Open Market JournalOpen book
EURUSDReviewCautious

EURUSD Session Review — August 26, 2026

Durable Goods Beat Drives a Six-Pip Miss at the Digestion Floor Ahead of Jackson Hole

EURUSD opened Wednesday at 1.16716 and closed at 1.16515, a 20-pip net decline that stayed just inside the week-old 1.16353-1.16824 digestion band -- the range/digestion co-lead (30%) technically held even as the session's directional character matched the hot-PCE breakdown branch (27%) far more closely. Core PCE and GDP printed dead on forecast, giving neither directional branch its confirming surprise, but a real beat in Durable Goods Orders (1.1% vs 0.8% forecast) supplied the dollar impulse the prep's driver stack flagged as a compounding risk. The durable move formed exactly where the priors predicted -- the 15:00-17:00 UTC delayed-resolution window -- driving price to a session low of 1.16416, just 6.3 pips short of confirming a held break of the 1.16353 floor before the day stabilized into the close.

Prep outcomepartial
Lead scenario30% · hit
Leanneutral · incorrect
Day typeevent-suspended → range
Surpriselow
Grade card7 of 9 correct

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

Session chart
EURUSD — 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
EURUSD Pre-Jackson Hole PCE Wednesday Session Review
Symbol
EURUSD
Window
00:00 - 23:59 UTC
Day type called
Event-suspended
Day type actual
Range/digestion
Lean outcome
Incorrect (neutral-incorrect: a real, catalyst-driven grind resolved close to structure)
Regime
Genuine event-driven test of the digestion floor that stopped 6.3 pips short of confirming
Preparation
Partially accurate
Surprises
Low

Grade card

7 of 9 correct
  1. Day-type callCorrect
    Called
    Event-suspended (compression pre-print, print-driven resolution)
    Actual
    Compressed ahead of the print, then a genuine catalyst-driven leg formed post-print -- resolved into one of the call's own mapped branches (range/digestion)
  2. Lean (unconditional)Incorrect · neutral-incorrect
    Called
    Neutral/Wait
    Actual
    Close 1.16515 vs open 1.16716 -- a 20.1-pip net decline, with a coherent 30-pip intraday break in the delayed-resolution window that stalled 6.3 pips from confirming structure
  3. Conditional lean -- long triggerCorrect · no trigger
    Called
    Long on a held, displaced H1 close above 1.16824 surviving through 16:00 UTC
    Actual
    Never approached -- session high (1.16771) fell 5.3 pips short of the trigger
  4. Conditional lean -- short triggerCorrect · no trigger, near miss
    Called
    Short on a held, displaced H1 close below 1.16353 surviving through 16:00 UTC
    Actual
    Near miss -- session low (1.16416) came within 6.3 pips but no H1 close ever printed below 1.16353
  5. Lead scenario -- range/digestion co-lead (30%, no operative lead vs the 28% soft-PCE branch)Correct · co-lead fired; map quality below
    Called
    Neither directional trigger fires; price continues rotating inside 1.16353-1.16824
    Actual
    Technically true (no trigger fired, price stayed in-band), but the session traced a one-directional 30-pip grind toward the floor, not two-way rotation
  6. Hot-PCE breakdown branch (27%, the branch the session's direction most resembled)Partial · direction matched, trigger unconfirmed
    Called
    A hot Core PCE and/or Durable Goods beat produces a held, displaced close below 1.16353
    Actual
    Durable Goods beat as flagged, but PCE itself was inline; direction matched the branch's thesis (EUR down) but the trigger never confirmed
  7. Key level 1.16824 (resistance/pivot, soft-PCE confirmation trigger)Correct
    Called
    Expected to cap upside absent a soft-PCE surprise
    Actual
    Never tested -- session high 1.16771, 5.3 pips short
  8. Key level 1.16507 (Tuesday's low, freshest support)Correct · mechanism fired, timing shifted
    Called
    "A clean loss on volume ahead of the print would open a fresh leg toward 1.16353"
    Actual
    Lost cleanly, but in the delayed-resolution window (17:00 UTC) rather than ahead of the print -- and it did open the leg toward 1.16353 as described
  9. Key level 1.16353 (structural floor, hot-PCE confirmation trigger)Correct · narrow survival
    Called
    First real support; a held loss negates the range read
    Actual
    Held -- low 1.16416, 6.3 pips above, the narrowest margin the floor has survived by across the digestion band's life

The tape

  1. Asian / pre-London (00:00-07:00 UTC)

    Opened 1.16716, ticked up to the day's eventual high of 1.16771 within the first two hours, then reversed into a slow grind lower through the rest of the Asian session and into early Europe, closing the 06:00 hour at 1.16643 -- the day's high was set and abandoned before London even opened.

