WKAP Radar Feed - 2026 - 09 - 29

WKAP Radar Feed

*September 29, 2026*

*AI Capital Headlines: Who Pays, Who Owns, Who Earns?*

*3 Thesis Objects: NVDA, AMD, KKR*

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 NVDA, AMD, KKR by research confidence, not expected return. Three
September 28 announcements change capital allocation, but none alone
changes per-share earnings. Nvidia's new authorization needs executed
share retirement; AMD's stock-funded World Labs purchase spends
ownership; Samsung's Helix commitment requires a bridge to KKR's own
economics. Nvidia's historical half-year cash coverage is stronger
than its latest quarter, making period selection consequential. The
weakest shared assumption is that committed capital becomes value for
existing shareholders. Next filings can reorder the map through net
shares and cash coverage, acquisition issuance and commercial proof,
or disclosed fund fees and parent exposure. Keep unverified strategic
benefits outside the base case.

Suggested Prompt:

“Update these three objects using a cash/ownership/attribution ledger.
Preserve periods and evidence classes. Identify the first model input
that can legitimately change, challenge NVDA > AMD > KKR, and return a
decisive datum, reversal rule and 7-day/30-day retrieval queue 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 capital
authorization or commitment → execution → operating benefit →
attributable economics → value per listed share. Each object tests a
different missing link. Do not add ecosystem funding, supplier sales
and manager fees as independent demand.

Process each as trackable_wow. Primary evidence changes status; X
changes questions. Preserve entity, currency, period, denominator and
actual-versus-expectation status. Missing terms are unknown, not zero.
A zero base-case acquisition uplift is a modeling discipline, not a
claim that the asset has no value. Retrieval failure does not
invalidate a thesis. Portfolio context is absent; withhold sizing and
suitability conclusions.

Do not treat this Feed as investment advice.

## AGENT_OUTPUT_SPEC

1. Today in One Sentence: identify which capital headline can change
which model input.
2. Thesis Object Summary: previous → current evidence state.
3. Fundamental Evidence vs Attention Flow: separate issuer terms from
independent interpretation.
4. Evidence-State Audit: REPORTED_FACT / MANAGEMENT_EXPECTATION /
INTERPRETATION / UNRESOLVED_QUESTION.
5. Capital-Conversion Comparison: cash returned, ownership issued,
fund capital committed.
6. Cross-Object Ranking: test NVDA > AMD > KKR and name adjacent-pair
reversal evidence.
7. Business-Proof Comparison: ownership bridge, commercial payback,
public-parent attribution.
8. Attention Stage / Crowding Risk: distinguish post age from measured
positioning.
9. Weakest Assumption by Object: identify the causal link most exposed
to failure.
10. Most Important Data Point: exact source, unit, period and denominator.
11. Attention-to-Rerating Conditions: require economic proof and a
separate valuation bridge.
12. Attention Fade Conditions: separate conversation decay from
deteriorating economics.
13. 7-Day Research Workflow: retrieve, reconcile, classify; preserve
unavailable inputs.
14. 30-Day Research Workflow and WKAP Daily Top 3 Source Follow-Up:
Validate/Upgrade/Downgrade/Invalidate.
15. Portfolio Fit: only with supplied context; do not infer holdings
from reading interest.

## TODAY_SUMMARY

### Part 1 — Main Market Thesis

NEW FACT — REPORTED_FACT: September 28 brought Nvidia's additional
$150 billion repurchase authorization, AMD's approximately $8.2
billion all-stock World Labs agreement, and Samsung's $1 billion Helix
commitment. These are different instruments, not three interchangeable
cash flows.

UNCHANGED FACT — REPORTED_FACT: The historical quarterly releases
remain operating baselines. INTERPRETATION: Nvidia has the most
observable endpoint; AMD and KKR need progressively more attribution.
UNRESOLVED_QUESTION: what economic benefit reaches each existing
share, at what cost and purchase valuation?

