13 NOVEMBER 2026 · 66 HAZEM SIGNAL · 5M SELL · SWEEP
A Gold SWEEP Reported TP3 While Two Opposing BUY Setups Reported Stops
A five-minute gold SWEEP SELL reported all three targets in sixteen minutes. Its activation-time price discrepancy and two overlapping BUY stop reports show why a successful endpoint still needs a careful audit.
Open CyGuru Gold Terminal to inspect signal plans and lifecycle events →
The production evidence
This selected favorable example is an indicator-reported reference observation. The production gateway retained a separate activation report and an actual TP3_HIT with the same signal identity. Engine 66-HAZEM-V1, payload schema 1, OANDA:XAUUSD, 5-minute timeframe, SELL and SWEEP were recorded. A separate script build hash was unavailable.
Signal ID 66-HAZEM-V1|OANDA:XAUUSD|5|1794602100000
TP3 event ID 66-HAZEM-V1|OANDA:XAUUSD|5|1794602100000|1794604260177|TARGET
Production feed observed through .
All 5 production messages retained the same original Entry, SL and three targets, along with the same creation and expiry timestamps and setup fields. The event trail contains no earlier stop or invalidation for this identity. The receiver timestamps below come from the matching stored rows.
| Field | Level | Distance | Original risk multiple |
|---|---|---|---|
| Entry | 3877.937 | — | — |
| Original SL | 3880.161 | 2.224 adverse | 1.000R |
| TP1 | 3875.390 | 2.547 favorable | 1.145R |
| TP2 | 3872.555 | 5.382 favorable | 2.420R |
| TP3 | 3870.527 | 7.410 favorable | 3.332R |
One reference R is the absolute distance from the original entry to the original stop. Each target multiple divides the distance from entry by that fixed denominator. The ratios describe the proposal’s price levels. Broker fills, commissions, spread, slippage and P&L are unverified.
| Recorded stage | Source time | Receiver time |
|---|---|---|
| SIGNAL_CREATED | ||
| SIGNAL_UPDATED · activation | ||
| TP1_HIT | ||
| TP2_HIT | ||
| TP3_HIT |
Replay of the site’s retained one-minute ranges
The signal-specific collector audit returned 17 OANDA:XAUUSD one-minute bars. The audited window has no missing minute slots. For a SELL, a low at or below a target counts as a nominal touch and a high at or above the original stop counts as a nominal stop. An earlier stop contradicts target-first ordering. A bar touching both requires finer verified observations to resolve the order. The window runs from the activation minute through the TP3 report minute. The activation minute is partial relative to the event clock and contains neither the stop nor TP1.
The maximum retained high in that window is 3879.750, below the original stop. No stop-before-target or same-bar stop/target conflict appears in these ranges. The site’s collector and the indicator both rely on TradingView inputs. An independently authenticated matching OHLC series was unavailable, so the replay remains a diagnostic on the collector data.
| Threshold | Minute open, UTC | High | Low |
|---|---|---|---|
| TP1 | 2026-11-13T21:01:00+00:00 | 3876.085 | 3873.975 |
| TP2 | 2026-11-13T21:08:00+00:00 | 3873.190 | 3872.120 |
| TP3 | 2026-11-13T21:10:00+00:00 | 3871.720 | 3869.530 |
The reported activation and the entry range do not coincide
The five-minute SWEEP SELL was created at 20:35 UTC on 13 November 2026. The script reported activation at 20:55:00.178, after a twenty-minute wait. Yet the retained 20:55 minute ranges from 3878.360 to 3879.750, entirely above the proposed SELL entry of 3877.937. The next minute also stays above entry. The first post-report collector range containing entry opens at 20:57, with a low of 3877.535 and a high of 3879.190.
This discrepancy limits what activation can mean in an evidence article. It confirms that the indicator sent an activation message; it does not authenticate a broker fill at the original entry. Different evaluation logic or inputs could explain the discrepancy, but the available records do not establish its cause. The event identifier contains FILL, while its payload type is SIGNAL_UPDATED with activation_reported=true. Neither wording supplies an execution receipt.
