Open Market JournalOpen book
XAUUSDReviewCautious

GOLD Session Review — August 11, 2026

Overnight Breakout Trap Reverses Into a Close Near the Lows

Gold's overnight breakout to a fresh $4,435.00 multi-week high failed to hold and fully reversed, breaking every mapped support level to close at $4,367.23 — down $21.09 (-0.48%) on the day and within $4.47 of the prep's secondary $4,352 pivot. The prep's 50%-weighted continuation lead did not fire, but its lower-weighted give-back branch and every named support level played out almost exactly as mapped; the carry-forward is to weight overnight-already-extended breakouts toward reversal risk rather than continuation, especially when the driver stack reads only partially confirming.

Prep outcomepartial
Lead scenario50% · missed
Leanneutral · incorrect
Day typetrend → trap
Surprisemoderate
Grade card7 of 10 correct

Every prep call is graded like this. See the full track record →

Session chart
XAUUSD — session vs prep levelsH1
Shaded band: the reviewed session. Price lines: the preparation's key levels — see which held and which broke.
Powered by Cortiq

Graded against live session data from Cortiq

Every call in this review is checked against the session data captured in Cortiq — the AI trading workspace that powers our daily preparation, review, and calibration loop.

See how Cortiq works
Session
GOLD Overnight-Breakout Reversal Session
Symbol
XAUUSD
Window
00:00 – 23:00 UTC
Day type called
Trend (continuation)
Day type actual
Trap
Lean outcome
Incorrect (Neutral/Wait lean; session resolved directionally lower)
Regime
False breakout reversal — broke every mapped support level, closed near the day's low
Preparation
Partially accurate
Surprises
Moderate

Grade card

7 of 10 correct
  1. Day-type callIncorrect
    Called
    Trend (continuation)
    Actual
    Trap — the overnight spike above $4,419-4,420 never held a second H1 close and fully reversed into a breakdown through every mapped support
  2. LeanIncorrect · neutral-incorrect
    Called
    Neutral/Wait
    Actual
    Session resolved directionally, closing -$21.09 (-0.48%) near the day's low
  3. Conditional lean — long legCorrect · no trigger
    Called
    Long on a held H1 close above $4,419-4,420
    Actual
    Only the 04:00 UTC close ($4,429.55) cleared the trigger; the very next hour closed back below it ($4,410.68) — never held
  4. Conditional lean — short/give-back legCorrect
    Called
    Short/give-back on a held H1 close below $4,388
    Actual
    Trigger fired and held (08:00 → 09:00 UTC closes at $4,375.30 / $4,366.11); price extended to $4,356.53 and closed at $4,367.23
  5. Lead scenario (Continuation, 50%)Incorrect
    Called
    Overnight breakout extends through NY, targets $4,450/$4,470-4,480
    Actual
    Breakout printed intraday but never held; price reversed into the give-back path instead
  6. Key level $4,419-4,420 (resistance/pivot)Correct
    Called
    The critical breakout-continuation test
    Actual
    Touched and exceeded intraday (spike to $4,435) but never closed above on a held basis — rejected exactly as a resistance test, not confirmed as a break
  7. Key level $4,393 (support)Correct
    Called
    Earliest tell of stalling momentum
    Actual
    Broken and held below from 08:00 UTC onward
  8. Key level $4,388 (support, today's open)Correct
    Called
    First genuine reversal signal if lost on a held close
    Actual
    Lost on a held close (08:00 → 09:00 UTC) — the reversal materialized exactly as flagged
  9. Key level $4,371.60 (Friday's high/support)Correct
    Called
    Live only if $4,388 fails on a held close
    Actual
    $4,388 failed; price traded through $4,371.60, bounced, then closed back below it at $4,367.23
  10. Key level $4,352 (secondary pivot)Correct · within noise
    Called
    Live only on a genuine give-back reversal
    Actual
    Session low $4,356.53 — within $4.47 of the level, effectively tested

The tape

  1. Open (first 30–60 minutes)

    The session opened at $4,388.32 and immediately extended the prior three sessions' gains, pushing through $4,400 within the first hour (01:00 UTC close $4,399.15). Price kept climbing through the Asian book with no pause at the $4,393 pullback level from the prep's own key-levels table.

  2. Mid-session

    The rally peaked at $4,435.00 during the 04:00-05:00 UTC window — beyond the prep's $4,420 continuation trigger and closing in on the $4,450 target — but only a single H1 close ($4,429.55 at 04:00) cleared the trigger before the next hour rolled back below it. The reversal accelerated during the London open: the 08:00-09:00 UTC window, which the prep explicitly flagged as a no-trade Judas-prone zone, is where price broke and held below both $4,393 and $4,388, dropping to a session low of $4,356.53 by 10:00 UTC — within $4.47 of the prep's secondary $4,352 pivot. A modest recovery into the 13:00-15:00 UTC NY/COMEX window pushed price back above $4,388 and tested $4,400 (high $4,403.96 at 16:00 UTC) but never closed above it, leaving the give-back scenario's own invalidation level untouched.

  3. Late / close

    The bounce failed and price rolled over again into the New York afternoon, printing a fresh intraday low close to the session low (18:00 UTC low $4,369.02) before chopping through the flagged 22:00 UTC "Asian-resume" trap window without renewed continuation in either direction. The session closed at $4,367.23 — below the open, below $4,393, below $4,388, and below $4,371.60, with only the $4,352 pivot and the deeper $4,300-4,320 flip zone left unbroken.

What we learned

The session unfolded broadly within the prep's own scenario map — the give-back branch and every key level it named worked — but the headline day-type and lead-scenario calls were wrong, and three specific gaps drove that miss:

  1. Day-type precondition needs to weigh "already extended overnight" against "partial driver confirmation" more heavily

    The prep called Trend/continuation off a real-yield driver that agreed and a dollar driver that had only stabilized — explicitly "partial" alignment — while price had already broken to a fresh high before London liquidity arrived. Trend-day calls should require full driver-stack confirmation, not one confirming leg plus a merely stabilizing one, when the move is already extended before the session's primary engine opens. This routes to re-examining the day-type precondition check.

  2. The breakout base rate was applied to the wrong kind of setup

    GOLD's "H1 ranges break out, not revert" prior is built for fresh in-session coils, not for a move that already ran through the Asian session unchecked. Applying it here inflated the continuation branch's weight (50% vs. the give-back branch's 20%) on a setup that was structurally closer to an overextended spike than a fresh range. This routes to a priors refinement: distinguish a fresh breakout candidate from an already-extended overnight one, and shade the latter's weighting toward reversal.

  3. The right branch was under-weighted, not missing

    The give-back scenario named the exact trigger ($4,388 lost on a held close), the exact path ($4,371.60 then $4,352), and it fired almost to the dollar. A driver stack that reads "partial alignment, weaker than the prior day's read" on top of an already-extended overnight move should shift weight toward the reversal branch rather than simply capping the lead's conviction at 50%. This routes to tightening the tradability/weighting standard for partial-alignment driver stacks. Footer: Reviewed prep: 2026-08-11-xauusd-session-preparation