Mountain Frog Research

Sessions most like 2026-09-04, as of 11:30 ET

A quiet, top-heavy morning: the high went in early and the tape has spent the rest of the session leaking away from it, under VWAP.

What would separate them: The separator in these sessions was what the break of the morning low did: whether it drew expanding volume and acceptance underneath, or spiked and got reclaimed inside a few bars. Watch also whether price can take VWAP back and hold there, and whether the early high stays untouched as the afternoon develops.

What these sessions share is shape more than outcome: an early high, a late morning spent beneath it, and a downside probe once the cutoff had passed. Most of them traded through the morning low at some stage afterward; fresh session highs were the exception. Today arrives at that shape with less fuel behind it — participation is lighter than a normal session and lighter than most of the cohort, and the daily range is narrower than the corpus median — while the pullback so far has not gone as deep as typical, so more of the morning's structure is still standing here than in most of the analogs. Where the group separates is entirely after the cutoff, and it separates widely.

Break and press 4 of 12

These kept going the way the morning pointed: the break of the morning low drew follow-through rather than a bounce, and they closed in the lower part of the day's range. The most extreme of them became a sustained trend lower, far beyond the rest of the cohort in scale.

new high 0 new low 4 median to close -0.32 ATR
Reversed and ran 2 of 12

These gave back little or nothing below the cutoff before turning hard, spending the afternoon working up and away from where the cutoff found them and closing near the top of the day. Upside travel here was the strongest in the cohort — the morning weakness ended up marking the session low.

new high 1 new low 2 median to close +0.60 ATR
Ground back up 2 of 12

The same turn, slower: ground was lost first, then recovered through the afternoon as a grind rather than an impulse, closing above the cutoff price. The repair was steady and modest, with no clean trend leg in either direction.

new high 0 new low 2 median to close +0.23 ATR
Flushed then flat 4 of 12

These delivered the break and then bought it back, ending the day about where the cutoff left them. The downside excursion was real — in the deepest case a sharp flush — but it paid nothing by the close, and the afternoon was chop around the morning's middle.

new high 0 new low 4 median to close -0.03 ATR

Written from the computed figures only, by a model that is not permitted to write a number — every count above is rendered from the same data as the table. Camps are one reading of these sessions, not a forecast: the matched features have been measured to add nothing over price position alone, so a cohort rate at this sample size is descriptive.

2026-09-04 — TODAY vwap -0.36· hi +0.20 @09:36 · IBgap -0.24· vwapΔ -0.33 · retr 13%/32%max in progress
All twelve, and today, on one set of axes hover a line to isolate it · use the plot column in the table below to drop a session — the scale refits to what is left, so hiding an outlier gives the rest the whole frame
Faint left of the cutoff: those sessions were selected for resembling today there, so the agreement is by construction. Everything that matters is to the right. Outcome is in the geometry as well as the colour — a new high is a line clearing its own earlier high — because red and green are the hardest pair to tell apart with colour-blindness.
What these became
11 of 12 expanded downward
after the cutoff, by which extremes they took
Room past the level
-0.48 ATR
median run from the cutoff price
Heat before it got there
+0.12 ATR
worst excursion against you between the cutoff and the extreme — not the whole session · worst +0.35 ATR · typically reached 14:05
What would catch it off guard
1 did something else
1 went both ways · worst close +0.60 ATR

What kind of session these became — expansion one way, both ways, or balance — on the same four-way split the corpus study uses, where balance is a third of all sessions. Deliberately not a hit rate: which extreme gets taken largely restates where price already sat at the cutoff, and nobody holds to the close anyway. How far it ran, what it cost to sit through, and when it got there are the parts a trade is built on. n=12.

