2026-09-29 · midday

Net cash / NLV still red at -0.111, $16,579 past the -0.10 line; no fills today. Red means Rule 3: tier 1 sales. Gross leverage 1.747x red; long-leg band still over. GOOGL $341.62, nearest written level tranche 3 at $360, +5.38%. MU $1,064.92 prints tomorrow after the close. Today's forecasting pipeline run refused at pre-flight again.

What to do

net cash / NLV -0.1112 (cash -$164,661.14 on NLV $1,480,822.46), $16,579 past the -0.10 red line.
SELL 200 AMZN, SELL 200 EQT, SELL 1 TSLA, SELL 1 DIS (limit near last, day)
Tier 1 in full at midday prints. Which tier-1 names is Gabriel's pick; the smallest tier-1 set that clears red is EQT 200 + TSLA 1 + DIS 1 + AMZN 26 ($16,639, lands -0.09996). AMZN 25 is not enough. · raises ~$59,622 (AMZN $49,406, EQT $9,758, TSLA $353, DIS $105); lands net cash / NLV at -0.071
Your playbook: “"Red means: execute Rule 3 until the metric is back to amber. Choose the sales myself, at my own pace, before the broker chooses them for me." Rule 3 tier 1: AMZN, BRK.B, EQT, DIS, TSLA, UPRO.”
EQT sells at a ~$1,045 loss (avg $54.01); last EQT buy 2026-08-12, outside the 30-day window. 100 of the AMZN shares were bought 2026-09-25 at $248.76: pick the older lots or any loss on them washes on an AMZN buy within 30 days.

Still over, unchanged

These fire every run until you act on them. No order attached.

rail	now	limit
long_leg_band	280P x5, 300P x8	$230-$260 band	13 contracts outside
gross_leverage	1.7472x	1.5x red (1.3x amber)	0.2472x over red
top2_name_delta_ratio	0.63x	amber 0.70x, red 0.85x	clear
total_delta_ratio	0.96x	amber 1.30x, red 1.50x	clear

Why each one reads the way it does

long_leg_band
Standing repair (BUY TO CLOSE 5 Dec-28 360P / SELL TO CLOSE 5 Dec-28 280P) is $18,180.91 debit at midday marks (36.362/spread). A cash debit while net cash is red, so the two rails conflict and it is not proposed.
gross_leverage
Gross $2,587,267 on NLV $1,480,822.

delta by name

Delta notional per name as of 2026-09-29: stock at delta one, each option leg at the Black-Scholes delta implied by its own mark. Top two: GOOG, MU.

name	delta notional	of NLV
GOOG	$763,364	51.6%	1543 sh, 18 legs
MU	$163,142	11.0%	525 sh, 2 legs
VOO	$121,517	8.2%	173 sh, 0 legs
META	$105,888	7.2%	144 sh, 0 legs
NVDA	$97,653	6.6%	485 sh, 2 legs
RDDT	$73,995	5.0%	513 sh, 0 legs
AMZN	$49,406	3.3%	200 sh, 0 legs
INTC	$34,707	2.3%	300 sh, 0 legs
EQT	$9,758	0.7%	200 sh, 0 legs
CRWD	$2,618	0.2%	10 sh, 0 legs
PANW	$1,939	0.1%	5 sh, 0 legs
TSLA	$353	0.0%	1 sh, 0 legs
HD	$289	0.0%	1 sh, 0 legs
DIS	$105	0.0%	1 sh, 0 legs

Everything else clear

Nearest trigger is GOOGL - tranche 3 confirmation (nearest written level) at $360, 5% away.

All 27 triggers swept, with today's number

price trigger	now	fires at	distance
GOOGL - tranche 3 confirmation (nearest written level)	$341.62	≥ $360	5.4% up
GOOGL - tranche 2 weakness	$341.62	≤ $317	7.2% down
MU - collar roll level	$1,064.92	≥ $1,250	17.4% up
GOOGL - thesis break, close regardless of price	$341.62	≤ $280	18.0% down
GOOGL - washout, stop selling spreads and buy stock	$341.62	≤ $260	23.9% down
GOOGL - Rule 8, deploy the $100k crash reserve	$341.62	≤ $260	23.9% down
GOOGL - no chase past $380-390 short	$341.62	≥ $430	25.9% up
GOOGL - Rule 6, close every short put	$341.62	≥ $434	27.0% up
GOOGL - long spreads near max, would-I-buy-it-fresh test	$341.62	≥ $500	46.4% up
GOOGL - shares capped by the 600Cs	$341.62	≥ $600	75.6% up