### Part 2 — Today’s Thesis Objects

NVDA — Fundamental: remaining authorization is $235 billion;
management expects execution through fiscal 2028. Attention:
@StockSavvyShay interprets confidence in demand and pricing. Intent
cannot establish execution. Primary:
https://nvidianews.nvidia.com/news/nvidia-announces-a-150-billion-share-repurchase-authorization-increase

AMD — Fundamental: stock-funded capability acquisition, with closing
expected by year-end 2026 subject to conditions. Attention:
@PatrickMoorhead's leadership interview supplies a roadmap mechanism,
not earnings accretion. Primary:
https://ir.amd.com/news-events/press-releases/detail/1299/amd-to-acquire-world-labs-to-advance-the-future-of-ai-compute

KKR — Fundamental: additional third-party commitment to Helix;
public-parent fee/ownership allocation unquantified. Attention:
@ShanuMathew93's June 14 framework explains coordination friction, not
current momentum. Primary:
https://media.kkr.com/news-details/?news_id=ea6af31b-ab59-43ac-b92f-1cb5c126118d

### Part 3 — Attention Flow Today

Two September 28 posts and one explicitly older June 14 operating
framework direct three different tests. The AMD interview is a
mechanism lead:
https://x.com/PatrickMoorhead/status/2104687060535652471 . Post
visibility is not measured capital flow, and these sources do not
independently corroborate one another's earnings claims. No holding
disclosure was visible in the three selected posts; that does not
establish absence of holdings.

### Part 4 — The Better Question

Surface question: which announcement is largest? Better question:
which claim survives the transition from a capital headline to
attributable value per existing share? A larger authorization,
acquisition value or fund pool can increase the work required without
yet increasing the justified forecast.

## MARKET_REGIME

RISK_TONE: INTERPRETATION — Selective capital-conversion research; no
index-level risk-on/risk-off forecast.
MAIN_DRIVER: NEW FACT / REPORTED_FACT — Three issuer announcements
change permissions, ownership commitments and funding availability.
MARKET_CONTEXT: UNCHANGED FACT — Historical cash-flow and share-count
periods differ. INTERPRETATION — AI spending is an exposure, not proof
of incremental end demand or investment returns.
ATTENTION_ENVIRONMENT: INTERPRETATION — Buyback confidence,
strategic-capability debate and an older infrastructure-coordination
framework. Positioning and crowding are unmeasured.
WKAP_VIEW: INTERPRETATION — Upgrade the evidence ledger before the
earnings estimate. UNRESOLVED_QUESTION — What execution, dilution or
fee capture is already priced? Timestamped quotes below are reference
observations, not valuation conclusions.

## STATE_CHANGE_LEDGER

| Object | Previous State | Evidence Change | Current State | Decision
Impact | Next Status-Changing Observation |
|---|---|---|---|---|---|
| NVDA | Existing cash-return program and historical cash generation |
Additional $150bn authorization | Rank 1; execution audit | Record
capacity, not executed retirement | Net outstanding-share bridge plus
rolling cash coverage |
| AMD | Standalone operating roadmap | Approximately $8.2bn
stock-funded agreement | Rank 2; capability/ownership gate | Keep
strategic uplift outside base case | Issuance terms and measurable
commercial milestone |
| KKR | Helix launched with over $10bn commitments | Samsung commits
$1bn | Rank 3; attribution gate | Upgrade funding evidence only | Fee
basis, parent capital exposure and contracted deployment |

Previous states are analytical baselines, not claims that each thesis
appeared in earlier Radar coverage. Rankings and transition rules are
INTERPRETATION; first-added dates are archive metadata, not purchase
dates.

## CROSS_OBJECT_EVIDENCE_MATRIX

| Audit Axis | NVDA | AMD | KKR |
|---|---|---|---|
| Capital instrument | Repurchase permission | Equity consideration |
Third-party fund commitment |
| What can be recorded now | Authorization balance | Announced
transaction terms | Committed capital evidence |
| What cannot be booked from headline | Executed shares retired |
Acquisition revenue/accretion | Parent earnings or supplier orders |
| Historical economic anchor | Q2 FY2027 FCF $21.341bn; H1 $69.895bn |
Q2 2026 FCF $1.558bn | Q2 2026 firmwide FRE about $1.21bn |
| Ownership/attribution anchor | Weighted-average diluted shares
24.285bn versus 24.532bn year earlier | Weighted-average diluted
shares 1.659bn versus 1.630bn year earlier | Anchor financed by KKR
balance sheet and managed vehicles; split undisclosed |
| Missing bridge | Period-end outstanding shares and all capital uses
| Exact issued shares, retention and integration costs; commercial
return | Fee-paying basis, fees, ownership, calls, attributable
project cash |
| Next admissible model change | Observed denominator/cash execution,
with valuation kept separate | Disclosed ownership terms; benefit only
with supported mechanism and scale | Disclosed attributable economics,
not commitment multiplied by guessed fee |

REPORTED_FACT anchors have different fiscal periods and definitions;
FRE is KKR's non-GAAP adjusted measure, not free cash flow. Do not
compare their dollar magnitudes as a profitability league table.

