EURUSDReviewCautious

EURUSD Session Review — August 12, 2026

CPI Fakes Above 1.1550, Resolves Through 1.15300

EURUSD opened 1.15388 and closed 1.15217, a 17-pip down day built from a 43-pip round trip around the 12:30 UTC US CPI print. A mixed release -- headline m/m missed (0.1% vs 0.2% forecast) while y/y beat sharply (3.4% vs 2.7% forecast) -- first sparked a genuine rally through 1.1550 into the 1.15550-1.15600 soft-CPI zone, then fully reversed and held a displaced close below 1.15300 from 19:00 UTC into the close, confirming the prep's 32%-weighted hot-CPI branch over its 40%-weighted lead. The prep's Neutral/Wait lean resolved directionally; its named levels and both directional triggers worked exactly as built.

Prep outcomepartial

Lead scenario40% · missed

Leanneutral · incorrect

Day typeevent-suspended → whipsaw

Surprisemoderate

Grade card6 of 11 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.

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 US CPI Resolution Review
Symbol: EURUSD
Window: 00:00 - 23:00 UTC (full session, keyed to the 12:30 UTC US CPI print)
Day type called: Event-suspended
Day type actual: Whipsaw
Lean outcome: Incorrect (Neutral/Wait; session resolved directionally)
Regime: Two-sided post-CPI whipsaw, resolving lower into the close
Preparation: Partially accurate
Surprises: Moderate

Grade card

6 of 11 correct

Day-type callCorrect
Called
Event-suspended (tier-1 CPI at 12:30 UTC)
Actual
Compressed 8-pip pre-print drift, exactly as called, then a genuine two-sided post-print resolution
LeanIncorrect
Called
Neutral/Wait, no operative lean (55% bar uncleared)
Actual
Session closed 17 pips lower than the open, a directional resolution
Conditional lean (short side)Correct
Called
Short on a held, displaced H1 close below 1.15300 clearing Tuesday's 1.15308 low
Actual
Trigger fired at 19:00 UTC (close 1.15277); price extended to a 1.15194 low and held below 1.15300 for five straight H1 closes into the 1.15217 close
Conditional lean (long side)Correct · trigger failed to hold, as the fade-window methodology anticipated
Called
Long on a held, displaced H1 close above 1.1550 clearing 1.15550-1.15600
Actual
Two H1 closes held above 1.1550 (1.15590, then 1.15518) but the reclaim reversed by the 17:00 close (1.15425) before the map's displacement condition was durably satisfied
Lead scenario -- Inline CPI / range-chop (40%)Incorrect
Called
Price stays inside 1.15300-1.1550 without a decisive resolution
Actual
Price broke and held outside both edges of the band
Hot-CPI branch (32%, the branch that fired)Correct · fired, target approached not cleared
Called
Held, displaced close below 1.15300, extends toward 1.15167 then 1.1500
Actual
Held from 19:00 UTC, low 1.15194 (2.7 pips shy of 1.15167), closed 1.15217
Soft-CPI branch (28%)Incorrect
Called
Held, displaced close above 1.1550, extends toward 1.15659 then 1.15805
Actual
Reclaimed 1.1550 for two hours (high 1.15624) but never cleared 1.15659 and fully reversed
Key level 1.1550 (pivot)Partial
Called
A held, displaced reclaim flips the read bullish
Actual
Reclaimed for two H1 closes, then reversed -- a real reaction, not a durable flip
Key level 1.15308/1.15300 (support, hot-CPI trigger)Correct
Called
A held, displaced loss confirms the hot-CPI branch
Actual
Lost and held from 19:00 UTC onward
Key level 1.15167 (hot-CPI first target)Partial
Called
First real target on a confirmed hot-CPI break
Actual
Session low 1.15194, 2.7 pips short of the level, before the close
Key level 1.15659 / 1.15805 (soft-CPI confirmation levels)Correct
Called
Clearing these validates the soft-CPI branch
Actual
Day's high 1.15624 never reached 1.15659

