Type: WKAP Radar Feed

WKAP Radar Feed - 2026 - 10 - 01

Market date: 2026-10-01
Canonical URL: https://wkap.ai/radar/wkap-radar-feed-2026-10-01.html

WKAP Radar Feed

*October 1, 2026*

*AI Growth: Cash Earned, Speed Delivered, Capacity Still Owed*

*3 Thesis Objects: MU, CRWV, ORCL*

Preheader:

Make your AI track X alpha.

WKAP Feed compresses earnings inflections, KOL flow, supply-chain
bottlenecks, and short-duration attention into daily AI-ready context.

Set up your AI to read this Feed.

Ask:

“What did I miss?”

## HUMAN USER — START HERE

Rank MU, CRWV, ORCL for research, not expected returns. September 30
disclosures establish Micron's cash generation, CoreWeave's production
capability and an unchanged schedule at one Oracle site. Those are
different endpoints. Micron's $43.973 billion quarterly operating cash
covers $11.110 billion gross cash PPE; CoreWeave's workload benchmark
proves neither utilization nor profit; Oracle's financing-adjusted
outlay is not its free-cash-flow denominator. The decision is where to
promote diligence and where to withhold an incremental AI earnings
assumption. The weakest shared assumption is that current demand
survives renewal, reinvestment and financing costs. Next same-period
cash results, attributable deployment economics or dated capacity
delivery can change this map. A project-specific update cannot settle
another site's risk.

Suggested Prompt:
“Audit the cash-classification and attribution ledgers. Separate
reported facts, management expectations, interpretation and unknowns.
Challenge MU > CRWV > ORCL; name the evidence that reverses each
adjacent pair. Return one decisive data request per object and
executable 7-day/30-day workflows without buy/sell advice.”

Or simply ask: “What did I miss?”

The goal is to turn today's information into trackable thesis objects
and explicit status-changing tests.

## FOR YOUR AI AGENT

This is a structured market-cognition input file. Follow AI demand →
equipment spending → installed/billable capacity → revenue and
operating cash → reinvestment/financing → attributable shareholder
economics. The disclosures do not establish direct transactions among
these three companies. Do not sum supplier revenue, operator
commitments and customer funding as separate final demand.

Process these as trackable_wow objects, not binary signals with
invented resolution dates. Preserve issuer, reporting period, metric
definition, currency, cash-flow category and actual-versus-guidance
status. A whole-company cash bridge cannot identify incremental
project returns. Rank evidence readiness separately from balance-sheet
safety, valuation and portfolio fit. Missing data is not zero;
retrieval failure is not business failure. Record attention
independently and infer no user holdings.

Do not treat this Feed as investment advice.

## AGENT_OUTPUT_SPEC

1. Today in One Sentence: which claim became cash evidence, production
evidence or site-status evidence?
2. Thesis Object Summary: previous → current state and the research
job that changed.
3. Fundamental Evidence vs Attention Flow: issuer measurements versus
independent hypotheses.
4. Evidence-State Audit: REPORTED_FACT / MANAGEMENT_EXPECTATION /
INTERPRETATION / UNRESOLVED_QUESTION.
5. AI Cash-Conversion Comparison: operating cash, investment,
financing and attributable service economics.
6. Cross-Object Ranking: test MU > CRWV > ORCL and explain
adjacent-pair reversals.
7. Business-Proof Comparison: same-period cash coverage, billable
utilization, project-level delivery.
8. Attention Stage / Crowding Risk: narrative shortcuts versus
measured positioning; do not invent the latter.
9. Weakest Assumption by Object: isolate one causal link, not a
generic macro risk list.
10. Most Important Data Point: source, unit, period, denominator and
model cell affected.
11. Attention-to-Rerating Conditions: business proof followed by a
separate valuation test.
12. Attention Fade Conditions: distinguish fewer posts from
deteriorating economics.
13. 7-Day Research Workflow: retrieve, classify, reconcile, preserve
NO_NEW_DATA.
14. 30-Day Research Workflow and WKAP Daily Top 3 Source Follow-Up:
Validate/Upgrade/Downgrade/Invalidate.
15. Portfolio Fit: only if context is supplied; otherwise withhold
suitability and sizing.

## TODAY_SUMMARY

### Part 1 — Main Market Thesis

NEW FACT — REPORTED_FACT: Micron released fiscal Q4 results and
CoreWeave announced Cognition production workloads on Vera Rubin NVL72
on September 30. Oracle/STACK reported no critical-infrastructure
damage after flooding reached New Mexico's Project Jupiter.
MANAGEMENT_EXPECTATION: that site's planned schedule remains
unchanged; this is not completed delivery. UNCHANGED FACT: CoreWeave's
June-quarter financials and Oracle's fiscal-Q1 cash baseline remain
the financial anchors. INTERPRETATION: upgrade only the link each
disclosure measures. UNRESOLVED_QUESTION: which earnings survive
pricing normalization and the capital required to provide service?
Financing costs and renewal prices are causal exposures here, not an
index-level macro forecast.