Cash worksheet — INTERPRETATION arithmetic, USD billions: Nvidia Q2
distributions = $19.732 repurchases + $6.047 dividends = $25.779,
exceeding FCF by $4.438. H1 distributions = $39.044 + $6.290 =
$45.334, below H1 FCF by $24.561 before other capital uses. Q2 also
contains $24.896 net debt issuance and $15.822 equity-security
purchases. Neither the quarterly deficit nor the half-year excess
assigns particular debt dollars to repurchases. Reconcile the whole
cash statement.

Ownership worksheet — INTERPRETATION, not forecast: let S>0 be
pre-deal shares on a consistent basis and N≥0 the actual new
deal/retention shares. Existing owners retain S/(S+N). For positive
comparable earnings E, incremental earnings must be at least E×N/S to
avoid EPS dilution; equality gives neutrality and exceeding it gives
accretion before financing/integration effects. EPS neutrality is not
value creation. Neither AMD's weighted-average 1.659bn nor today's
quote establishes final S or N; keep the worksheet symbolic until
terms are disclosed.

Attribution worksheet — INTERPRETATION: distinguish fee-eligible
capital × disclosed fee rate, net of incremental expenses, from
balance-sheet investment returns and realized performance income.
Apply each stream's actual ownership and accounting treatment. Do not
apply KKR's firmwide average economics to Samsung's commitment. The
fund principal is not a fourth earnings stream.

## RANKED_UNANSWERED_QUESTIONS

1. NVDA: did gross repurchases become net ownership reduction without
degrading funding flexibility? Retrieve outstanding-share and issuance
reconciliations alongside the full cash statement.
2. AMD: what precisely is being issued, and which commercial milestone
can earn it back? Retrieve transaction exhibits and named
product/workload evidence; do not manufacture an exchange ratio.
3. KKR: which entity receives fees, carries investment risk and makes
capital calls? Retrieve Helix terms before inserting a public-parent
contribution.
4. All three: what is already required by valuation? Build a separate
scenario range after attribution, not a target price from announcement
size. This unresolved step prevents turning the research order into a
trade ranking.

## SCENARIO_BRANCHES

Base — Keep NVDA > AMD > KKR while new announcements remain
unconverted. Preserve existing operating forecasts; create separate
contingent rows for strategic benefits.
Ownership proof — AMD supplies issuance terms and measurable
commercial improvement sufficient to test a per-share benefit, while
NVDA's next execution bridge fails cash/share tests: AMD can outrank
NVDA. A product name alone does not satisfy the branch.
Attribution proof — KKR discloses material, contracted and
attributable economics with controlled parent capital needs while AMD
remains commercially untestable: KKR can outrank AMD. Another
commitment alone cannot reorder them.
Negative execution — Net issuance offsets Nvidia's repurchases, AMD's
measurable costs grow without differentiation, or Helix calls precede
supported deployment economics: demote the affected thesis
independently. Do not interpret missing disclosure as an observed
loss.

## ATTENTION_TRADE_BOARD

Attention Trade Board

| Object | Attention Stage | Attention Source | Why Today | Hard
Evidence | Narrative Gap | Crowding Risk | Likely Window | Fade Signal
|
|---|---|---|---|---|---|---|---|---|
| NVDA | Authorization interpretation | @StockSavvyShay, Sep28 | New
capacity for returns | Issuer authorization and historical cash |
Intent versus net retirement | Unmeasured; confidence shortcut | 1–7
days, then filing | Repetition without execution evidence |
| AMD | Capability interpretation | @PatrickMoorhead, Sep28 |
Stock-funded acquisition | Definitive agreement | Engineering
mechanism versus per-share return | Unmeasured; strategic halo | 7–30
days and closing | Roadmap language without measurable benefit |
| KKR | Funding update/older context | @ShanuMathew93, Jun14 | Samsung
commitment | New capital and founding terms | Fund size versus parent
economics | Unmeasured; ecosystem double-counting | 7–30 days,
deployment longer | Funding headlines without attribution |

WKAP Attention View

The clearest prospective ownership change is AMD's agreement; the most
measurable historical endpoint is NVDA's cash/share record. KKR has
the largest attribution gap. The cleanest research asymmetry is
separating instrument types before modeling. No most-crowded stock is
established. Durable attention requires execution evidence; narrative
fade alone does not prove business deterioration.