The tape

Open (00:00-07:00 UTC, Asian)
Opened 1.15388 and stayed compressed exactly as the prep's event-suspended read called for -- an 8-pip range through most of the window (1.15345-1.15443), drifting down to ~1.15354 by 07:00. No directional tell.
Mid-session (07:00-12:30 UTC, pre-print)
London and the EU calendar window added no resolution -- price chopped 1.15318-1.15397, still inside the band. The 12:00-13:00 UTC candle carried the print itself: an immediate spike to 1.15495 (just short of 1.1550) that fully faded back to a 1.15357 close by 13:00, then continued fading to a 1.15321 low in the following hour -- a textbook sweep-fade of the first post-print tick, exactly as the prep's no-trade window anticipated.
Second decision window (14:00-16:00 UTC)
A genuine second wave developed here, not in the immediate aftermath -- price pushed from 1.15344 through 1.1550 and into the soft-CPI trigger zone, printing a 1.15624 high and closing the 15:00 hour at 1.15590, comfortably inside 1.15550-1.15600. The 16:00 close (1.15518) held above 1.1550 as well -- two consecutive H1 closes above the pivot, a real reclaim, not a wick.
Late session (17:00-23:00 UTC)
The reclaim did not survive. The 17:00 close (1.15425) dropped back below 1.1550, and by 19:00 UTC -- roughly 6.5 hours after the print, well past even the extended second-decision window the priors describe -- price broke and held below 1.15300 (close 1.15277), then extended to a 1.15194 low by 21:00. Five consecutive H1 closes held under 1.15300 into the 23:00 close of 1.15217, near the session low and 17 pips below the open.
Closing posture
Closed 1.15217, below the open, below the 1.15300 hot-CPI trigger, and within 2.7 pips of the branch's 1.15167 target -- a held, displaced resolution to the downside that only became visible in the last third of the session.

What we learned

The session unfolded with two distinct surprises, neither of which broke the map's structure but both of which argue for calibration edits.

[Driver-stack read -- propose a priors edit] CPI printed with genuine internal divergence, and the market ultimately traded the divergent measure, not the consensus one
Headline m/m missed forecast (0.1% vs 0.2%) while y/y beat sharply (3.4% vs 2.7%). The first market reaction (the 14:00-16:00 rally through 1.1550) traded the softer m/m miss; the durable, held move traded the y/y beat instead. When a forecast set already implies a large gap between the m/m and y/y readings, as today's did (m/m expected to decelerate sharply from -0.4% to 0.2%, y/y expected to decelerate from 3.5% to 2.7%), the scenario map should flag a fourth, explicit "mixed print" contingency rather than only inline/clean-hot/clean-soft branches.
[Driver-stack read -- propose a priors edit] The durable post-CPI move took materially longer than the priors' 1-4h delayed-resolution window
The held break below 1.15300 didn't establish until 19:00 UTC, about 6.5 hours post-print -- beyond even the extended "second decision window" (13:00-16:00 UTC) the live-review amendments describe. This session's own second-wave rally (14:00-16:00) looked like it was completing the delayed-resolution pattern, then itself reversed. Worth widening EURUSD's post-CPI resolution window on mixed-signal prints specifically, rather than treating 13:00-16:00 as the outer bound.
[Tradability standard -- confirmed working, no fix needed] The hot-CPI branch's trigger definition performed exactly as designed
A held, displaced H1 close below 1.15300 that survived the sweep-fade window did fire, and price followed through five straight hours to within 2.7 pips of its stated target. The displacement-plus-hold-through-fade-window standard is validated by this session, including the false trigger at 1.1550 that the same standard correctly filtered out (two closes above the pivot were real, but never met the sustained-hold bar the branch required).
[Lean discipline -- no fix needed, note for calibration ledger] The Neutral/Wait outright lean was the honest call given an 8-point lead margin under the 55% bar, but the session still resolved with real distance (17 pips, 30% of daily ATR)
This is consistent with the ledger's standing finding that roughly half of Neutral/Wait sessions resolve directionally -- today's conditional-lean framing correctly captured that resolution even though the outright lean did not. --- Reviewed prep: 2026-08-12-eurusd-session-preparation