SP500ReviewCautious

SP500 Session Review — August 28, 2026

A Jackson Hole Whipsaw Blows Through Both Sides of the Range

SP500 did not have the quiet pre-weekend digestion day the preparation called. The session round-tripped more than 70 points — 131% of the 14-day ATR — ripping through the 7,735.22-7,746.43 resistance band to a fresh high of 7,770.27 in the 17:00-18:00 UTC window before fully reversing to a 7,699.77 low, then settling to close at 7,707.97, just 13 points below the open. No scheduled tier-1 US print explains the move; it lines up with the ongoing Jackson Hole Economic Symposium. The Range day-type call was wrong; the map's own 23%-weighted whipsaw branch was the closer read, and the next preparation should not let 'no tier-1 print today' stand alone as the case for a Range day when a live Symposium sits inside the window.

Prep outcomepartial
Lead scenario44% · missed
Leanneutral · correct
Day typerange → whipsaw
Surprisehigh
Grade card5 of 10 correct
Session chart
SP500 — 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
SP500 Friday Pre-Weekend Whipsaw During the Jackson Hole Symposium
Symbol
SP500
Window
00:00 – 23:59 UTC (14:30 UTC US cash open as the dominant engine)
Day type called
Range
Day type actual
Whipsaw
Lean outcome
Partial
Regime
Violent two-sided whipsaw, closing near-flat versus the open
Preparation
Partially accurate
Surprises
High

Grade card

5 of 10 correct
  1. Day-type callIncorrect
    Called
    Range — no tier-1 print, pre-weekend digestion
    Actual
    Whipsaw — both edges broke intraday, neither held, 131% of ATR range
  2. LeanCorrect · neutral, no durable resolution
    Called
    Neutral/Wait
    Actual
    Session round-tripped over 70 points but closed only 13 points from the open — no durable resolution
  3. Conditional lean — long trigger (H1 close above 7,735.22, survive cash open + power hour)Incorrect · fired, did not hold
    Called
    Confirms continuation toward 7,746.43 / 7,775.60-7,805.39
    Actual
    Fired twice (15:00 close 7,737.52; 17:00 close 7,767.02) with real displacement past the level, but reversed within the same or next hour and never survived the power hour
  4. Conditional lean — short trigger (H1 close below 7,714.35, survive cash open + power hour)Correct
    Called
    Confirms range/rejection, reopens 7,695.40/7,696.21
    Actual
    Fired at 19:00 (close 7,713.77) and held through the power hour and the close (7,707.97)
  5. Lead scenario — Range/rejection (44%)Incorrect · wrong mechanism, coincidental close
    Called
    Confirmed close below 7,714.35 without a sustained close above 7,728.30-7,735.22 first; retest of 7,695.40/7,696.21
    Actual
    Price closed below 7,714.35 by the end, but only after a sustained, displaced push to 7,770.27 — the scenario's own defined path was violated even though the closing level rhymes with it
  6. Key level 7,746.43 (resistance)Incorrect · broken, then failed to hold either way
    Called
    Next ceiling if the band finally breaks
    Actual
    Blown through by more than 20 points to 7,770.27 before fully reversing
  7. Key level 7,735.22 (immediate resistance)Incorrect
    Called
    Needs cash-open confirmation, not just an overnight print, to be trusted
    Actual
    Got two separate confirmed-and-displaced closes above it, and still didn't hold
  8. Key level 7,714.35 (support/pivot)Correct · eventual break matches the close
    Called
    A hold above keeps range/rejection honest; a break reopens the downside case
    Actual
    Broken decisively from 19:00 UTC onward, day closed below it
  9. Key level 7,695.40 / 7,696.21 (support)Correct · held, within noise
    Called
    First real support test below the open
    Actual
    Wicked to 7,699.77 (4.37 points above) — never confirmed through
  10. Key level 7,680.12 (flipped pivot)Correct
    Called
    This week's central level; a loss here signals the reclaim failing
    Actual
    Never seriously tested (day low 7,699.77 stayed well clear)

The tape

  1. Open (00:00-14:30 UTC overnight book into the cash open)

    Consistent with the prep's own read — the overnight book stayed contained, testing the 7,735.22 pivot twice (05:00 and 06:00 UTC highs both at 7,735.22) without confirming through it, then drifted back toward the 7,719-7,731 area heading into the 14:00 UTC Michigan print, which landed in line (51.7 vs 51.0 forecast) and produced no reaction.