## RADAR_OBJECT_INDEX

THESIS_OBJECT_1: NVDA
THEME: AI cash conversion / capital return
STATUS: active_trackable — rank 1, execution audit
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $230.49 USD (Yahoo Finance, NasdaqGS delayed quote,
U.S. premarket, September 29, 08:44:43 EDT / 19:44:43 ICT;
https://finance.yahoo.com/quote/NVDA/)
DATE_FIRST_ADDED_TO_RADAR: 2026-08-25
SETUP_TYPE: Authorization-to-ownership bridge
ATTENTION_STAGE: Authorization interpretation
ATTENTION_WINDOW: 1–7 days; financial confirmation at next filing,
date not established here
KEY_QUESTION: Does execution improve net ownership without weakening
cash flexibility?

THESIS_OBJECT_2: AMD
THEME: AI capability / stock-funded acquisition
STATUS: active_trackable — rank 2, commercial-return gate
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $614.42 USD (Yahoo Finance, NasdaqGS delayed quote,
U.S. premarket, September 29, 08:44:17 EDT / 19:44:17 ICT;
https://finance.yahoo.com/quote/AMD/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-29
SETUP_TYPE: Capability-to-per-share-return bridge
ATTENTION_STAGE: Agreement and engineering interpretation
ATTENTION_WINDOW: 7–30 days; closing expected by year-end 2026, not assured
KEY_QUESTION: Which measured benefit clears the ownership hurdle?

THESIS_OBJECT_3: KKR
THEME: AI infrastructure / attributable fund economics
STATUS: active_trackable — rank 3, attribution gate
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $94.15 USD (Yahoo Finance, NYSE—Nasdaq Real Time
Price, U.S. premarket, September 29, 08:14:01 EDT / 19:14:01 ICT;
older last trade, rechecked near 19:46 ICT;
https://finance.yahoo.com/quote/KKR/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-29
SETUP_TYPE: Commitment-to-parent-economics bridge
ATTENTION_STAGE: Fresh funding plus older operating framework
ATTENTION_WINDOW: 7–30 days for terms; deployment can take longer
KEY_QUESTION: What does KKR earn and risk rather than merely manage?

## THESIS OBJECTS

### THESIS_OBJECT_1 — NVDA

CARD_ID: WKAP-RADAR-2026-09-29-NVDA
CARD_TITLE: Audit Retirement, Not the Authorization Ceiling
TYPE: trackable_wow
THEME: Cash conversion / net ownership
STATUS: active_trackable — first research priority
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $230.49 USD (Yahoo Finance, NasdaqGS delayed quote,
U.S. premarket, September 29, 08:44:43 EDT / 19:44:43 ICT;
https://finance.yahoo.com/quote/NVDA/)
DATE_FIRST_ADDED_TO_RADAR: 2026-08-25
ATTENTION_STAGE: Confidence interpretation entering execution audit
ATTENTION_WINDOW: 1–7 days; next filing for financial confirmation

WHY_A_READER_CARES_NOW: A larger permission can be mistaken for a
completed per-share benefit.
PREVIOUS_STATE: Historical cash generation and an existing repurchase program.
CURRENT_STATE: Higher authorized capacity; new execution still unobserved.
DECISION_OR_MONITORING_PAYOFF: Reconcile net ownership and
quarter/half-year cash before changing the capital-return assumption.

THESIS_SUMMARY

Nvidia leads because there is a reported cash-and-share endpoint to
test. The latest quarter and half-year tell different coverage
stories; neither is sufficient alone.

WKAP_ANGLE

The surface-level frame: buybacks demonstrate undervaluation.
The alternative frame: spending permission is neither retirement nor
evidence of attractive purchase prices.
The key research question: what net ownership change survives
compensation, other issuance and competing cash uses?

CORE_THESIS

REPORTED_FACT: The matrix preserves historical cash distributions and
diluted-share averages. Q2 FY2027 ended July 26. Weighted-average
shares fell about 1% year over year, but that EPS denominator is not
the period-end retirement ledger. MANAGEMENT_EXPECTATION: execution
through fiscal 2028. INTERPRETATION: reported conversion earns
research priority, not an automatic earnings uplift from the new
ceiling.