### Part 2 — Today’s Thesis Objects

MU — Fundamental: fresh cash-flow tables separate operating cash,
gross investment, adjusted FCF and financing deposits. Attention:
@JonahLupton's forward-multiple bull case makes the normalized
earnings denominator the research priority. Primary:
https://investors.micron.com/news/press-release/2026/Micron-Technology-Inc--Reports-Record-Fiscal-Fourth-Quarter-and-Full-Year-2026-Results/default.aspx

CRWV — Fundamental: production and workload-specific throughput are
demonstrated; financial conversion remains open. Attention:
@StockSavvyShay connects concurrency to future demand, which needs a
billable-economics bridge. Primary:
https://www.coreweave.com/news/coreweave-delivers-nvidia-vera-rubin-nvl72-performance-at-production-scale-starting-with-cognition

ORCL — Fundamental: one site's status is updated; Q1 financing and
contracted demand do not guarantee energized capacity. Attention:
@negligible_cap raises a separate Wisconsin delivery-risk question,
not a confirmed issuer delay. Primary:
https://www.oracle.com/news/announcement/project-jupiter-statement-on-southern-new-mexico-floods-2026-09-30/

### Part 3 — Attention Flow Today

Three September 30 posts frame valuation optimism, deployment optimism
and delivery skepticism. Micron's post appeared October 1 Vietnam
time: https://x.com/JonahLupton/status/2105392084949536789 . The
author explicitly disclosed long positions in Micron, SK Square and
Kioxia. The other selected posts showed no holding disclosure, which
does not establish no holding. These are curated questions, not
measured market-wide capital flows. Neither the bullish earnings
forecast nor the third-party power timetable becomes a primary fact by
repetition.

### Part 4 — The Better Question

Surface question: which AI headline is most impressive? Better
question: which model input may change now without confusing
capability, funding or a promise with earned cash? MU permits a
historical cash update; CRWV permits an execution update; ORCL permits
a site-specific schedule-status update. None alone supplies a
justified stock price.

## MARKET_REGIME

RISK_TONE — INTERPRETATION: Selective evidence promotion, not a broad
risk-on call.
MAIN_DRIVER — NEW REPORTED_FACT: Memory cash, production capability
and physical-site status have different evidentiary endpoints.
MARKET_CONTEXT — UNCHANGED FACT: The financial periods differ; current
operational news does not refresh older cash statements.
ATTENTION_ENVIRONMENT — INTERPRETATION: Low-multiple, faster-compute
and power-delay narratives can each jump a missing causal link.
Positioning remains unmeasured.
WKAP_VIEW — UNRESOLVED_QUESTION: What sustainable, attributable cash
is already required by the prevailing valuation? Quotes below record
prices, not answers.

## STATE_CHANGE_LEDGER

MU — Previous: cash durability awaited the new quarterly disclosure.
New reported evidence: September 30 cash-flow tables. Current: rank 1,
promote cash-durability diligence. Decision impact: close the
historical cash-generation question and open normalization. Next
changer: Q1 actual margin, CFO and same-quarter gross cash PPE. This
is an analytical transition, not a claim that the entire memory thesis
is newly proven.

CRWV — Previous: new-platform performance awaited named production
evidence. New reported evidence: Cognition workloads in production.
Current: rank 2, execution upgraded; financial verdict unchanged.
Decision impact: move from “does it work?” to “who pays, how much and
for how long?” Next changer: attributable utilization, realized
pricing and cash economics. A smaller capex bill alone cannot count as
monetization.

ORCL — Previous: delivery and financial conversion remained
unresolved. New reported evidence: New Mexico flood-impact statement;
MANAGEMENT_EXPECTATION: unchanged site schedule. Current: rank 3,
site-status context updated, incremental delivery gate still open.
Next changer: issuer-dated energized capacity and attributable
economic contribution. Wisconsin remains a separate unresolved
question. Prior states describe research baselines; archive
first-added dates below are not purchase dates or public CRM
promotions.

## CROSS_OBJECT_EVIDENCE_MATRIX

Primary financial anchors: MU results linked above; CRWV:
https://investors.coreweave.com/news/news-details/2026/CoreWeave-Reports-Strong-Second-Quarter-2026-Results/default.aspx
; ORCL: https://www.oracle.com/news/announcement/q1fy27-earnings-release-2026-09-10/
and official slide 8:
https://s23.q4cdn.com/440135859/files/content_files/Q1-FY27-Oracle-Earnings-Slides.pdf
. These unlike periods and investment definitions are not a
profitability league table or project-return calculation.