plot sessiondistATRIB /ATRRVOL retr nowretr max to closemax upmax down new highnew low
2026-09-04 — today0.0043.20.500.8713%32%in progress
2025-10-070.4746.50.391.0212%42%-0.03+0.05-0.35YES
2023-09-260.4934.00.851.142%23%-0.28+0.35-0.53YES
2023-05-310.5538.00.741.46-0%23%+0.44+0.76-0.01YES
2023-09-190.5631.80.701.0018%19%+0.60+0.78-0.31YESYES
2023-05-120.5636.00.520.895%24%+0.23+0.24-0.50YES
2024-07-300.5750.20.511.3618%28%-0.04+0.32-0.73YES
2023-02-210.5748.50.551.274%20%-0.32+0.07-0.41YES
2024-12-130.5835.50.610.3420%31%+0.22+0.32-0.23YES
2026-06-050.5955.50.851.2712%26%-1.90+0.06-2.27YES
2023-09-150.6534.00.571.3526%26%-0.51+0.12-0.62YES
2024-08-260.6659.80.410.8216%39%+0.03+0.18-0.27YES
2023-10-130.6748.50.481.4715%16%-0.12+0.26-0.48YES
cohort low31.80.390.34-0%16%-1.90+0.05-2.27
cohort median46.50.571.2715%26%-0.03+0.26-0.41
cohort high59.80.851.4726%42%+0.60+0.78-0.01
cohort hit rate1/1212/12
CORPUS median (n=967)50.50.511.0034%60%+0.05+0.30-0.28
CORPUS base rate56%50%

All ATR-normalised and measured from the 11:30 cutoff. max up / max down are forward excursions after the cutoff — not the session extremes, which in most of these were already set. new high / new low ask whether the post-cutoff move exceeded the pre-cutoff extreme. Click any row to expand that session's chart to full width and jump to it; click again to collapse.

2025-10-07 d=0.47 vwap -0.38· hi +0.10 @09:34 · IBgap -0.28· vwapΔ -0.36 · retr 12%/42%max closed -0.03 ATR
2023-09-26 d=0.49 vwap -0.43· hi +0.23 @09:44 · IBgap -0.23· vwapΔ -0.30 · retr 2%/23%max closed -0.28 ATR
2023-05-31 d=0.55 vwap -0.32· hi +0.26 @09:45 · IBgap -0.23· vwapΔ -0.26 · retr -0%/23%max closed +0.44 ATR
2023-09-19 d=0.56 vwap -0.25· hi +0.18 @09:45 · IBgap -0.21· vwapΔ -0.32 · retr 18%/19%max closed +0.60 ATR
2023-05-12 d=0.56 vwap -0.37· hi +0.07 @09:43 · IBgap -0.20· vwapΔ -0.35 · retr 5%/24%max closed +0.23 ATR
2024-07-30 d=0.57 vwap -0.48· hi +0.22 @09:42 · IBgap -0.20· vwapΔ -0.39 · retr 18%/28%max closed -0.04 ATR
2023-02-21 d=0.57 vwap -0.32· hi +0.23 @09:45 · IBgap -0.28· vwapΔ -0.25 · retr 4%/20%max closed -0.32 ATR
2024-12-13 d=0.58 vwap -0.37· hi +0.23 @09:47 · IBgap -0.34· vwapΔ -0.34 · retr 20%/31%max closed +0.22 ATR
2026-06-05 d=0.59 vwap -0.31· hi +0.09 @09:33 · IBgap -0.24· vwapΔ -0.44 · retr 12%/26%max closed -1.90 ATR
2023-09-15 d=0.65 vwap -0.30· hi +0.10 @09:30 · IBgap -0.15· vwapΔ -0.33 · retr 26%/26%max closed -0.51 ATR
2024-08-26 d=0.66 vwap -0.23· hi +0.13 @09:33 · IBgap -0.20· vwapΔ -0.25 · retr 16%/39%max closed +0.03 ATR
2023-10-13 d=0.67 vwap -0.45· hi +0.26 @09:41 · IBgap -0.17· vwapΔ -0.27 · retr 15%/16%max closed -0.12 ATR

How these were matched

Matched on seven measures, all ATR-normalised and z-scored: distance from VWAP, magnitude of the high off the open, the clock time of that high, expansion of the price-VWAP gap across IB creation, how far VWAP itself was dragged, and how much of the drive has been retraced — both currently and at its deepest. The IB-gap and VWAP terms separate a day that trended from one that chopped to the same place; the retracement terms separate a day still pinned near its high from one that has round-tripped to the open. Ordered nearest first. Solid = up to 11:30, the part matched on. Faded = the rest of the day, the part to read. Shaded band left of the 09:30 marker is premarket from 06:00 ET. The y-axis is a shared ATR ruler with the session open at 0, identical on every tile, so magnitudes are directly comparable.

n=12. This is a case-inspection tool for rare behaviour — sort the tiles into camps by eye. It is not an inference tool, and a percentage taken off this many charts would not mean anything.

Charts rendered with Lightweight Charts™, Copyright © 2026 TradingView, Inc. (Apache-2.0).