Intraday snapshot · S&P 500 E-mini · Wed 16 Sep 2026
The 7699 extension is given back; inside the initial balance at the day's control
Ninety minutes after the third push cleared 7687, the 12.5-point extension is gone: 7679.75, inside the initial balance, on the session control. The day is now a two-node profile inside 14 Sep's 7665–7703 value, and neither morning trigger has fired.
Who holds territory
Sellers, since 12:00
The extension above the 7686.5 initial-balance high failed: 7699 held one bar, and every close from 12:00 to 13:00 was lower, the 13:00 low 7674. That is the afternoon's structure. The sellers have not taken the 7669 initial-balance low, so this is control of the rotation, not the day.
Who is offside
Both sides; the 11:45 buyers are nearest
Whoever bought the 11:45 extension between 7687 and 7699 is 7 to 19 points offside at 7679.75. Yesterday's sellers, control 7657.25 on 14,433 net sold, are still 22 points underwater. Both cohorts sit inside 25 points of last; whichever the auction reaches first pays.
Gap up, sold, bought back: an initial balance inside 14 Sep's value
This is the excursion out of the 20–26 Aug balance, begun 27 Aug, and today's 7669–7699 range sits entirely under the composite's 7720 value low, in the thin stretch above yesterday's 7643.5 low.
The open at 7676.5 gapped 20.5 over the 7656 close. The 9:30 bar sold to 7669, the 9:45 bar bought to 7686.5 on +2,301, and the initial balance settled at 7669–7686.5, 17.5 points, inside 14 Sep's 7665–7703 value with both morning triggers four points outside its edges.
29 Jul – 15 Sep · contains the 2 beside it, not today
5 sessions
13 sessions
developing
Bars are bucketed to approximate distribution shape only. Every labelled price — POC, VAH, VAL, and the levels at right — is the exact tick-level value, not a bucket midpoint.
Today's 7669–7699 range sits inside 14 Sep's 7665–7703 value, the morning triggers four points outside each edge. The 7699 high stopped under the 7699.75–7703 cluster; 7657.25, yesterday's control, is untouched below.
Extension to 7699 at 11:45, all of it returned by 12:30
The 10:15 bar retested 7669.5 and held it. The 10:30 to 11:00 bars bought 7670 to 7692.75, the 11:00 bar clearing yesterday's 7687 high on +721. The 11:15 and 11:30 bars held 7682–7690.5 on flat delta; the 11:45 bar extended 12.5 points above the initial balance to 7699 — the overnight high to the tick, under the 7699.75 July swing high and yesterday's 7701 overnight high — on +587, and closed back at 7689.
That was the high. The 12:00 bar closed 7685.75, 12:15 at 7683.25, 12:30 at 7680.75 — back inside the initial balance — and the 12:45 bar sold −1,007 in a 4.75-point range. The 13:00 bar printed 7674, the afternoon's low, and closed 7678.75. Extension above: 12.5, all given back. Below: none.
Three thousand contracts net bought for three points
Cumulative delta ran +540 at 9:30, +2,841 by 9:45, +3,515 when 7687 cleared at 11:00. The 12:00 bar closed the peak at +4,229 — one bar after the price high, on a close 13 points below it. The 12:45 bar's −1,007 was the session's largest sell and moved price 1.5 points close to close, absorbed at the 7680 session control.
Delta stands +3,250 at the 13:15 partial; price stands 3.25 points above the open. The buying has been met the whole way up, and the 11:45 buyers are the cohort holding it.
Nothing triggered; 14 Sep's value traded edge to edge
The alternate needed acceptance above 7703; the high was 7699, four points short, one bar. The primary needs 7665 lost; the low is 7669, held at 9:30 and 10:15. The bear's 7643.5 is 36 points away.
The 11:59 snapshot said 7687 cleared and held. It held one more bar: the 12:00 close was 7685.75 and price has not been over 7687 since. The 08:05 read's line that yesterday's sellers were the nearest trap has not paid — 7657.25 has not been touched.
Sellers own the rotation, nobody owns the day
Two nodes, 7670.5 and 7685.75, last at the 7680 control between them. The failed extension says the rotation goes to the other initial-balance edge.
What argued for the primary
- Extension above failed; a failed extension rotates to the other edge, 7669
- Delta rolled +4,229 to +3,250 on the day's largest sell bar
What argued against
- 7669.5 held at 10:15; the 13:00 low of 7674 left the lower node untested
- Delta is still +3,250
What decided it
A failed extension outweighs delta that bought three points.
The rest of the session
Initial-balance low lost, into yesterday's value high
PrimaryA close under 7669 after the failed extension puts the whole day's buying offside. 7665 is 14 Sep's value low and the morning primary's trigger; 7659 is yesterday's value high, where yesterday's sellers begin getting flat.
Upper node reclaimed, second run at 7699 and 7703
AlternateBack above 7690.5, the 11:15 high, the 11:45 buyers are whole again and the 7699–7703 cluster is the only thing between them and the morning alternate's 7720.
Through yesterday's value high, the gap fully rejected
BearBelow 7659 the day goes from balance to rejection of the whole gap. Yesterday's 7657.25 control is the first stop, 7649 its value low, 7643.5 the low that held five hours yesterday.
What to watch
- 7665.00 14 Sep's value lowFour points under the initial-balance low, held twice today. Lost, the morning primary is live and the 11:45 buyers are 22 points offside.
- 7687.00 Yesterday's highCleared at 11:00, lost at 12:00. Reclaimed, the afternoon's sellers are the offside cohort and 7699 is the test.
- 7699.00 Overnight high, today's highTwo touches now, night and 11:45, both sold. A third that holds is acceptance under the 7703 trigger rather than rejection.
What this is, and what it is not
This is a context gauge — the climate a discretionary decision gets made inside. It is not a trigger, a signal, or a trade.
Mechanical order-flow entry signals do not survive out-of-sample testing, and nothing here should be treated as one. The value of a read like this is subtractive: it tells you which levels matter, which side is already committed, and when to stand down — not when to click.
Every figure above is computed rather than estimated. The synthesis is machine-written from a fixed analytical framework and reviewed before publication — how these are made. Framework v1.5 (fb07b9e+)