  2. Mid-session (14:30-19:00 UTC)

    The cash open produced an immediate, confirmed break: the 15:00 UTC H1 candle closed at 7,737.52, above the 7,735.22 pivot on its first post-open hour. The 16:00 candle stalled at 7,745.52 before fading back to 7,734.08, right at the pivot — an early tell the break wasn't clean. Then the 17:00 UTC hour delivered the session's defining move: a sweep down to 7,714.52 followed by a violent repricing higher to a fresh high of 7,767.77, closing the hour at 7,767.02 — a single-hour range of more than 53 points, larger than the full day's ATR. No scheduled tier-1 event sits in this window; the timing lines up with the ongoing Jackson Hole Economic Symposium, whose exact speaker schedule the calendar feed does not itemize.

  3. Late / close (19:00-23:59 UTC, power hour into the Friday close)

    The entire spike reversed within the next hour — the 18:00 UTC candle posted the session high (7,770.27) early and closed at 7,727.27, giving back nearly the whole move. From 19:00 UTC onward price broke and held below 7,714.35, extending to the session low of 7,699.77 at 20:00 UTC (4.37 points above the 7,695.40 support), then chopped in a tight 7,700-7,716 band into the close, finishing at 7,707.97 — 13 points below the open and inside the week's established balance.

What we learned

**Surprises:** The 17:00-18:00 UTC round trip was the session's defining surprise — a move of more than 53 points in a single hour, unexplained by any scheduled US print, most plausibly tied to Jackson Hole Symposium commentary landing without a listed exact time. The Michigan Consumer Sentiment print at 14:00 UTC was in line and demonstrably not the catalyst. The magnitude (131% of the 14-day ATR, on a day the prep expected 60-80% of ATR at most) exceeded even the map's own whipsaw branch, which anticipated both edges testing but not a displaced break of this size on either side. The long conditional trigger firing twice with real displacement and still failing both times is new and worth flagging.

  1. Chronically wrong day-type calls → re-examine precondition checks

    The prep's own carried-forward footer already showed 38% day-type accuracy over 16 sessions; this session adds another miss. The precondition check that treats "no scheduled tier-1 print" as sufficient grounds for a Range call needs to explicitly exclude sessions where a multi-day Symposium or speaker event is live inside the window, even without an exact listed time — that combination has now produced a materially underweighted whipsaw.

  2. Right scenario, weight too low → raise the whipsaw floor when the driver stack itself flags contradiction plus live Symposium/speaker risk

    The map correctly named whipsaw at 23% and its own base-rate reasoning cited "the contradictory driver stack" — but defaulted the day type to Range anyway. When both signals are present together (explicit driver disagreement AND an unscheduled-but-live macro event inside the window), whipsaw should compete for the day-type call itself, not just sit as a third scenario.

  3. Right scenario, untradable trigger → tighten the tradability standard for confirmed-and-displaced breaks during live event risk

    A single confirmed H1 close with real displacement past a level failed twice on the same day. Add a second-consecutive-close requirement specifically when a Symposium/speaker event is live in the window, rather than treating one displaced close as sufficient confirmation.

  4. The no-trade discipline worked as designed

    The prep's own "survives the cash open and the power hour" requirement is exactly what kept a disciplined trader out of the long fakeout — the break never held that long. This is worth preserving even as the day-type and weighting issues above get fixed. --- Reviewed prep: 2026-08-28-sp500-session-preparation