level trigger	now	fires at	distance
Net cash / NLV, now red	-0.111	back above -0.10 (amber)	$16,579 of sales at midday NLV.
Every long call spread against the 90%-of-width sell rule	42.80% is the highest	90% of width	Dec-28 300/400 42.80%, Sep-27 300/440 39.54%, Sep-27 300/445 38.86%, Dec-28 350/500 30.23%, Dec-28 350/600 24.37%.
Dec-28 put IV, the alternate tranche-2 trigger	35.66%	43% (Rule 8 at 50%)	7.34 points. Midpoint IV valid at midday (implied_vol 35.92%).
GOOGL Dec-28 340P extrinsic, the assignment watch	$55.09	$5	Put is $1.62 out of the money; the whole mark is extrinsic.
Program cap	$175,059	$200,000	$24,941 of room.
Excess liquidity against the Part 4 amber	75.74% of NLV	40% amber, 25% red	$1,121,503 against a $650,000 floor.
NLV on the week against the 72-hour freeze row	-4.03%	a FALL of 8% in one week	Struck against 2026-09-22's open, $1,542,979.22.

calendar	date	away
MU fiscal Q4 print, after the close	2026-09-30	1d
Earliest clean sale date for the 6/05 GOOG loss lots if only bought calls count as replacement property	2026-10-01	2d
Earliest clean sale date if a WRITTEN call is also replacement property	2026-10-15	16d
GOOG print, first scheduled thesis adjudicator	2026-10-28	29d
Deploy the $100k reserve at market if no crash has come	2026-10-29	30d
GOOG capex guide, second scheduled adjudicator	2027-02-03	127d
GOOG lots vest long-term, add collar put leg 9 (+66 sh)	2027-03-13	165d
GOOG lots vest long-term, add collar put legs 10-11 (+200 sh)	2027-04-02	185d
GOOG lots vest long-term, legs 12-14 (+350 sh)	2027-06-05	249d
GOOG lots vest long-term, leg 15 (+50 sh)	2027-06-25	269d

Spot prices

	last	day
META	$735.33	+2.8%
MU	$1,064.92	+1.0%
CRWD	$261.80	+1.0%
RDDT	$144.24	+0.8%
AMZN	$247.03	+0.4%
VOO	$702.41	-0.2%
GOOG	$338.36	-0.2%
INTC	$115.69	-0.3%
GOOGL	$341.62	-0.3%
DIS	$105.17	-0.4%
HD	$288.69	-0.4%
NVDA	$227.73	-0.5%
PANW	$387.76	-1.1%
TSLA	$352.96	-1.3%
EQT	$48.79	-2.2%

What other people said

D.A. Davidson / Gil Luria on MU: buy, reiterated, $2,000.
Morgan Stanley / Joseph Moore on MU: maintains, $1,200.
6 new calls filed, none came due to score.

Every call filed today

who		action	target	when
D.A. Davidson / Gil Luria Reiterated.	MU	Buy, reiterated	$2,000	2026-09-28
Morgan Stanley / Joseph Moore Maintained.	MU	Maintains	$1,200	2026-09-28
Bernstein / Mark Li Maintained; same $1,300 as 09-01, 09-04, 09-08.	MU	Maintains	$1,300	2026-09-29
Seeking Alpha / Ricardo Fernandez "25% upside to $430 by YE27 and 50% to $515 by 2028"; TPU 44% cost advantage over Nvidia.	GOOGL	Buy	$430 by YE27	2026-09-29
Seeking Alpha / Hunting Alphas $514B cloud backlog, commercial TPU revenue from 2027; DMA remedies a risk; above long-term forward P/E.	GOOGL	Buy	none	2026-09-28
Seeking Alpha / Tech Stock Pros Chinese non-HBM DRAM supply threatens pricing; DRAM price acceleration slowing; memory price collapse coming, timing hard.	MU	Strong Sell into the print	none	2026-09-29

consensus	source	target	analysts
GOOGL	stockanalysis.com	$429.46	61 analysts	Unchanged.
GOOGL	marketbeat.com	$425.60	54 analysts	Unchanged.
GOOGL	finviz.com	$426.50	Recom 1.25	Down from $426.66.
MU	stockanalysis.com	$1,515.00	49 analysts	Unchanged.
MU	marketbeat.com	$1,348.03	45 analysts	Unchanged.
MU	finviz.com	$1,580.88	Recom 1.26	Unchanged.

