts, ¥1,147,040 total) is summarized in The complete record of Funded7's ¥2M account.

TL;DR: not a violation, but a "hold" — and the cause was "too consistent" lot sizing

  • On September 8, 2026, a Funded7 payout request came back "DELAYED" the next day. The reason: just one item — OREF Rule 1's R1A (position-sizing consistency) — wasn't met. Rule 2, Rule 3, and max DD all PASSed, and zero profit was invalidated
  • R1A's allowable threshold is MIN(median×2.5, Q3 + 1.5×IQR). The more uniform the lot sizes, the smaller the IQR shrinks, and this time the IQR was only 1.4% of the median, pushing the threshold to just 2.7% above the median
  • Sitting there was a position (BTCUSD) 4.3% above the median, making it 30 outlier trades → FAIL. A fixed-lot EA — which should be "the most consistent" — gets caught here precisely because notional value moves with price
  • Rule 1 is a review criterion, not a violation. Profit isn't confiscated, it doesn't count as a strike, and the entire target period is recalculated on the next review. The pending payout amount (¥54,774) is still sitting there, untouched
  • Passing requires widening the IQR until the threshold exceeds the largest position. Ironically, deliberately adding variance to lot sizes is the correct move here

What happened: two emails the day after the request

Date/time (JST)Content
September 8Payout requested. Auto-reply: "We have received your payout request — it is currently being processed"
September 9, 20:50Automated notice: "Important Notice Regarding Your Payout Request" (subject line: Payout Rejected. Only states that the team will send details)
September 9, 20:50Manual detail email from support: "Payout Review Result"

The second email's subject line, "Payout Rejected," is heart-stopping, but the content is boilerplate saying only "the team will send details." The real substance came in the third email, which included every calculated value used in the review — median, Q3, IQR, threshold, even the ID of the position that exceeded it. Among the 19 firms this site covers, we've never seen a payout review email disclose numbers this openly.

The account is a two-phase (JPY) ¥2M account, with roughly ¥1M already withdrawn. This review's target period was the 8 trading days from the day after the previous payout (August 31) to the request date (September 9).

What OREF is: Funded7's payout review runs on four rules

In 2026, Funded7 published a review framework called OREF (Optimal Risk Evaluation Framework). When we wrote our Funded7 complete guide, we assessed that "there's a consistency clause but no quantitative standard" — but it's now on the official FAQ with the formula spelled out. This article is also a correction of that earlier assessment.

RuleWhat it checksResult of a failure/violation
Rule 1: ConsistencyR1A position sizing / R1B trading activity / R1C profit quality (QC)Not a violation. The payout is held (Pro accounts); if you pass an evaluation without meeting it, you start at the Silver tier. No profit confiscation
Rule 2: Position riskWhether risk on a single position, correlated bucket, or the whole account exceeds the tier cap (Gold 3% / Silver 2% / Bronze 1%). Trades without an SL are estimated via ATR(14)×1.96Only the profit from the offending trade is voided (the loss stays). Repeated violations lead to a tier downgrade
Rule 3: Execution standardWhether closes under 15 seconds exceed 2% of all trades, or closes under 30 seconds exceed 3%Voids the profit involved and moves the account to Silver/Bronze
Rule 4: Prohibited strategiesMartingale, arbitrage, copying, account sharing, etc.Grounds for account closure

The key point is that Rule 1 is a different kind of thing entirely. Rules 2–4 are "things you must not do," but Rule 1 is "a state you should be in before withdrawing." If you don't meet it, you just wait — your profit doesn't shrink. That's exactly what happened here too: zero trades were subject to Rule 2 or Rule 3, and ¥0 in profit was invalidated.

The R1A formula: MIN(median×2.5, Q3+1.5×IQR)

Here's the official FAQ's definition.

Allowable Threshold = MIN (Median Criterion, IQR Criterion) Median Criterion = Median × 2.5 IQR Criterion = Q3 + (1.5 × IQR)

  • The target isn't lot count but notional value. A 1-lot EUR/USD position and a 1-lot gold position carry different exposure, so everything is converted to a monetary amount before comparison
  • Take the notional value of every position in the target period, and compute the median, Q1 (25th percentile), and Q3 (75th percentile). IQR = Q3 − Q1
  • The threshold is the smaller of "2.5× the median" and "Q3 + 1.5×IQR." Anything above the threshold is an "Inconsistent Outlier"

Reading "up to 2.5x is allowed" makes this sound lenient. We initially read it that way too, and had described Funded7's consistency rule as "loose." The catch is the second formula. Q3 + 1.5×IQR gets smaller the more uniform your lot sizes are.

