Paper 08 · Resurrection file 02 · July 30, 2026

Tide lives.
The claim got stricter.

V-ENT was the candidate under test: wait for a real intraday washout, then buy the recovery. It passed its placebo at the 97.24th percentile and was still REJECTED for promotion — it added −$40 across 2017–2021 against +$3,614 across 2022–2026, so it fails both-half stability. This paper records that verdict. It is not the strategy being run forward.

Open the strategy Read the verdict →

CandidateTIDE V-ENT

VerdictREJECTED

Research size1 MNQ

Live ordersOFF

2026 · JAN 01 — JUL 29

Hypothetical one-MNQ replay of the REJECTED V-ENT candidate, after modeled commissions. The strategy actually being run forward is the Tide core — see the research record at /run.

$1,744.50SIMULATED NET

2.43PROFIT FACTOR

44.23%TRADE HIT RATE

−$275.50MAX CLOSED DD

The actual strategy

Buy recovery after proof of weakness.

Tide is a one-minute NQ signal evaluated with one MNQ in research. The V-ENT gate prevents the core and reclaim entries from buying a day that only extended upward: price must have closed below same-session VWAP on an earlier completed bar.

01 / CONTEXT

Washout first

At least one prior RTH one-minute close must print below the running VWAP. The signal bar cannot create its own permission.

02 / TRIGGER

Recovery confirms

The fast EMA crosses above the slow EMA inside locked time and VWAP-distance gates. Weak-range and open-chop filters still apply.

03 / EXECUTION

Enter next bar

No same-close fill. The engine enters on the following bar's open and gives the stop priority when stop and target collide.

04 / RISK

Fixed first, managed later

The base long starts with an 80-point target and 20-point stop. Existing managed-stop, max-hold, failure, and 15:58 flatten rules remain intact.

Evidence gates

Good enough to resurrect.
Not cleared for live.

The frozen 2017–2026 tape remains strong, and the fresh May–July replay matches the reference trade-for-trade. The random-control test is below our own threshold, so the status stays watch / paper-shadow.

FROZEN LIFETIME BACKTEST895 trades

May 2017 — June 2026 · NQ reference simulation

Simulated net: $117,670
Profit factor: 1.63
Trade hit rate: 40.67%
Start-zero DD: −$8,517

Fresh replay parity

16 / 16 exact

Fresh calendar violations

0 trades

IS and OOS improvement

Both positive

Leave-one-year-out

10 / 10

Random-control percentile

92.7th / needs 95th

Forward resurrection paper

0 / needs 60 sessions

Every month pays $250 + fee

FRESH BAR FILE63,094 one-minute bars

REPLAY WINDOWMay 26 — July 29

TRADE PARITY16 / 16 exact

BLACKOUT VIOLATIONS0

Subscription economics

The average works.
The promise does not.

One MNQ averaged $249.21 gross per month in the seven observed 2026 months. After a hypothetical $100 subscription that becomes $149.21, with March and June below zero. A fee can be evaluated; a $250-every-month promise cannot be supported.

MONTH1 MNQ GROSSAFTER $100RESULT

JAN$372$272CLEARS $250

FEB$174$74POSITIVE

MAR−$210−$310NEGATIVE

APR$420.50$320.50CLEARS $250

MAY$746.50$646.50CLEARS $250

JUN$19.50−$80.50NEGATIVE

JUL$222$122POSITIVE

DEFAULT PAPER SIZE1 MNQ

$249.21 average simulated gross · $149.21 after a $100 fee · 3 of 7 months clear $250 after the fee.

SENSITIVITY — NOT A RECOMMENDATION2 MNQ

$498.43 average simulated gross · $398.43 after a $100 fee · 4 of 7 months clear $250 · worst month −$520.

COMMERCIAL RULENo guarantee.

We do not increase size to manufacture the subscription claim. Pricing follows validated value and full loss disclosure.

Sharpened with iron

New ideas had to earn entry.

Two pre-registered coverage modules attacked Tide-idle days using completed five-minute trend and one-minute triggers, next-bar fills, five-bar maximum holds, and same-gate random controls. Both lost. They stay out.

5M CONTINUATION / 1M BREAK−$29.00

Trades: 34
PF: 0.95
Null pctile: 71.1

REJECT FOR DEPLOYMENT

FIRST-HOUR BREAKOUT / 5M TREND−$366.50

Trades: 28
PF: 0.37
Null pctile: 3.85

REJECT FOR DEPLOYMENT

ORB-G-10 / OLD SLEEVE−$192.00

Tide DD: −$275.50
Book DD: −$652.50

DISABLED · PROBATION

The autonomous loop

Research can nominate.
It cannot promote itself.

Every daily run begins with immutable inputs and ends with a written decision. Any model can operate the loop because fill rules, null tests, rejection gates, and the authority boundary are explicit.

Freeze

Hash the bars, calendar, engine, manifest, and date before analysis.

Replay

Match every paper signal, entry, exit, and dollar result. Any mismatch stops the loop.

Diagnose

Classify wins, losses, avoided losses, no-trade days, and missed moves with causal context.

Lock

Write one exact hypothesis and its rejection threshold before seeing the result.

Attack

Run friction, split, regime, neighbor, leave-one-out, and 2,000 random-control tests.

Decide

Reject, keep researching, or nominate for human review. The loop cannot self-promote code.

Final verdict

Resurrect the edge.
Keep the certainty dead.

Tide V-ENT is materially better than the recent Tide + ORB research book, its fresh replay is exact, and its 2026 risk-adjusted simulation supports a serious paper campaign. It has not passed the 95th-percentile placebo gate, has zero new forward sessions, and cannot support a universal $250-per-month claim.

CHAMPION RESTORED ORB DISABLED LIVE PROMOTION BLOCKED INCOME GUARANTEE BLOCKED