Classification worksheet — REPORTED_FACT: Micron's $12.747bn full-year
customer-contract deposits sit in financing cash flow. CoreWeave's Q2
$0.790bn deferred-revenue change contributes to operating cash flow.
Oracle's supplemental outlay deducts financing-related capex flows and
qualifying customer prepayments, while its $20bn Q1 equity issuance is
financing. INTERPRETATION: never apply one issuer's “customer cash”
adjustment to all three. Micron's annual deposits cannot be subtracted
from its quarterly CFO; CoreWeave's deferred-revenue movement should
not be relabeled financing or recurring earned revenue. Retrieve
reversals and obligations before normalizing cash.

Sensitivity worksheet — INTERPRETATION, not forecast: MU coverage is
43.973/11.110 = 3.96×. Let c be proportional CFO decline and k be
proportional gross-PPE growth from this same-quarter baseline. Simple
break-even is 43.973×(1−c) = 11.110×(1+k). At k=0, c≈74.7%; at an
illustrative k=20%, c≈69.7%. Neither input is a prediction. This
variable-investment version prevents the fixed-capex cushion from
becoming a permanent safety claim. It excludes acquisitions,
distributions, debt obligations and equity valuation.

Attribution worksheet — UNRESOLVED_QUESTION: For CRWV, collect named
cohort capacity, billable utilization, realized net price, service
costs, capital invested and financing terms; do not multiply a
benchmark gain by company revenue. For ORCL, map site, energized date,
capacity, customer acceptance, attributable recognized revenue and
economic contribution. Company-wide cash trends corroborate these
bridges but cannot identify them. Unknown cohort/site fields remain
unquantified, not zero-return assumptions.

## RANKED_UNANSWERED_QUESTIONS

1. MU: how far can margin and working-capital benefits normalize
before self-funding changes? Retrieve the next actual cash statement
and gross-PPE line together, then test both variable inputs above. The
Q1 non-GAAP margin guide of 86.25% versus Q4's 87.0% is an
expectation, not a reported decline.
2. CRWV: does the new deployment earn attractive realized economics
beyond the initial contract window? Retrieve
issuer/customer-attributable utilization and pricing, plus later
financials. September's approximately $40m/MW annualized revenue
metric applies to new three-to-six-month contracts; renewal is the
missing durability test.
3. ORCL: which named capacity becomes service on which date? Retrieve
issuer project updates and reconcile delivered capacity to revenue,
separating New Mexico from Wisconsin. A positive site statement cannot
clear every project's schedule.
4. All three: what normalized cash and reinvestment path would justify
today's price? Only after the classification/attribution work,
construct a valuation range. The research order must not silently
become an expected-return ranking.

## SCENARIO_BRANCHES

Confirm — MU retains same-period CFO/gross-PPE coverage above 1×
without a material margin disappointment; CRWV remains
production-proven but economically unquantified; ORCL remains
delivery-gated. Preserve MU > CRWV > ORCL. This validates research
emphasis, not a purchase.

Mixed — MU margin softens but cash still covers actual investment.
Reduce normalized earnings assumptions as warranted without
automatically removing self-funding. CRWV's gap narrows solely because
investment falls: update financing need, not monetization status. ORCL
receives another funding commitment: update financing availability,
not energized capacity.

Break/reorder — A MU margin miss combined with coverage below 1×
reverses its promotion. CRWV can lead the diligence queue only if new
attributable utilization/pricing/cash evidence is more decision-ready;
its rank must not improve by pretending its own gate passed. ORCL
moves ahead of CRWV when verified capacity conversion and attributable
economics exceed a still benchmark-only CRWV case. A confirmed site
delay without offsetting capacity weakens ORCL's near-term conversion
assumption, not every Oracle business.

No new data — Retain absolute status and the last verified reporting
period. Return the missing disclosure, next retrieval date and
dependency. Silence is neither delivered capacity nor failed
economics; repeat social coverage does not reset evidence age.

## ATTENTION_TRADE_BOARD

Attention Trade Board

WKAP Attention View

Strongest financial evidence change: MU. Cleanest
evidence-to-attention asymmetry: separating MU's financing deposits
from CFO before evaluating durability. Largest
production-to-financial-proof gap: CRWV. Most crowded: undetermined
without positioning data. Highest qualitative narrative-fade risk:
CRWV if benchmark repetition supplies no economic bridge; this is not
a measured probability. Best durable-thesis candidate: MU conditional
on normalized cash and valuation, with neither operator excluded from
later promotion.

