A couple of guys talking daily market data for ES/NQ futures traders. Not financial advice.
Options based market read at 9:50ET. Grade after market close.
Send me any trading claim — a thread, a backtest, a rule from a course. I'll run it through the same placebo battery I use on my own ideas and publish the result, win or lose. Most claims die. The ones that survive, you'll want to know about.
It's incredibly helpful. You can CRANK through hypotheses to test and get things built much faster than without (especially personal tools), but if you're not pretty familiar with statistics and whatnot be very careful.
Working with AI on tools is hilarious...
"And I should be plain about my own record here. In this session I produced four confident findings that dissolved under controls, and separately re-killed a real result by repeating a predecessor's error. Today's catches came from you twice, from a pre-registration once, and from Fable once. None came from me checking my own work unprompted. That's worth knowing about how much weight to put on my first pass at anything."
@TailThatWagsDog Love your framing. It blows my mind how helpful AI has been, especially the last 6 months with model improvement.
My biggest lesson has been to make sure I'm not relying on it to do the read for me. Building deterministic tools with it that I can verify and debug is critical.
Today's hedge flow is charm.
as of 10am MMs have ~$2.7bn of delta per day to SELL with the index at 7720.. and 92% of it comes from the position expiring TODAY.
here's where it comes from, how to see it in VS3D, and how to go from dollars per day to contracts-
(short thread)
Band today: 27 to 57 points, rest of day from a late 10:47 anchor. Realized 19.4.
Miss, low side. Record 6 from 9.
That is the 1st percentile of my sample, the quietest rest-of-day I have on record. The whole range was made in six minutes after the anchor and nothing after 13:37 got outside it.
Second low-side miss in a row. Both days the tape delivered less than even the compressed band expected.
Late read, so this is anchored 10:46 and graded from 10:47.
Spot 7794. Absorbers heavy, p95, the heaviest share I have logged. Ceiling about 6 points overhead, buyer shelf ~20 below.
Band: 27 to 57 points of range into the close, scaled for the session that is left.
Small budget, and no view on which way it gets used.
Grade at 4:15.
Late read, so this is anchored 10:46 and graded from 10:47.
Spot 7794. Absorbers heavy, p95, the heaviest share I have logged. Ceiling about 6 points overhead, buyer shelf ~20 below.
Band: 27 to 57 points of range into the close, scaled for the session that is left.
Small budget, and no view on which way it gets used.
Grade at 4:15.
Band today: 28 to 61 points, rest of day. Realized 39.8.
Hit, mid-band. Record 6 from 8.
This one was a test of the map against a free read of the tape. The map called a small budget off heavy density. The opening range, 30 points in the first twenty minutes, called it a busy day.
The range came in at the 25th percentile of my sample. Map wins this one.
The day did not feel quiet. It tagged the upper band at 10:32, slid 40 points into noon, then closed 36 off the low.
Terrain read, 9:51. Spot 7792, already up 30 off the open, which is over half the day's expected move.
Absorbers heavy, p95. Ceiling about 8 points overhead, buyer shelf ~50 below.
Band: 28 to 61 points rest-of-day range. Small budget, and no view on which way it gets used.
Grade at 4:15.
Terrain read, 9:51. Spot 7792, already up 30 off the open, which is over half the day's expected move.
Absorbers heavy, p95. Ceiling about 8 points overhead, buyer shelf ~50 below.
Band: 28 to 61 points rest-of-day range. Small budget, and no view on which way it gets used.
Grade at 4:15.
@jam_croissant The 30 year is holding a level it last touched in July 2007. Meanwhile 30 day SPX implied vol is running about a point and a half below trailing realized.
Question for you: Can the front end price a slow structural repricing at all, or does it only ever price realized path and jump risk?
Bah. We knew it'd be a compressed day, but that's prettttty compressed. 3rd percentile of days in my sample.
A very quiet day broke it from below. My two misses now point opposite ways. Record: 5/7
No call on character today. Absorbers mid range at p45, and the ceiling is a slab rather than a level. Two reasons to withhold the word, so I am withholding it.
Band 32 to 84 pts rest of day. Still graded.
Sellers ~26 overhead, shelf ~49 below.
Live: band 5/6. Grade at 4:15
No call on character today. Absorbers mid range at p45, and the ceiling is a slab rather than a level. Two reasons to withhold the word, so I am withholding it.
Band 32 to 84 pts rest of day. Still graded.
Sellers ~26 overhead, shelf ~49 below.
Live: band 5/6. Grade at 4:15
Band held. 46.72 pts from the 09:56 anchor against 28 to 61.
But the word was half right. Range compressed, path trended: high at 10:01, bleed to a 14:29 low, closed 39 under the open.
Didn’t really move enough to test the moves stall call.
Live: band 5/6.
Absorbers heavy again, ~80th percentile for this hour, second day running. Sellers ~28 pts overhead, shelf ~27 below, tight box. Boundary sign positive (context).
Band 28 to 61 pts rest of day.
So what: moves stall. Take points when you have them.
Live: band 4/5. Grade at 4:15
The market calm is real but rented.
The quiet is measured, not imagined: heavy absorbers, overpriced movement, a book that dampens. It can run for months, and the market has priced it running until roughly mid October.
the mid-September recompute of the election box when the November weeklies list. Plus checkpoint dates: Labor Day, mid-September, November, year-end. If it's wrong, the record will show exactly when and which claim broke.
The market calm is real but rented.
The quiet is measured, not imagined: heavy absorbers, overpriced movement, a book that dampens. It can run for months, and the market has priced it running until roughly mid October.
Notes for grading this down the road: There are five tripwires registered above, and each we can falsify: skew crossing its one-year median, yen strength and 30-year yields rising together, BoJ hike repricing, our own daily bands missing high twice in a week, and