ATTENTION_TRADE_FRAME

Attention Source: @StockSavvyShay, September 28;
https://x.com/StockSavvyShay/status/2104535870560940501
Why Today: New repurchase capacity activates a testable confidence claim.
Attention Stage: Announcement recognition → execution audit.
Attention vs Evidence — Hard evidence: issuer authorization and
reported financial tables. Primary sources:
https://nvidianews.nvidia.com/news/nvidia-announces-a-150-billion-share-repurchase-authorization-increase
; https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-second-quarter-fiscal-2027
Attention / interpretation: GPU-demand and pricing confidence, not
proof of either; no holding disclosed in the selected post.
Attention Path: Authorization → spending → net shares → cash/value per share.
Attention Asymmetry: Different coverage periods correct both an overly
bullish ceiling argument and an overly bearish single-quarter funding
inference.
Crowding Risk: Unmeasured; a familiar buyback narrative can hide issuance.
What Could Sustain Attention: Reconciled net retirement and durable
operating cash conversion.
What Could Make Attention Fade: No execution detail beyond the authorization.
Attention-to-Thesis Conversion: Net ownership improves with credible
cash coverage, followed by valuation work on purchase prices.

WEAKEST_ASSUMPTION: Working-capital absorption normalizes sufficiently
without sacrificing investment flexibility.
MOST_IMPORTANT_DATA_POINT: Outstanding-share/issuance bridge paired
with rolling cash coverage.
NEXT_DATA_POINT: Next filed quarterly cash and equity disclosures;
date not established in this Feed. Demote if repurchases fail to
offset issuance or distributions persistently exceed generation
without credible financing. Working-capital outflow alone is not
customer default.

### THESIS_OBJECT_2 — AMD

CARD_ID: WKAP-RADAR-2026-09-29-AMD
CARD_TITLE: A Cash-Light Purchase Still Spends Ownership
TYPE: trackable_wow
THEME: Spatial intelligence / commercial return per share
STATUS: active_trackable — second priority, no automatic acquisition uplift
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $614.42 USD (Yahoo Finance, NasdaqGS delayed quote,
U.S. premarket, September 29, 08:44:17 EDT / 19:44:17 ICT;
https://finance.yahoo.com/quote/AMD/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-29
ATTENTION_STAGE: Capability recognition entering ownership test
ATTENTION_WINDOW: 7–30 days; conditional year-end closing target

WHY_A_READER_CARES_NOW: The announced consideration changes how
existing owners should assess strategic upside.
PREVIOUS_STATE: Standalone AI roadmap and historical diluted-share growth.
CURRENT_STATE: Stock-funded World Labs agreement; exact ownership and
payback unquantified.
DECISION_OR_MONITORING_PAYOFF: Hold strategic uplift outside the base
case until terms and a measurable benefit support a per-share test.

THESIS_SUMMARY

Internal world-model expertise could improve compute design. The
investment case requires more than capability: retained talent,
integration costs, product translation and customer value must clear
the ownership issued.

WKAP_ANGLE

The surface-level frame: all-stock funding protects cash.
The alternative frame: cash is preserved by sharing future economics
with new owners.
The key research question: what differentiated workload outcome is
large enough to justify that transfer?

CORE_THESIS

REPORTED_FACT: Approximately $8.2bn all-stock agreement; no exchange
ratio in the announcement. Historical Q2 ended June 27, with FCF
$1.558bn and weighted-average diluted shares 1.659bn.
MANAGEMENT_EXPECTATION: closing by year-end, subject to approvals and
customary conditions. INTERPRETATION: preserve the existing operating
thesis separately from a speculative acquisition contribution.

ATTENTION_TRADE_FRAME

Attention Source: @PatrickMoorhead, September 28 U.S. date;
https://x.com/PatrickMoorhead/status/2104687060535652471
Why Today: The interview offers an engineering mechanism behind the agreement.
Attention Stage: Strategic rationale → testable product milestone.
Attention vs Evidence — Hard evidence: agreement and historical
results. Primary sources:
https://ir.amd.com/news-events/press-releases/detail/1299/amd-to-acquire-world-labs-to-advance-the-future-of-ai-compute
; https://ir.amd.com/news-events/press-releases/detail/1295/amd-reports-second-quarter-2026-financial-results
Attention / interpretation: Internal experimentation may inform
hardware/software roadmaps; the post does not identify a first product
or customer. No holding disclosed in the selected post.
Attention Path: Research insight → product choice → measurable
workload advantage → adoption → attributable return.
Attention Asymmetry: A plausible mechanism is worth investigating
without converting it prematurely into revenue.
Crowding Risk: Unmeasured; strategic enthusiasm can outrun a dilution
calculation.
What Could Sustain Attention: Published issuance/retention terms and
named, measurable commercial milestones.
What Could Make Attention Fade: Repeated vision statements without a
first observable benefit.
Attention-to-Thesis Conversion: Demonstrated benefit can be evaluated
against ownership, costs and opportunity cost—not merely EPS
neutrality.