## RADAR_OBJECT_INDEX

THESIS_OBJECT_1: MU
THEME: AI memory / cash durability
STATUS: active_trackable — rank 1, promote cash diligence
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $1,054.50 USD (Yahoo Finance, delayed quote, U.S.
premarket, October 1, 08:43:39 EDT / 19:43:39 ICT;
https://finance.yahoo.com/quote/MU/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-16
SETUP_TYPE: Reported cash → normalized coverage
ATTENTION_STAGE: Post-result valuation interpretation
ATTENTION_WINDOW: 1–7 days for digestion; next financial disclosure
for durability
KEY_QUESTION: What operating cash survives changing margins and actual
investment?

THESIS_OBJECT_2: CRWV
THEME: AI cloud / production-to-economic conversion
STATUS: active_trackable — rank 2, execution upgrade only
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $88.00 USD (Yahoo Finance, delayed quote, U.S.
premarket, October 1, 08:43:15 EDT / 19:43:15 ICT;
https://finance.yahoo.com/quote/CRWV/)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-01
SETUP_TYPE: Production proof → billable utilization and pricing
ATTENTION_STAGE: Workload benchmark interpretation
ATTENTION_WINDOW: 1–7 days for attention; 7–30 days for retrieval,
contracts extend longer
KEY_QUESTION: Does deployment generate attributable cash, not merely
faster output?

THESIS_OBJECT_3: ORCL
THEME: AI capacity / project delivery and financing
STATUS: active_trackable — rank 3, incremental delivery gate
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $139.80 USD (Yahoo Finance, Nasdaq Real Time Price,
U.S. premarket, October 1, 08:42:57 EDT / 19:42:57 ICT;
https://finance.yahoo.com/quote/ORCL/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-11
SETUP_TYPE: Funded/contracted capacity → attributable economics
ATTENTION_STAGE: Site-status evidence meets third-party risk interpretation
ATTENTION_WINDOW: 7–30 days for disclosure review; physical delivery
may take longer
KEY_QUESTION: Which capacity is energized, accepted and economically productive?

Timestamped prices are reference observations, not entry
recommendations. Research-review horizons below are not asserted
earnings or delivery dates.

First-added dates follow review of all 68 prior public issues. MU's
standalone card begins September 16:
https://wkap.ai/radar/wkap-radar-feed-2026-09-16.html ; ORCL's begins
September 11: https://wkap.ai/radar/wkap-radar-feed-2026-09-11.html .
CRWV is a new standalone object today; a July 31 basket-context
mention is not an earlier individual thesis card.

## THESIS OBJECTS

### THESIS_OBJECT_1 — MU

CARD_ID: WKAP-RADAR-2026-10-01-MU
CARD_TITLE: Normalize the Cash, Not the Financing Label
TYPE: trackable_wow
THEME: Memory cash capture / durability
STATUS: active_trackable — first research priority
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $1,054.50 USD (Yahoo Finance, delayed quote, U.S.
premarket, October 1, 08:43:39 EDT / 19:43:39 ICT;
https://finance.yahoo.com/quote/MU/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-16
ATTENTION_STAGE: Earnings recognition entering normalization
ATTENTION_WINDOW: 1–7 days; next issuer cash statement for confirmation

WHY_A_READER_CARES_NOW: Fresh cash evidence closes one uncertainty
while an exceptional earnings denominator opens another.
PREVIOUS_STATE: Latest quarter's cash coverage awaited disclosure.
CURRENT_STATE: Cash coverage reported; sustainable level unresolved.
DECISION_OR_MONITORING_PAYOFF: Replace demand speculation with a
same-period cash/investment stress test.

THESIS_SUMMARY

Micron leads because current operating cash is observable, not because
current margins should be permanent. Classification discipline rejects
both a mechanical deposit subtraction and an uncritical peak-earnings
multiple.

WKAP_ANGLE

The surface-level frame: extraordinary earnings make the stock cheap.
The alternative frame: durable earnings and investment requirements
determine whether the denominator is useful.
The key research question: what cash survives normalization without
shifting obligations outside the analysis?

CORE_THESIS

REPORTED_FACT: the evidence matrix reconciles gross cash investment to
company-adjusted FCF. MANAGEMENT_EXPECTATION: Q1 revenue of $61.5bn
±$1.5bn and non-GAAP gross margin of 86.25%; neither is an actual.
INTERPRETATION: higher revenue guidance does not imply margin
expansion. Track operating cash separately from financing deposits and
customer obligations; a correct category does not eliminate economic
obligations.

