EURUSDReviewCautious
EURUSD Session Review — August 17, 2026
A Real Breakout Above 1.1600 That Didn't Survive New York
EURUSD delivered the genuine, displaced breakout above 1.1600 the prep's continuation branch needed -- a clean H1 close at 1.16062 and a push to 1.16138 during the London ignition window -- but it failed to hold through the New York data window, fading on a stronger-than-expected Empire State print and a large TIC inflow beat to a session-low retest of 1.15706, one pip above the breakout-and-fail confirmation level, before closing at 1.15791, back inside the pre-existing range. Neither of the prep's two directional triggers durably fired; the Neutral/Wait lean and the caution around the thin, competitive three-way map both proved correct, though the leading 45%-weighted scenario's own target zone was never reached.
Every prep call is graded like this. See the full track record →
Powered by Cortiq
See how Cortiq worksGraded 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.
- Session
- EURUSD Range-Top Breakout-and-Reversion Session
- Symbol
- EURUSD
- Window
- 00:00 - 23:00 UTC
- Day type called
- Trap (breakout-and-fail)
- Day type actual
- Trap (breakout-and-fail)
- Lean outcome
- Correct (neutral-correct)
- Regime
- Displaced breakout above a thrice-rejected shelf, fully reverted into range
- Preparation
- Partially accurate
- Surprises
- Moderate
Grade card
7 of 11 correct- Day-type callCorrect
- Called
- Trap (breakout-and-fail)
- Actual
- A genuine, displaced break above 1.1600 that failed to hold through NY and fully reverted into the pre-existing range
- LeanCorrect · neutral-correct
- Called
- Neutral/Wait -- neither trigger cleared the 55% combined-weight bar
- Actual
- Session round-tripped a violent breakout spike and closed back inside 1.1550-1.1600, resolving nothing durably in either direction
- Conditional lean -- long triggerCorrect · no durable trigger
- Called
- Long on a held, displaced H1 close above 1.1600
- Actual
- 10:00 UTC H1 closed 1.16062 (past 1.1600 with 15p+ displacement, touching 1.16138), but the very next H1 bar closed back under 1.1600 -- never held
- Conditional lean -- short triggerCorrect · no durable trigger
- Called
- Short on a held, displaced H1 close back below 1.15697, surviving London open and NY window
- Actual
- Session low came within one pip of 1.15697 (1.15706 at 21:00 UTC) but never closed below it
- Lead scenario -- breakout-and-fail (45%)Incorrect
- Called
- Held H1 close below 1.15697, fading to 1.15550-1.15600 then 1.15300
- Actual
- Price never closed below 1.15697; the round-trip target zone was never reached
- Runner-up -- breakout continuation (30%)Incorrect
- Called
- Held H1 close above 1.1600, extending to 1.16217
- Actual
- Cleared 1.1600 with real displacement (to 1.16138) but reversed within the hour and closed the day well below both 1.1600 and 1.1585
- Scenario Map qualityPartial
- Called
- Third-ranked branch (range/chop, 25%) trailed a 15-point lead
- Actual
- The eventual close (1.15791, inside 1.1550-1.1600) best matches the trailing, lowest-weighted branch -- the map named the outcome but underweighted it
- Key level 1.1600 (breakout confirmation)Correct · rejected as expected
- Called
- Sweep target; a held close beyond confirms breakout
- Actual
- Touched and closed above intrasession (1.16062, high 1.16138), then rejected within the hour
- Key level 1.15850 (shelf, 4th test)Partial
- Called
- Fourth test of a level rejected three times prior; needs a held close above 1.1600 to call it cleared
- Actual
- Traded above it for roughly 12 hours (04:00-16:00 UTC) -- the most decisive clearance of the four tests -- but closed the day back below it at 1.15791
- Key level 1.15697 (Friday's close / fail trigger)Correct · held as support
- Called
- A held, displaced loss here confirms breakout-and-fail
- Actual
- Dipped to 1.15706 in the late session but held as support, never closing below
- Key level 1.15113 (deepest invalidation)Correct · untouched, structure intact
- Called
- A loss here would negate the entire bullish leg
- Actual
- Not tested; session low was 1.15576, well above
The tape
- Open (00:00-01:00 UTC)
Opened 1.15607, a ~9-pip gap down from Friday's 1.15697 close, exactly as the prep flagged. Swept immediately to 1.15576 in the first hour -- the Asian-session low the prep had already noted as "in progress" at file time.