WEAKEST_ASSUMPTION: Proprietary research yields commercial
differentiation that customers value more than its acquisition and
integration cost.
MOST_IMPORTANT_DATA_POINT: Actual issued-share/retention terms
together with the first measurable workload or product milestone.
NEXT_DATA_POINT: Transaction exhibits, closing documents and issuer
product disclosures. If costs/dilution become measurable while
differentiation remains untestable, demote the incremental thesis;
failed closing removes this acquisition path, not AMD's entire
operating case.

### THESIS_OBJECT_3 — KKR

CARD_ID: WKAP-RADAR-2026-09-29-KKR
CARD_TITLE: Fund Capital Is Not Public-Parent Earnings
TYPE: trackable_wow
THEME: Coordinated AI infrastructure / fee and capital attribution
STATUS: active_trackable — third priority, funding improved/earnings unproven
POSITION_CONTEXT: [not provided]
PRICE_AT_PUBLISH: $94.15 USD (Yahoo Finance, NYSE—Nasdaq Real Time
Price, U.S. premarket, September 29, 08:14:01 EDT / 19:14:01 ICT;
older last trade, rechecked near 19:46 ICT;
https://finance.yahoo.com/quote/KKR/)
DATE_FIRST_ADDED_TO_RADAR: 2026-09-29
ATTENTION_STAGE: Funding update with older operating context
ATTENTION_WINDOW: 7–30 days for terms; project conversion longer

WHY_A_READER_CARES_NOW: A new strategic investor validates funding
availability without specifying what public shareholders earn.
PREVIOUS_STATE: Helix launch commitments and a mixed
balance-sheet/managed-vehicle anchor.
CURRENT_STATE: Samsung adds $1bn; parent attribution remains unresolved.
DECISION_OR_MONITORING_PAYOFF: Request fee basis, ownership, calls and
contracted project economics before changing earnings.

THESIS_SUMMARY

Coordinating land, power and connectivity can address real delivery
friction. Outside capital finances the effort; it is not itself
recurring KKR income or supplier revenue.

WKAP_ANGLE

The surface-level frame: more AI funding means more KKR earnings.
The alternative frame: the recipient, fee basis, capital at risk and
timing determine shareholder benefit.
The key research question: which economic stream is attributable to
the listed parent, net of its obligations?

CORE_THESIS

REPORTED_FACT: Samsung's $1bn commitment builds on over $10bn at
launch; KKR's anchor combines balance-sheet and managed-vehicle
capital. Strategic investors may receive priority/first-look supply
rights, not guaranteed purchase orders. Historical Q2 firmwide FRE was
about $1.21bn; it includes more than management fees and provides no
Helix-specific forecast. INTERPRETATION: funding evidence improves
while earnings attribution remains the gating uncertainty.

ATTENTION_TRADE_FRAME

Attention Source: @ShanuMathew93, June 14 — older operating context,
not fresh September momentum;
https://x.com/ShanuMathew93/status/2066170841163853912
Why Today: Samsung's new commitment makes the coordination model relevant again.
Attention Stage: Old mechanism revisited against a new funding fact.
Attention vs Evidence — Hard evidence: new commitment and founding
allocation language. Primary sources:
https://media.kkr.com/news-details/?news_id=ea6af31b-ab59-43ac-b92f-1cb5c126118d
; https://media.kkr.com/news-details?news_id=6a26acd6-83b8-4377-84be-b1dadb847806
; firmwide baseline:
https://ir.kkr.com/media/document/15a25fa9-8226-4da1-9ff3-027b9803d436/assets/KKR_Q226_Earnings_Release.pdf?disposition=inline
Attention / interpretation: Fragmented procurement creates delivery
friction; the framework proves no utilization, fees or investment
returns. No holding disclosed in the selected post.
Attention Path: Capital → coordinated delivery → contracted
utilization → attributable fees/returns.
Attention Asymmetry: The useful gap is mapping economic rights, not
claiming hidden earnings.
Crowding Risk: Unmeasured; count overlapping ecosystem capital only once.
What Could Sustain Attention: Disclosed fee-paying basis, manageable
parent calls and binding project/customer terms.
What Could Make Attention Fade: Larger commitments without deployment
or attribution detail.
Attention-to-Thesis Conversion: Observable, material parent economics
with controlled capital requirements.