ATTENTION_TRADE_FRAME

Attention Source: @JonahLupton;
https://x.com/JonahLupton/status/2105392084949536789
Why Today: A fresh bullish valuation argument follows the new results.
Attention Stage: Result recognition → normalized earnings debate.
Attention vs Evidence — Hard evidence: issuer cash-flow statements and guidance.
Primary sources:
https://investors.micron.com/news/press-release/2026/Micron-Technology-Inc--Reports-Record-Fiscal-Fourth-Quarter-and-Full-Year-2026-Results/default.aspx
Attention / interpretation: Forward earnings and multiple opinion;
explicitly long MU, SK Square and Kioxia. The post's forecast and
unverified cash-return claim are not adopted.
Attention Path: Cash result → normalized margin/working capital →
investment needs → valuation.
Attention Asymmetry: A bearish cash-quality shortcut can be wrong
while bullish earnings extrapolation is also unsafe.
Crowding Risk: Positioning unmeasured; denominator certainty is the
narrative risk.
What Could Sustain Attention: Actual cash coverage despite less
favorable margins and continuing investment.
What Could Make Attention Fade: Repeated forecast multiples without
new normalization evidence.
Attention-to-Thesis Conversion: Durable cash after investment supports
a separately tested equity valuation.

WEAKEST_ASSUMPTION: Current cash capture persists enough as margins,
working capital and investment change.
MOST_IMPORTANT_DATA_POINT: Next-quarter actual CFO divided by that
quarter's gross cash PPE, reconciled with margin and working-capital
movements.
NEXT_DATA_POINT: Next issuer financial release/filing; date not
established here. Coverage below 1× removes self-funding; a margin
miss alongside lost coverage reverses promotion. Preserve the narrower
conclusion if only one condition changes.

### THESIS_OBJECT_2 — CRWV

CARD_ID: WKAP-RADAR-2026-10-01-CRWV
CARD_TITLE: A Faster Workload Must Become a Paying Cohort
TYPE: trackable_wow
THEME: AI service monetization / funding needs
STATUS: active_trackable — second priority, execution strengthened
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $88.00 USD (Yahoo Finance, delayed quote, U.S.
premarket, October 1, 08:43:15 EDT / 19:43:15 ICT;
https://finance.yahoo.com/quote/CRWV/)
DATE_FIRST_ADDED_TO_RADAR: 2026-10-01
ATTENTION_STAGE: Production evidence entering monetization test
ATTENTION_WINDOW: 1–7 days for attention; financial proof at later disclosures

WHY_A_READER_CARES_NOW: The new production milestone changes execution
risk without resolving the funding burden.
PREVIOUS_STATE: Named new-platform production awaited confirmation.
CURRENT_STATE: Cognition production confirmed; attributable cash
economics unquantified.
DECISION_OR_MONITORING_PAYOFF: Request billable utilization and
realized pricing rather than another speed comparison.

THESIS_SUMMARY

Capability is necessary for service economics but is not a financial
measurement. The company-wide funding gap is a constraint to
investigate, not proof that the newly installed cohort destroys value.

WKAP_ANGLE

The surface-level frame: more tokens per GPU means more revenue and profit.
The alternative frame: utilization, price realization, costs and
renewal determine who captures efficiency.
The key research question: does the provider retain economic benefit
after the customer receives faster service?

CORE_THESIS

REPORTED_FACT: up to 4.8× total token throughput for SWE-2 inference
versus GB200 NVL72 is a specific workload comparison. September's
approximately $40m/MW figure is annualized revenue divided by related
cluster power for new three-to-six-month contracts, not a fleet margin
or perpetual cash annuity. Primary pricing source:
https://investors.coreweave.com/news/news-details/2026/CoreWeave-Continues-to-Contract-New-Compute-Capacity-at-Higher-Prices/default.aspx
. INTERPRETATION: request renewal economics and attributable cash
before promoting the financial thesis.

ATTENTION_TRADE_FRAME

Attention Source: @StockSavvyShay;
https://x.com/StockSavvyShay/status/2105283154663870712
Why Today: The author connects Cognition's performance to more
concurrent Devin sessions and future demand.
Attention Stage: Deployment recognition → commercial hypothesis.
Attention vs Evidence — Hard evidence: issuer production claim and
June-quarter financial baseline, not the post's monetization
inference.
Primary sources:
https://www.coreweave.com/news/coreweave-delivers-nvidia-vera-rubin-nvl72-performance-at-production-scale-starting-with-cognition
; https://investors.coreweave.com/news/news-details/2026/CoreWeave-Reports-Strong-Second-Quarter-2026-Results/default.aspx
Attention / interpretation: Concurrency can support demand, but does
not measure realized revenue. No holding disclosure visible in the
selected post.
Attention Path: Workload advantage → billable utilization → realized
price → service cash after costs/investment.
Attention Asymmetry: Execution certainty can improve while financial
uncertainty stays unchanged.
Crowding Risk: Unmeasured; applying a benchmark multiplier to the
income statement is the principal shortcut.
What Could Sustain Attention: Attributable deployment revenue,
utilization, renewals and cash contribution.
What Could Make Attention Fade: Benchmark repetition while economic
disclosure remains absent.
Attention-to-Thesis Conversion: Evidence that deployment improves
realized economics, not merely that lower investment narrows the
financing gap.