Breaking down the actual review figures

ItemValuevs. median
Median notional value¥1,240,946-
Q1 (estimated: Q3 − IQR)¥1,231,323−0.8%
Q3¥1,248,417+0.6%
IQR¥17,0941.4%
Median × 2.5¥3,102,365+150%
Q3 + 1.5×IQR¥1,274,058+2.7%
Threshold adopted (the smaller one)¥1,274,058+2.7%
Largest notional value (BTCUSD, September 3)¥1,294,723+4.3%
Excess amount¥20,665-
Outliers30 / 1,005 trades3.0%

Median × 2.5 would have left ¥3.1M of headroom. But since the IQR was only ¥17,000, Q3 + 1.5×IQR came out to ¥1.274M — a ceiling just 2.7% above the median — and a position 4.3% larger than that got flagged with 30 outliers. FAIL.

Why "the most consistent" fixed-lot approach fails

This account was running BTCUSD with a fixed-lot EA. By lot count, it's perfectly consistent. But the review measures notional value.

Notional value = lot size × contract size × price (× JPY conversion rate)

If the lot size is fixed, notional value moves directly with price. If BTC swings 4% in a week, a same-lot position becomes 4% larger in notional value. Meanwhile, the IQR — "how spread out the middle 50% of positions are" — shrinks to 1–2% if most positions are clustered around the same price level. The tighter the distribution, the lower the ceiling, and only positions from a moment when price moved a bit end up piercing it. That's exactly the structure at play here.

  • Major FX pairs: weekly price moves of 1–2% tend to keep IQR and price variation on the same scale, even with a fixed lot
  • Gold, indices, BTC: 3–5% weekly moves are normal. A fixed lot lets price movement exceed the IQR and create outliers
  • Balance-proportional lots: sizing that scales up as the balance grows clusters the second half of the period's positions on the high side, which is also a breeding ground for outliers

In other words, R1A is designed to "prevent oversized gambling," but it has the side effect that the more uniform your lot sizing, the more you get caught by tiny wobbles in notional value.

R1B and R1C had passed

In the same email, the other two items under Rule 1 both PASSed. For reference, here are the numbers.

Sub-ruleStandardThis time
R1B trading activity10+ closed trades, 3+ trading days1,005 trades, 8 days
R1C profit quality (QC)(2nd-largest profit + 3rd-largest profit) ÷ 1st-largest profit ≥ 0.90(3,268 + 3,219) ÷ 3,300 = 1.97

The R1C QC score measures "whether you're relying on one big win," and multiple positions opened close together on the same symbol are bundled and calculated as a single trade (per the official FAQ's Trade Cluster definition). For EAs that stack up small wins, this generally isn't a problem.

How the pending payout amount was calculated

Even on hold, the money hasn't disappeared. Here's the email's calculation.

ItemAmount
Cumulative net P&L on MT5+¥1,048,366
Invalidated by Rule 2/Rule 3−¥0
Past withdrawals−¥998,800
Profit carried over from cTrader+¥5,208
Current amount available for payout¥54,774

This meets the minimum payout threshold, and the official FAQ states that once Rule 1 clears, the amount is "released immediately." The profit hasn't been taken away — it's been put in a queue, which is the accurate way to describe the state.

For reference, the review materials' Account Summary listed "Final Compliant P&L: ¥54,774" while the Decision Summary listed "Compliant P&L (All-Time): ¥1,048,366 / Available for Payout: ¥54,774" — two separate figures, and the email body carefully footnoted that "the former is cumulative, the latter is this payout's available amount."

Passing the next review: widen the IQR, or push up Q3

Rule 1 is recalculated at the time of the next request, using every trade in the payout target period. These 1,005 trades don't disappear. The question is what to add so the ceiling exceeds the largest position.

The condition is Q3 + 1.5×IQR ≥ largest notional value. With Q3 unchanged, the IQR would need to be roughly ¥31,000 (2.5% of the median) — 1.8x the current figure.

Option 1: deliberately add variance to lot sizes

Randomizing the lot size by ±10–15% around a base lot brings the IQR to roughly a tenth of the base, opening the threshold to Q3 + 15% ≈ +20% above the median. That would absorb a largest position 10% larger even with ±5% price movement. The median×2.5 side stays far away, so that threshold wouldn't be hit either.

"Add variance to your lot sizes as a consistency-rule countermeasure" sounds counterintuitive, but under this formula it's the correct move. For a fixed-lot EA, it's as simple as adding one line of randomization to the lot calculation.