WEAKEST_ASSUMPTION: Coordination improves delivery enough to produce
attractive returns after power, construction and financing costs.
MOST_IMPORTANT_DATA_POINT: A Helix-specific parent
fee/ownership/capital-call bridge.
NEXT_DATA_POINT: Next issuer fund or project disclosure; no date
established here. Demote if capital calls outrun supported contracted
economics or project returns deteriorate. Missing terms remain
uncertainty, not proof of poor returns.

## CROSS_OBJECT_ATTENTION_COMPARISON

Cross-Object Attention Comparison

| Rank | Object | Attention Asymmetry | Evidence Quality | Catalyst
Clarity | Crowding Risk | Attention Window | Conversion Potential |
|---|---|---|---|---|---|---|---|
| 1 | NVDA | Ceiling versus cash and ownership | Direct historical
statements; new execution absent | Authorization clear; next filing
date not established | Unmeasured | Days, then filing | Net retirement
with cash/valuation discipline |
| 2 | AMD | Capability versus ownership cost | Agreement clear; payoff
unquantified | Conditional closing target | Unmeasured | Weeks, then
closing | Commercial proof exceeding dilution and costs |
| 3 | KKR | Funding versus attributable economics | Commitment clear;
parent bridge missing | Deployment timing unresolved | Unmeasured |
Weeks to longer deployment | Fees/returns with controlled capital
exposure |

Cleanest Attention Trade: instrument classification, a research
asymmetry rather than a prescribed trade. Most Evidence-Backed
Attention Trade: NVDA's historical endpoint. Most Crowded Attention
Trade: undetermined. Highest Fade Risk: AMD if only roadmap language
follows, and KKR if only commitments follow; no measured probability
ranking. Best Candidate to Become a Durable Thesis: NVDA, conditional
on execution. These are analytical priorities, not security-return
predictions.

## 7_DAY_RESEARCH_WORKFLOW

REVIEW_AT: 2026-10-06, Vietnam date. This is a research deadline, not
a claimed earnings date.

### NVDA — 7-Day Checks

Retrieve latest equity and cash-flow tables; record period-end shares,
issuance, repurchases and compensation separately; reconcile quarter
and half-year distributions against FCF and other capital uses. Never
convert stock-compensation expense dollars into shares. Record
remaining authorization separately from executed dollars.

### AMD — 7-Day Checks

Retrieve any transaction exhibit, exchange formula, actual share
issuance and retention arrangements; record conditions and closing
status; identify one named product/workload outcome with a comparable
performance/cost measure. Keep expected benefits separate from
reported results. If terms remain absent, preserve symbolic dilution
inputs.

### KKR — 7-Day Checks

Retrieve fund/project disclosures; map each capital provider to the
fee recipient, investment owner and caller of capital; distinguish
committed, called, deployed and fee-paying capital. Find binding
customer/project terms before treating strategic supply rights as
orders. Reconcile any new claim with parent disclosure.

### Cross-Object — 7-Day Checks

Return one observation row per changed input: object | source URL |
publication/retrieval time | reporting period | entity | metric/unit |
old value → new value | evidence class | model cell affected | rank
effect | unresolved dependency. Reject rows that change an earnings
cell using only an attention source. If nothing changes, report “no
new admissible model input” and retain the queue.

Primary-source retrieval registry: NVDA:
https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-second-quarter-fiscal-2027
; AMD: https://ir.amd.com/news-events/press-releases/detail/1299/amd-to-acquire-world-labs-to-advance-the-future-of-ai-compute
; KKR: https://media.kkr.com/news-details/?news_id=ea6af31b-ab59-43ac-b92f-1cb5c126118d
; roles: cash/equity baseline, transaction terms, funding/attribution;
retrieve later issuer filings before updating.

Attention-source retrieval registry: NVDA:
https://x.com/StockSavvyShay/status/2104535870560940501 ; AMD:
https://x.com/PatrickMoorhead/status/2104687060535652471 ; KKR:
https://x.com/ShanuMathew93/status/2066170841163853912 ; roles:
confidence hypothesis, engineering mechanism, older coordination
framework; never replace primary evidence.