WEAKEST_ASSUMPTION: Efficiency becomes paid demand and retained
economics rather than lower prices or idle capacity.
MOST_IMPORTANT_DATA_POINT: A cohort-level utilization/realized-price
bridge corroborated by later operating cash, cash investment and
financing costs.
NEXT_DATA_POINT: Subsequent issuer customer/deployment disclosure and
financial release; dates not established here. Weaker realized pricing
or persistently greater funding dependence despite throughput progress
weakens the monetization case, not the demonstrated hardware result.

### THESIS_OBJECT_3 — ORCL

CARD_ID: WKAP-RADAR-2026-10-01-ORCL
CARD_TITLE: Track the Site Before Crediting the Cash
TYPE: trackable_wow
THEME: Contract-to-capacity conversion / financing classification
STATUS: active_trackable — third priority, delivery verification
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $139.80 USD (Yahoo Finance, Nasdaq Real Time Price,
U.S. premarket, October 1, 08:42:57 EDT / 19:42:57 ICT;
https://finance.yahoo.com/quote/ORCL/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-11
ATTENTION_STAGE: Physical-delivery debate with distinct site evidence
ATTENTION_WINDOW: 7–30 days for retrieval; project completion may be longer

WHY_A_READER_CARES_NOW: Site-specific reassurance and a different
site's risk can be incorrectly netted against each other.
PREVIOUS_STATE: Contract conversion and delivery economics unresolved.
CURRENT_STATE: New Mexico schedule expectation maintained; Wisconsin
concern remains third-party interpretation.
DECISION_OR_MONITORING_PAYOFF: Build a site-level delivery ledger and
preserve the gross-capex cash bridge.

THESIS_SUMMARY

Financing enables construction; it neither energizes a campus nor
establishes its return. Oracle's established businesses also
contribute company cash, so aggregate CFO cannot prove incremental
AI-site economics.

WKAP_ANGLE

The surface-level frame: backlog plus funding assures growth, or a
social delay estimate disproves it.
The alternative frame: each project has its own physical milestones
and economic attribution.
The key research question: what named delivered capacity produces
revenue and acceptable contribution on schedule?

CORE_THESIS

REPORTED_FACT: Q1 RPO was $664bn and additional capacity delivered was
850MW; these measure different stages, not an all-energized backlog.
The matrix preserves reported FCF and the separate funding-adjusted
outlay. MANAGEMENT_EXPECTATION: New Mexico's planned schedule is
unchanged. INTERPRETATION: this statement limits a flood-delay
inference at that site, not a power-delivery concern elsewhere.

ATTENTION_TRADE_FRAME

Attention Source: @negligible_cap;
https://x.com/negligible_cap/status/2105272931303485535
Why Today: September 30 discussion reactivates an attributed Wisconsin
power-delivery risk estimate.
Attention Stage: Third-party schedule interpretation, not a newly
confirmed issuer delay.
Attention vs Evidence — Hard evidence: Oracle's Q1 tables and separate
New Mexico statement.
Primary sources:
https://www.oracle.com/news/announcement/q1fy27-earnings-release-2026-09-10/
; https://s23.q4cdn.com/440135859/files/content_files/Q1-FY27-Oracle-Earnings-Slides.pdf
; https://www.oracle.com/news/announcement/project-jupiter-statement-on-southern-new-mexico-floods-2026-09-30/
Attention / interpretation: A useful risk lead; the author also agreed
in discussion that the stock already discounts problems. This is not
an unqualified sell thesis. No holding disclosure visible.
Attention Path: Site power/construction → energization → customer
acceptance → attributable revenue/economics.
Attention Asymmetry: Contract visibility can be strong while physical
timing and retained cash remain uncertain.
Crowding Risk: Unmeasured; geography conflation can distort both
bullish and bearish cases.
What Could Sustain Attention: Dated issuer capacity milestones and
attributable monetization.
What Could Make Attention Fade: Repeated backlog or third-party
timetables without site-specific updates.
Attention-to-Thesis Conversion: Verified delivery plus economic
contribution; company cash corroborates rather than substitutes for
attribution.

WEAKEST_ASSUMPTION: Power, construction and customer acceptance align
early enough to monetize financed capacity on acceptable terms.
MOST_IMPORTANT_DATA_POINT: Issuer-dated energized/accepted capacity
with attributable revenue and economic contribution, identified by
site.
NEXT_DATA_POINT: Next issuer project or financial disclosure, date not
established here. Confirmed slippage without offsetting capacity
weakens near-term conversion; an unconfirmed forecast alone does not
trigger invalidation.