Option 2: stack up trades at the same size as the current largest

Adding positions to the upper end of the distribution raises Q3 and widens the IQR too. Running the numbers here, adding roughly 150–200 trades around the ¥1.29M size would push the ceiling past the largest position (a rough estimate assuming a linear distribution shape). At a pace of 125 trades/day, that's 1–2 days' worth.

That said, if BTC keeps rising while the lot stays fixed, a new largest position will emerge and you're back to square one. Combining this with Option 1 is the safer route.

Things to watch

  • Lower-end outliers: it's unclear from the email whether positions below Q1 − 1.5×IQR (roughly ¥1,205,682 this time) count as outliers too. Better not to create too-small positions either
  • Funded7 doesn't give individual advice: it explicitly states "this doesn't mean adjusting by exactly ¥20,665 guarantees a pass" and "approval on the next review isn't guaranteed." Assume every number is a moving target
  • Compatibility with Rule 3: while stacking up trades, don't exceed the 2%/3% caps on closes under 15/30 seconds

Checklist for EA traders running Funded7

  1. Don't fix your lot size. Add randomization of base lot × (0.85–1.15)
  2. Compute your own notional-value distribution. Pull "lot × contract size × execution price × JPY rate" from MT5 history and check the median, Q1, Q3, and max. Calculate the threshold — MIN(median×2.5, Q3+1.5×IQR) — yourself
  3. Keep your largest position within 80% of the threshold. Leave headroom for price movement
  4. Be extra careful in weeks when a symbol moves a lot. Fixed-lot gold, BTC, and index positions create outliers from price movement alone
  5. Watch QC 0.90 and Rule 3 at the same time. Note that QC bundles closely spaced positions on the same symbol into one
  6. Run the above calculation before requesting a payout. A hold itself carries no penalty, but it wastes one review cycle

This site's P&L tracker collects trade history per account, but a notional-value distribution calculation isn't in there yet. For now, export your MT5 history and compute the median, Q1, and Q3 in a spreadsheet.

Assessment: high transparency, but the formula that punishes "too much consistency" has room to improve

The good part is disclosing every calculated value in full. With the median, Q3, IQR, threshold, and even the offending position ID all spelled out, traders can recalculate for themselves and take countermeasures. Compared to a firm that just says "consistency violation" and stops there, this is far more honest. Not treating a Rule 1 miss as a violation, and not shrinking profit, is also commendable design.

What's concerning is the formula itself. Q3 + 1.5×IQR is lifted straight from statistical outlier detection (Tukey's fences), and it gets stricter the tighter the distribution is. Against a stated goal of "preventing oversized gambling," it seems at odds with the design intent that the fixed-lot EA furthest from gambling gets caught first. Standardizing on median×2.5, or putting a floor on the IQR (say, 5% of the median), would fix this, and we'd suggest the improvement to Funded7.

Other firms' consistency rules mostly follow a "largest profit day ÷ total profit ≤ 30–50%" model, which looks at the skew of profits (complete guide to consistency rules). Funded7's R1A looks at skew in size, not profit — a different animal — so an EA that passed elsewhere could still get caught here.

FAQ

Q. If I FAIL Rule 1, is my profit confiscated?

No. The official FAQ explicitly states "not meeting Rule 1 is not a violation and doesn't count as a strike," and in this case, ¥0 was invalidated under Rule 1. The available payout amount is simply held and gets released once you meet the standard on the next review.

Q. How many outliers trigger a FAIL?

This time, the stated reason for FAIL was "the largest notional value exceeded the threshold by ¥20,665." Since the official FAQ also calls anything over the threshold an "Inconsistent Outlier (violation)," the safe reading is that even one position over the threshold triggers a FAIL.

Q. What's the target period for the review?

From the day after the previous payout to the current request date. This time it was August 31 to September 9. The period extends on the next review, and the recalculation includes the trades already covered.

Q. Does this matter for manual traders too?

Yes. The official FAQ's own example is aimed at manual traders: "someone whose threshold is $140,000 in EUR/USD who opens a $230,000 notional position in gold would be in violation." Watch out for the fact that switching symbols can spike your notional value even at the same lot size.

Q. Is Rule 1 checked during the challenge too?

It's assessed at the point you pass. Per the official FAQ, if you pass without meeting Rule 1, you start on a Funded account at Silver (2% risk cap) instead of Gold, and can be promoted after 3 payouts.

Q. What's the "ATR risk" under Rule 2?

For a position opened without an SL, the "statistical estimated loss" is calculated as ATR(14)×1.96×lot×contract size, and if that exceeds the tier cap (3% for Gold), it's a Rule 2 violation. If an SL is set, the calculation uses the distance to the SL, and a gap that fills worse than the SL doesn't count as a violation.