Cost of compute

H100 spot $2.33 against the $2.82 monthly composite. No tripwire.
H100 SXM $2.33 on Vast off 16 listings, 17% under the $2.82 composite; 24h avg $2.94, 7-day $2.94, 30-day $2.51. The -27.2% day follows Monday's +16.8% and Sunday's -12.7%: the median is swinging a dollar a day on a 16-listing book while the 24h average holds near $2.94, so do not read it as trend. Cheapest cloud H100 $5.00, hyperscaler $7.89, Akash $2.72, all flat on the day. B200 $9.38 on 4 and H200 $4.45 on 7 are thin books. A100 SXM4 $0.91 on 41, flat.
No tripwire, and none is computable before ~2026-11-16: the daily series is too short for a 90-day column.
Capture health: No WARNING. SFCompute unparsed on every capture since 2026-09-25 (five days empty).

Every rate, per chip and per source

Dollars per GPU-hour. Vast is a marketplace floor and runs well below contract rates, which is what the cloud and hyperscaler columns are there to show. Bold rows are current generation.

	vast	cloud	hyper	akash	n
H100 SXM	$2.33	$5.00	$7.89	$2.72	16
H100 NVL	$3.01	-	-	-	11
H200	$4.45	$4.50	$10.85	$4.45	7 thin
B200	$9.38	$7.88	$14.24	-	4 thin
A100 SXM4	$0.91	$2.00	$3.67	$1.71	41
A100 PCIE	$0.74	-	-	-	18
L40S	$0.80	-	-	-	29
RTX 4090	$0.47	-	-	-	64
RTX 5090	$0.59	-	-	-	64

H100 by source. SFCompute clearing price -. Composite is $2.82 as of 2026-04 (Still the 2026-04 H100 print at $2.82, the last free figure on gpu-index.semianalysis.com; later months paywalled (re-checked 2026-09-03). Next monthly append due on the first October open run.).

source	now	24h avg	day
vast med	$2.33	$2.94	-27.2%
sfcompute	-	-	-
getdeploying	$5.00	-	+0.0%
akash	$2.72	$2.72	+0.0%

The check itself

broken Forecasting pipeline refused at pre-flight again at 12:45 today; the gate has not run since Thursday.
ops/status.py --check exits 12: "latest daily record status is 'refused': pre-flight: 187 path(s) under data/ or dist/ already differed from HEAD" and "latest daily record has no gate_examined count: the gate did not run, or ran without reporting. This is the 171f531 shape." gate_examined: none. No COMMIT_REFUSED.
known rails.py still prices the GOOGL programme off GOOG.
GOOG $338.36 vs GOOGL $341.62 at midday. No verdict flips. rails.json not edited.

Rails that could not be evaluated

No written rule covers PANW or CRWD; both were bought while net cash was red.
The long-leg repair order is a cash debit while net cash is red - the two rails conflict and no doc says which wins.
stressed_excess_liquidity has never been measured; the amber stress re-run fell due today (2026-09-29) and is still unmeasured.
MU has no written level below spot, one day before its print.
SFCompute H100 empty five days running.
The 8x Dec-28 300P has no written repair order.

Where the numbers came from

Run label from the rails log: 2026-09-29 open entry present, so this is the MIDDAY run; GPU and forecast passes skipped. GPU and forecast sections carry this morning's open-run figures.
Book pulled ~19:47Z. 36 rows, cash -$164,661.14, unchanged since Monday. No working orders.
Prices: REALTIME get_price_snapshot last; prior close = last minus the snapshot's change, i.e. Monday 09-28 close.
IV: Dec-28 340P option-midpoint-iv 35.66% valid; implied_vol 35.92%; bid/ask 54.95/55.65.
Runbook Parts 2-4 swept on midday numbers: every row clear.