## CROSS_OBJECT_ATTENTION_COMPARISON

Cross-Object Attention Comparison

Cleanest Attention Trade: MU's cash-classification research setup.
Most Evidence-Backed Attention Trade: MU's historical financial
endpoint. Most Crowded Attention Trade: unknown. Highest Fade Risk:
CRWV if economic follow-through is absent from the conversation. Best
Candidate to Become a Durable Thesis: MU conditional on normalization
and valuation. These labels rank research readiness, not prescribe
trades or estimate security returns.

## 7_DAY_RESEARCH_WORKFLOW

REVIEW_AT: 2026-10-08, Vietnam date. Review deadlines are not asserted
financial-report dates. At each update, check these live issuer
indexes: MU: https://investors.micron.com/news/default.aspx ; CRWV:
https://investors.coreweave.com/news/default.aspx ; ORCL:
https://www.oracle.com/news/ . Follow newly dated disclosures,
preserving exact release URL and retrieval time. An unrelated product
announcement or non-binding MOU does not refresh cash, delivery or
contracted economics.

### MU — 7-Day Checks

- Build a typed cash worksheet: fiscal period, CFO, gross cash PPE,
asset proceeds, incentives, adjusted FCF and financing deposits.
Reconcile the issuer measure without mixing annual and quarterly data.
- Record Q1 margin/revenue guidance separately from actuals. Use the
two-variable sensitivity to identify required CFO at alternative
investment levels; do not call it downside valuation.
- Request working-capital and customer-obligation details before
treating the latest cash generation as a recurring run rate.

### CRWV — 7-Day Checks

- Save benchmark workload, input/output/total-token basis, comparator
and production customer. Reject universal speed or company-revenue
multipliers.
- Request cohort billable utilization, realized pricing, contract
duration and renewal terms. Retain “unquantified” when disclosure is
absent; do not infer zero utilization.
- Reconcile later CFO and cash PPE on the same period/basis; separate
deferred-revenue cash, financing proceeds and interest. Attribute gap
changes to operating improvement versus lower investment before
changing status.

### ORCL — 7-Day Checks

- Maintain separate New Mexico and Wisconsin rows with source class,
original milestone, current issuer statement, energized capacity,
acceptance and revenue attribution. Never merge projects based on
similar AI labels.
- Preserve $28bn total capex separately from $18bn net outlay.
Reconcile financing/prepayments and issuance without counting them as
earned cash.
- Retrieve subsequent issuer delivery disclosures; a funding or
backlog headline updates only its own field, leaving physical delivery
unchanged.

### Cross-Object — 7-Day Checks

Return one row per observation: ticker | site/cohort/entity |
release_date | retrieved_at | fiscal_period | metric/unit | definition
| cash-flow category | old → new | evidence_class | model_cell |
absolute_status_effect | relative_rank_effect | unresolved_dependency.
Include an alternative explanation for every apparent improvement. If
none is newly supported, return NO_NEW_DATA and preserve
status/financial period. Before October 8, retain that review date; on
or after October 8, set next_review_at to seven calendar days after
the actual review's Vietnam date, or an earlier newly confirmed
disclosure date if still future. Never reschedule to the current or a
past date, and do not manufacture a status event. Update valuation
assumptions only after economic attribution, not directly from posts.

Primary-source retrieval registry: MU:
https://investors.micron.com/news/press-release/2026/Micron-Technology-Inc--Reports-Record-Fiscal-Fourth-Quarter-and-Full-Year-2026-Results/default.aspx
; CRWV: https://investors.coreweave.com/news/news-details/2026/CoreWeave-Reports-Strong-Second-Quarter-2026-Results/default.aspx
; ORCL: https://www.oracle.com/news/announcement/q1fy27-earnings-release-2026-09-10/
; historical cash anchors, supplemented by exact
production/pricing/site sources in the cards; follow the issuers'
latest news and subsequent financial disclosures for forward updates.
Attention-source retrieval registry: MU:
https://x.com/JonahLupton/status/2105392084949536789 ; CRWV:
https://x.com/StockSavvyShay/status/2105283154663870712 ; ORCL:
https://x.com/negligible_cap/status/2105272931303485535 ; roles:
valuation hypothesis, monetization hypothesis, delivery-risk lead;
none substitutes for primary financial evidence.
Evidence-state snapshot: MU=cash reported/durability open;
CRWV=production confirmed/economics open; ORCL=site status
updated/delivery attribution open; classification=trackable_wow;
rule=update only the measured causal link; source-rule=primary
evidence changes business status, attention changes the research
question.