- Mid-session (02:00-11:00 UTC)
Price reversed hard off the sweep low and climbed steadily through the Asian close and into London: 1.15744 (02:00) -> 1.15798 (03:00) -> 1.15831 (04:00), clearing the 1.1585 shelf around the 04:00-06:00 window, then accelerating through the London open into 1.15952 (09:00) and exploding through the 1.1600 handle at 10:00 UTC -- an H1 close of 1.16062 with a high of 1.16115, extending to a fresh multi-week high of 1.16138 at 11:00 UTC. This was the genuine, displaced breakout the continuation branch (30%) needed: a full H1 close well past 1.1600 with 38 pips of clearance, arriving in the pair's primary London ignition window.
- Late / close (21:00-23:00 UTC)
Price bounced off 1.15706 through the 22:00-23:00 dead zone -- the window the prep itself flagged as uninformative -- closing the day at 1.15791, 18 pips above the open and comfortably back inside the pre-existing 1.1550-1.1600 range, just under the 1.1585 shelf, having fully round-tripped the session's 1.16138 high without a held close confirming either side of the map.
Full notes
What We Learned
Surprises: Moderate. The breakout was not a thin, dismissible wick -- it cleared 1.1600 with a full H1 close and 38 pips of displacement, comfortably past the standard's 15-pip/94% base-rate confirmation threshold, in the primary London ignition window. That it still failed to hold through New York is the session's real surprise: a fully confirmed breakout by the prep's own displacement rule that still round-tripped. The late-session dip to 1.15706 -- one pip from confirming the breakout-and-fail trigger -- was a genuine coin-flip near-miss, not a clean rejection of the lead scenario.
- Tighten the tradability standard. A single H1 close beyond a level with 15p+ displacement was necessary but not sufficient at a thrice-rejected shelf -- the close needs to survive at least one additional session window (here, the NY data window) before being treated as a live breakout. No-Trade Condition #3 ("a genuinely competitive three-way map is itself a caution signal... require full displacement confirmation") already argued for this; today is a confirming instance, not a contradiction, and it should stay codified as a standing check at heavily-defended levels.
- Re-examine the driver-stack precondition for stacked moderate-tier prints. The session was flagged as carrying no tier-1 EUR or USD event, and correctly so -- but Empire State, NAHB, two rounds of Treasury auctions, and TIC all landed on the same calendar day, and in aggregate they supplied the USD bid that stalled the breakout and drove the late fade. Propose a priors edit: three or more same-day moderate-tier USD releases can function as a de facto driver even absent a single headline mover, and the driver stack should flag that clustering rather than reading each print in isolation as sub-threshold.
- Loosen the binary trigger for near-miss confirmations. The breakout-and-fail branch's fail trigger (a held, displaced close below 1.15697) missed by a single pip on one H1 low (1.15706), which is enough to flip the branch's grade from a near-hit to a clean miss. Consider scoring a clean intrasession test-and-hold of the level as partial credit toward the branch, rather than an all-or-nothing close-below requirement, so the map's grading better reflects how close the call actually came.
- No action needed -- priors held. "The reaction at the level is the trade, not the level" and the lead-margin-is-signal amendment both worked as intended: the 45/30 (15-point) lead margin correctly flagged an operative-but-contestable call rather than a confident one, and it landed on the side of "no clean resolution," which is exactly the coin-flip-adjacent outcome a 15-point margin should be read as.
Session-analysis grading check: pass -- day-type call and lean graded per Rules 1-4 and 3b; the systematic miss (confirmed-but-unheld breakout) is routed to the tradability-standard category above per section 6 of the session-analysis framework.
Reviewed prep: 2026-08-17-eurusd-session-preparation