  2. London open (07:00-09:00 UTC)

    This pair's primary ignition window extended the overnight drift rather than reversing it, dipping to a pre-print low of 1.16594 (09:00 UTC) with no displacement or real follow-through -- consistent with the priors' caution that a push here, more than three hours ahead of the print, is positioning rather than signal.

  3. Pre-print recovery and rollover (09:00-12:00 UTC)

    Price recovered to 1.16766 by the 11:00 hour, approaching but not retesting the day's earlier high, then rolled over ahead of the 12:30 UTC print, closing the 12:00 hour at 1.16657.

  4. The print and immediate reaction (12:30-13:00 UTC)

    Core PCE m/m (0.2% vs 0.2% forecast), Core PCE y/y (3.3% vs 3.3% forecast), GDP q/q (1.5% vs 1.5% forecast), and GDP Sales q/q (2.2% vs 2.2% forecast) all landed dead on consensus. Durable Goods Orders beat clearly (1.1% vs 0.8% forecast, 0.3% prior). The immediate reaction was muted -- the 13:00 hour drifted from 1.16656 to a close of 1.16615, exactly the low-conviction first-30-minutes behavior the priors' sweep-fade discipline anticipates for a largely inline cluster.

  5. Delayed-resolution window (13:00-17:00 UTC)

    After a brief 14:00 bounce to 1.16685, the durable move began at 15:00 UTC with a break from 1.16686 to a low of 1.16551 (close 1.16556) -- the priors' NY-overlap fade zone (documented as a reversal zone, pullback bottoms continuing only 24-25%) instead became the epicenter of the session's sharpest extension. The break continued through 16:00 (close 1.16523) and into 17:00, where the session low printed at 1.16416 -- a clean loss of Tuesday's 1.16507 swing low, stalling 6.3 pips short of the 1.16353 structural floor.

  6. Late session (17:00-22:00 UTC)

    Price stabilized after the 17:00 low, recovering to 1.16501 by 18:00 and chopping in a tight 1.1649-1.1654 band through 22:00, with no fresh catalyst behind the recovery -- position-squaring ahead of the Jackson Hole eve, as the prep's late-session framing anticipated.

  7. Close (22:00-23:59 UTC)

    EURUSD closed the session at 1.16515, 20.1 pips below the open and comfortably inside 1.16416-1.16771 -- the lower half of the digestion band, with the floor tested but not broken.

What we learned

**Low surprise, but a real test.** The 12:30 UTC cluster was almost entirely inline -- only Durable Goods surprised -- yet the session still produced a coherent 30-pip directional break that pressed the digestion floor to within 6.3 pips of confirmation. This is not the calendar-quiet, catalyst-free grind of recent sessions; the print mattered, it just fell short of full structural confirmation.

  1. The hot/soft-PCE branch framing missed that the actual catalyst was the secondary print, not the lead indicator

    PCE was inline and gave no directional signal on its own; the entire dollar impulse came from the Durable Goods beat the driver stack flagged only as a "compounding" risk, not a scenario in its own right. Routes to: driver-stack read -- when a data cluster has a lead indicator (PCE) and a secondary indicator (Durable Goods) that can diverge, name a branch keyed to the secondary print's surprise explicitly, rather than folding all directional risk into the lead indicator's hot/soft split.

  2. The NY-overlap fade-zone prior (15:00-16:00 UTC, pullback bottoms continuing only 24-25%) was overridden by the delayed-resolution window on this event day, and the break accelerated through the zone rather than fading

    The two priors point in opposite directions on tier-1 event days, and this session the event-driven extension won cleanly. Routes to: driver-stack/day-type precondition -- propose a priors edit suspending the NY-overlap fade-zone read when it directly overlaps a live delayed-resolution window from a same-day tier-1 print.

  3. The 1.16353 invalidation trigger performed exactly as designed

    A 6.3-pip stall short of a held close kept the map from confirming a break that, given the late-session stabilization, would likely have round-tripped. This is the displacement requirement working, not a gap to fix -- no correction needed here.

  4. A binary triggered/not-triggered read on the Neutral/Wait lean undersold a near-miss directional session

    A 30-pip intraday swing that stalls 6.3 pips from confirmation is materially different from a session that never threatens structure at all, but both currently grade identically as "no trigger." Routes to: tradability standard -- tighten the lean-outcome grading to register near-miss magnitude (e.g., within roughly 10% of ATR of a trigger) as its own category, rather than collapsing every untriggered session into the same neutral-correct/neutral-incorrect binary. --- Reviewed prep: 2026-08-26-eurusd-session-preparation