References

prop-memo.com tools

Written by

Hosono P | the prop firm strategist

I buy challenges with my own money and record everything through to the payout. Recorded payouts: ¥6.1M in total from Fintokei, Fundora and Funded7, plus $4,776 from The5ers (as of September 2026). Author of the semi-discretionary EA "ELDRA".

Profile and payout recordX @hosono_p

📚Related articles

🛡️

Prop Firms That Won't Let You Change Strategy: Official Rules Checked | TradingCult Bans EAs Outright, Funded7 Disqualifies Running the Same EA Across Multiple Accounts [September 2026]

TradingCult bans EA and bot use outright, and a violation is a hard breach. The firm and FundedElite also explicitly ban 'account rolling.' At Funded7, even your own EA gets you disqualified for Group Trading if its results correlate too closely with another trader's. SuperFunded allows news trading freely during the evaluation but bans it within ±10 minutes only at the Funded stage, and FundedElite's strategy risk limit only applies at the funded stage — this article rounds up 9 firms whose rules change between the challenge and funded stages, based on each firm's official FAQ.

Risk management9/5/2026
🛡️

Can You Change Strategy Mid-Way? | I Checked 20 Firms' Consistency Rules and Found They Go in Opposite Directions [September 2026]

FundedNext explicitly states you must 'maintain the same strategy through your challenge and Funded account,' banning switching from an EA that passed the challenge to manual trading. Fintokei, meanwhile, states plainly it has no consistency rule and won't add one retroactively. Even within one firm, Hantec's Enhanced has no restriction while EnhancedX caps you at 35% — opposite treatment. Here's what to check before changing your method, across 20 firms.

Risk management9/4/2026
🛡️

When Should You Change Your Prop Account Settings? Don't Touch It During Evaluation — Always Change It the Moment You Get Funded

When a challenge isn't going the way you'd hoped, you want to raise your lot size. You want to change your trading hours. But that judgment call has no statistical basis. Telling a 40% pass-rate method apart from a 25% one needs 150 accounts per method, and most people only have one or two. Meanwhile, the moment you reach a funded account, rules that didn't exist during evaluation kick in all at once — Hola Prime caps risk per trade at 2%, Hantec treats the 3 minutes around a news release as a violation even to close a position. These are cases where you must change something. We sorted out what to keep fixed, what to change, and how to treat an account that's already deep in drawdown, using 40,000 Monte Carlo runs. As of September 18, 2026.

Risk management9/18/2026
🛡️

What Dan Cheung Means by "Rotation": It's Not About Doing More Trades — It's About Containing Losses to One Account

We looked into Dan Cheung's (London-based) account rotation, a term that comes up a lot in prop trading circles, based on his own public statements. It boils down to three lines: touch only one account at a time, close that account and move to the next once you hit your daily profit target or your stop, and never chase a losing account. People often mistake it for a way to trade more, but the real point is containing losses to a single account. We then checked whether this actually holds up under the terms of firms available from Japan, across four axes: consistency rules, minimum trading days, allocation caps, and multi-account clauses.

Risk management9/17/2026
🛡️

The Same Strategy Can Take 3x Longer Depending on Where You Run It: Testing 8 Firms' Rules Against a Mechanical Strategy [Analysis]

Profit targets and max DD line up almost identically across firms, yet running the same strategy through their rules stretches the cycle time from 17.2 days to 59.6 days — a 3.5x spread. The cause is the '1% risk per trade' rule. Pass rates barely differ (33.7–38.7%), while firms with a 10% Phase 1 target are 5 points worse off at the first stage, and a 60% profit split cuts expected funded-stage income by 40%. Also checks Hantec's 3-minute rule, which is actually a ratio condition — 'net profit from trades closed under 3 minutes is 30% or more of total' — against 3,826 real trades' holding-time distribution.

Risk management9/4/2026
🛡️

What Percentage of Prop Phase 1 Would a Coin Toss Pass? The Exact-Solution Answer Comes Out Above 50% [Analysis]

An exact calculation of the Phase 1 pass rate for a prop challenge using nothing but a ±¥200,000 coin toss. With zero fees, an 8% target / 10% max DD setup passes 55.6% of the time, and FTMO comes out at exactly 50.0%. With a 5% fee it drops to 43.2%, and it sinks into the 8% range for 1-step plans with a 3% daily DD. Compared with real-world pass rates of 5-14%, the average trader is losing to a coin toss.

Risk management8/31/2026