Evidence-state snapshot: NVDA=authorization expanded/execution
pending; AMD=agreement signed/ownership and payback unresolved;
KKR=funding improved/parent attribution unresolved;
classification=REPORTED_FACT|MANAGEMENT_EXPECTATION|INTERPRETATION|UNRESOLVED_QUESTION;
rule=update only the supported model cell; source-rule=primary sets
status, attention sets questions.

## 30_DAY_RESEARCH_WORKFLOW

REVIEW_AT: 2026-10-29, Vietnam date. Absence of a scheduled result is
not missed guidance. Longer execution/closing horizons remain open.

### NVDA — 30-Day Checks

Validate: reconcile new actual shares and cash with prior periods.
Upgrade: net ownership reduction and credible rolling coverage survive
other capital uses, then assess purchase valuation. Downgrade: rising
issuance or persistent coverage pressure weakens the bridge.
Invalidate this incremental capital-return thesis: executed spending
does not improve ownership economics and no credible repair is
established; do not equate this with invalidating every Nvidia
business thesis.

### AMD — 30-Day Checks

Validate: track closing conditions, issuance/retention terms and first
commercial evidence. Upgrade: measured differentiation permits a
credible benefit-per-share test. Downgrade: costs and dilution advance
while commercial milestones slip or remain untestable. Invalidate the
acquisition pathway: failed closing, or observed integration economics
that defeat the proposed advantage. Neither outcome alone proves the
standalone business failed.

### KKR — 30-Day Checks

Validate: distinguish commitment growth from calls, deployment and
fee-paying capital. Upgrade: disclosed attributable economics and
binding delivery/customer terms justify a model contribution.
Downgrade: parent calls or project costs rise ahead of supported
economics. Invalidate the Helix shareholder-capture thesis: observed
terms or project outcomes show the proposed benefit does not reach
shareholders on acceptable economics; missing disclosure alone is
insufficient.

### Cross-Object — 30-Day Checks

Revisit the three source questions, identify newly resolved
dependencies, and rerank only after testing the scenario branches.
Keep an unchanged row when evidence is absent. Finally build valuation
sensitivities for execution prices, acquisition benefits/issuance and
fund attribution; identify which assumption dominates downside before
considering portfolio fit.

## WKAP DAILY TOP 3

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

### 1. @StockSavvyShay — Buyback Confidence Needs an Execution Ledger

URL: https://x.com/StockSavvyShay/status/2104535870560940501
WKAP signal: September 28 opinion reads the authorization as
confidence in GPU demand and pricing; no holding disclosed in the
selected post.
Why it matters today: It provides the hypothesis to test against cash
and ownership, not independent proof of demand or value at repurchase
prices.
Themes/tickers: Capital return, cash conversion, net ownership / NVDA.
Question to ask: Does actual execution reduce net ownership claims
while preserving cash flexibility, and what valuation test remains
after that bridge passes?

### 2. @PatrickMoorhead — Internal Models Can Inform the Compute Roadmap

URL: https://x.com/PatrickMoorhead/status/2104687060535652471
WKAP signal: September 28 interview with AMD AI leadership connects
internal experimentation to hardware/software choices; no holding
disclosed in the selected post.
Why it matters today: A plausible engineering mechanism narrows the
research queue, but identifies neither a first commercial
product/customer nor an earnings timetable.
Themes/tickers: Spatial intelligence, acquisition dilution, commercial
differentiation / AMD.
Question to ask: Which named workload advantage, ownership terms and
integration costs permit a genuine per-share return test instead of a
strategic narrative?

### 3. @ShanuMathew93 — Coordinating Infrastructure Does Not Specify Fee Capture

URL: https://x.com/ShanuMathew93/status/2066170841163853912
WKAP signal: June 14 operating framework identifies fragmented land,
power and fiber procurement; explicitly older context, not fresh KOL
momentum. No holding disclosed in the selected post.
Why it matters today: Samsung's new commitment revives the execution
question while leaving utilization, public-parent attribution and
returns unresolved.
Themes/tickers: AI infrastructure coordination, capital commitments,
shareholder capture / KKR.
Question to ask: Which disclosed fee rights, capital obligations and
binding project economics show that coordinated funding becomes value
for the listed parent?