The stop-order diagnostic begins with the reported activation minute. That range starts 0.178 seconds before the event timestamp, making it a partial minute relative to activation. It contains neither the stop nor TP1. Starting instead at the first later collector range containing entry leaves the target-first finding unchanged in these retained bars. No intraminute fill sequence is reconstructed.
All three targets have separate reports
TP1 was reported at 21:02:00.175 UTC, TP2 at 21:09:00.206 and TP3 at 21:11:00.177. The activation-to-TP3 source interval is 15 minutes and 59.999 seconds. This case has no missing individual TP1 or TP2 timestamp. The first retained minutes containing those thresholds open at 21:01, 21:08 and 21:10 respectively. Each source report follows the close of the corresponding first containing minute by roughly two-tenths of a second.
That alignment is observable, but it does not prove the script runs exclusively at candle close. A bar gives a range over sixty seconds; the alert gives a reporting instant; the receiver clock gives the time the site accepted that message. These clocks answer different questions. A reader should not convert the minute opening into an exact target-touch time or treat the reporting interval as network latency.
The complete audit response includes a wider context window with missing slots after 22:00 UTC. The selected 20:55–21:11 replay window has all seventeen minute bars. Later gaps do not create missing observations inside that earlier diagnostic, and no market-closure cause is inferred from absent bars.
A frozen denominator and a narrow adverse margin
The original risk distance is 2.224 USD per troy ounce. Entry-to-target distances are 2.547, 5.382 and 7.410, corresponding to approximately 1.145R, 2.420R and 3.332R. All five production messages preserve these original levels. This freeze check prevents an apparently favorable target multiple from being manufactured by moving the stop or entry after the proposal.
The maximum retained high in the replay window is 3879.750, leaving 0.411 below the stop at 3880.161. That clearance is about 18.5% of the initial reference risk. The first TP3-containing minute, 21:10, has a low of 3869.530, below the 3870.527 target. These are quote-range comparisons. They do not establish executable bid/ask prices, position size, partial exits or money earned.
No independently authenticated matching OHLC source was available for this audit. The site collector identifies OANDA:XAUUSD one-minute inputs delivered through TradingView; the signal uses a five-minute setup on the same named symbol. Replaying those minute bars is a useful consistency check, but shared upstream inputs prevent describing the two records as independent verification.
Two opposing setups reported stops during this same episode
A five-minute Hazem PULLBACK BUY, full signal ID 66-HAZEM-V1|OANDA:XAUUSD|5|1794603300000, reported activation at 21:01:00.878 UTC and STOP_LOSS_HIT at 21:08:00.175. A different five-minute SWEEP BUY, 66-HAZEM-V1|OANDA:XAUUSD|5|1794602700000, reported activation at 21:09:00.204 and a stop at 21:11:00.174. The second stop report is only three milliseconds before this SELL’s TP3 report. That proximity does not establish causality between scripts or an execution ordering across accounts.
Three older SELL records also have overlapping reported active windows at the cutoff: CyGuru Indicator 15-minute SWEEP 1794592800000, Hazem 15-minute TREND 1794583800000 and Hazem 15-minute SWEEP 1794577500000. Their complete identities and available event trails are retained in the snapshot. An overlap means shared market exposure under the stated activation-to-terminal-event rule; it does not establish simultaneous broker positions or independently successful bets.
Gold Terminal makes this comparison practical by retaining the original plan, symbol, timeframe and event history for each identity. Use its signal records to distinguish an opposing setup from an update to the same setup, and use Research Lab to examine what the stored evidence can support. A journal should include the stopped BUY observations alongside the selected SELL success when reviewing this market episode. This article selects a favorable reported endpoint and cannot supply an overall win rate or a future-profit claim.
The snapshot preserves 5 other records with overlapping reported active windows on the same symbol. The selection and shared price exposure preclude an overall win rate, a claim of independence or an inference about future profits.
Download the dated JSON evidence and replay inputs →
Use Gold Terminal to compare the original plans, setup identities and reported outcomes →
66 Hazem Signal dossier · Research methodology · Gold trading journal guide · Gold Intelligence · Follow Research Lab