## 30_DAY_RESEARCH_WORKFLOW

REVIEW_AT: 2026-10-31, Vietnam date, or earlier on a relevant
disclosure. Do not presume all issuers report within the review
window. If no new results are available, retain status and the last
verified financial period and return NO_NEW_DATA. Before October 31,
retain that checkpoint; on or after October 31, set next_review_at to
seven calendar days after the actual review's Vietnam date, or an
earlier newly confirmed disclosure date if still future. Never return
the current or a past date as the next review. Calendar age is not a
financial miss.

### MU — 30-Day Checks

Validate: reconcile actual CFO, investment and issuer-adjusted FCF
when disclosed. Upgrade: durable cash survives normalization and
supports a separate valuation case. Downgrade: worsening
margin/working-capital cash or investment demands reduce the cushion.
Invalidate the self-funding designation: actual same-period coverage
below 1×; a margin miss alongside that loss reverses the broader
research promotion. A falling stock price alone is not this test.

### CRWV — 30-Day Checks

Validate: named production persists and billable-economics evidence
becomes available. Upgrade: attributable utilization/pricing and cash
contribution support financial conversion. Downgrade: lower realized
pricing, weak renewal terms or increased funding dependence contradict
expected benefit. Invalidate the monetization upgrade: performance
advances but disclosed service economics fail to retain value;
preserve the distinct production fact. Do not promote solely from
capex cuts or adjusted EBITDA while cash obligations remain
unexplained.

### ORCL — 30-Day Checks

Validate: dated site delivery/acceptance evidence matches the relevant
project's timeline. Upgrade: attributable revenue/economic
contribution supports conversion, with company cash as corroboration.
Downgrade: confirmed physical slippage or funding costs reduce
expected conversion. Invalidate the near-term delivery assumption:
confirmed slippage without offsetting capacity defeats the modeled
timing. Do not invalidate all Oracle operations from one project or
treat a third-party estimate as confirmation.

### Cross-Object — 30-Day Checks

Return prior → current rank, triggered rule, decisive source and the
strongest competing explanation. Re-test MU's two-variable cash
cushion; CRWV's customer-to-provider benefit allocation; ORCL's
project-level timing and attribution. Separate relative promotion from
an absolute business upgrade. Only then compare normalized equity cash
cases at prevailing prices. End with the single unresolved input most
capable of reversing each conclusion, not three requests to “watch
earnings.”

## WKAP DAILY TOP 3

Three market sources worth feeding into today’s market chat. Not
required reading — WKAP has already extracted the signal.

### 1. @JonahLupton — A Low Forward Multiple Needs a Durable Denominator

URL: https://x.com/JonahLupton/status/2105392084949536789
WKAP signal: September 30 U.S./October 1 Vietnam bullish earnings
interpretation; explicitly long Micron, SK Square and Kioxia.
Why it matters today: The newly reported cash result makes
normalization—not another demand headline—the central valuation
question. His earnings forecast is not issuer guidance.
Themes/tickers: Memory earnings durability, cash classification,
investment coverage / MU.
Question to ask: Which margin, working-capital and actual-investment
assumptions sustain the cash case, and which would reveal that the
apparently low multiple uses a cyclical peak?

### 2. @StockSavvyShay — Concurrency Is a Commercial Hypothesis, Not Revenue

URL: https://x.com/StockSavvyShay/status/2105283154663870712
WKAP signal: September 30 interpretation links Cognition performance
to more Devin sessions and future demand; no holding disclosure
visible.
Why it matters today: A named production milestone permits an
execution upgrade, while the primary workload definition constrains
claims and leaves service economics unproven.
Themes/tickers: AI cloud utilization, realized pricing, renewal and
funding / CRWV.
Question to ask: What attributable billable utilization and
realized-price evidence shows the provider retains economic benefit
rather than passing efficiency to customers or leaving capacity idle?

### 3. @negligible_cap — A Power-Risk Lead Needs a Site-Specific Answer

URL: https://x.com/negligible_cap/status/2105272931303485535
WKAP signal: September 30 attributed Wisconsin delivery-risk
discussion; the author also acknowledged that problems may already be
priced. No holding disclosure visible.
Why it matters today: Oracle's separate New Mexico statement cannot
answer the Wisconsin question, and neither a bearish timetable nor an
unchanged plan proves actual delivery.
Themes/tickers: Energized capacity, project geography, attributable
economics / ORCL.
Question to ask: Which issuer-dated site milestone and attributable
economic contribution would confirm or refute the relevant delivery
assumption without confusing financing, backlog and completed service?

WKAP Radar Feed - 2026 - 10 - 01

Agent-readable facts