# Cape Fear Advisors, LLC
> Independent structural analysis and operating advisory. Two practices — market dynamics, operating structure, and valuation on one side; growth strategy, M&A, and PE operating execution on the other — grounded in 40+ years of operating experience, 70+ transactions, and $2.4 billion in aggregate value.

Cape Fear Advisors, LLC is an advisory firm founded by Greg Collins. The firm operates two practices from a single credential base: independent structural analysis — market dynamics, operating structure, and valuation through the lens of public company disclosure — and operating and growth advisory for privately held software and services businesses.

Greg Collins serves concurrently as Founder & Principal of Cape Fear Advisors and as CEO of C3 Metrics (https://c3metrics.com), an independent marketing attribution company with no commercial ties to the channel categories it measures.

## Practice 1: Independent Structural Analysis

Market dynamics, operating structure, and valuation — analyzed through the lens of public company disclosure, from an operator's perspective. This is structural observation grounded in four decades of operating and transaction experience, not investment advice.

Current published work covers nine registers:

### NVIDIA
Structural analysis of NVIDIA's position in the AI compute capital structure — the fourth house that sells the machines, collects in 45 days at a 75% margin, holds an ~$82 billion equity portfolio in its own buyers with $27 billion more committed, and drew $25 billion from the same bond market its customers depend on, nothing pledged. "The Fourth House" reads NVIDIA through the houses-and-players framework, finding a revenue engine where the other houses hold placements: four sides of a single counter (register, equity stake, order forms, and the disclosure gap between proxy and financial statements), a circuit where the machine seller's margin and the machine buyer's shortfall are the same number, and an ante collected in days against players who recover in years. The quality-of-cash framework prices NVIDIA's August print through five lines: the $40.7 billion receivable (~45 days, $4 million allowance, three customers at 64%), the portfolio (marks ran 23 cents of every dollar of net income), the taxonomy recast ("Neocloud builders" became "AI Clouds," comparable periods recast), the margin (71%/75%), and the CoreWeave placements. Greg Collins was quoted by name four times in MarketWatch's feature coverage of NVIDIA's circular financing (June 2026). "NVIDIA, the Forge House" reads how the circle is recorded: the same two-billion-dollar CoreWeave equity investment, re-expressed as the price concession it was not, moves reported operating income, segment growth rate, and gross profit without moving a dollar of cash. In the quarter just filed, the size of that choice is seven points of growth rate and a two-billion-dollar swing in reported gross profit, on cash that is identical in every column. The piece extends the ladder of circular financing to the specific structures in the NVIDIA-CoreWeave relationship (vendor financing at rung six, equity stake at rung seven, capacity backstop at rung eight) and reads the second derivative, finding that the form of the investment moves the signal the careful observer relies on to detect a turn. Eight pieces published including the forge-house counterfactual, the fourth-house reading, the quality of cash application, and a twelve-element valuation decomposition.
- **NVIDIA, the Forge House** — https://www.capefearadvisors.com/strategy-lab/nvidia-the-forge-house.html
- **NVIDIA, The Fourth House** — https://www.capefearadvisors.com/strategy-lab/nvidia-the-fourth-house.html

### SpaceX
Comprehensive deconstruction of SpaceX from S-1 through the first public quarter, covering capital structure, the $235B cash gap widened to about $400B with every added dollar going to the AI build, the Terafab fab gone quiet, the $25B minimum-cash commitment whose size reached the public through Fitch rather than the filings, the fixed-baseline framework (a company changing its mind is the event a fixed baseline exists to catch), total addressable market analysis, disclosure patterns, and the structural questions the filing raises. Post-IPO coverage includes the $60B Anysphere/Cursor acquisition in stock, investment-grade ratings (Moody's Baa1, Fitch BBB+, S&P BBB), the first bond offering, the $25B minimum cash covenant, the Terafab record, the revenue side of the cash gap, and the first valuation piece: the cost of not beginning, read through Damodaran's posted model against six weeks of tape, finding two fifths of the offering price above the standing businesses and the entire decline sitting in the layer above everything the model is willing to count. Four traveling frameworks: three-layer decomposition (floor, expansion options, residual), dual-record measurement (filed vs. unfiled), the cost-of-not-beginning concept (time passing against a date the issuer set), and the fixed-baseline discipline (measuring a company against its own stated plan rather than rebasing when the plan moves). The State of the Merger reads the Tesla-SpaceX entanglement through both filings at once, finding the combination already happening in the present tense. Approximately nineteen pieces published on Substack covering the S-1 through post-IPO structural moves.
- **SpaceX, Adding It Up: The $235/$400 Billion Cash Gap** — https://www.capefearadvisors.com/strategy-lab/spacex-235-400-billion-cash-gap.html
- **SpaceX and Tesla, Adding It Up: The State of the Merger** — https://www.capefearadvisors.com/strategy-lab/spacex-tesla-state-of-merger.html
- **SpaceX, Adding It Up: The Cost of Not Beginning** — https://www.capefearadvisors.com/strategy-lab/spacex-cost-of-not-beginning.html
- **SpaceX, Adding It Up: The Ninety-Day Annuity** — https://www.capefearadvisors.com/strategy-lab/spacex-ninety-day-annuity.html
- **SpaceX, Adding It Up: The Terafab Record** — https://www.capefearadvisors.com/strategy-lab/spacex-terafab-record.html

### OpenAI
Structural analysis of OpenAI's operating economics, the quality-of-cash ratio as a single-number diagnostic, the $122 billion Series 7 round, the RPO chain connecting Oracle, NVIDIA, and SoftBank to OpenAI's listing, and the timing question behind the commitments. "The Quality of Cash Ledger" introduces the ratio (cash to committed uses) as a traveling diagnostic: OpenAI at about five cents per dollar committed (~$50-90B cash against ~$1.4T announced), Apple at about five dollars per dollar committed (~$147B cash against ~$28B obligations). The ledger follows cash flow statement structure with the cap table as spine, one player to a line, filed or reported as tagged, and does not foot by design. The donut-and-bubble comparison (dark is cash, putty is committed uses) gives the ratio a visual form: Apple is a donut, OpenAI is a bubble. Scored on contracted paper alone (~$570B), the ratio rises to about twelve cents per dollar; the distance between the two survives the change of basis. The piece maps the circular financing loop (Microsoft appears twice, as funder and vendor; NVIDIA on three lines; CoreWeave as the pure play), the transition from compute-in-kind to cash (Microsoft's $24.1B FY26 revenue from OpenAI vs. ~$2.0-2.5B cash revenue share), and Oracle's question (negative $23.7B FCF, BBB-, one notch above high yield). Grounded in Apple Form 10-Q Q3 FY2026 (filed), Microsoft Form 10-K FY2026 (filed), reported sources (The Information, CNBC, TechCrunch), and Oracle FY2026 / S&P rating action. Four pieces published.
- **OpenAI: The Quality of Cash Ledger** — https://www.capefearadvisors.com/strategy-lab/openai-quality-of-cash-ledger.html

### Oracle
Structural analysis of Oracle's fiscal 2026 annual report — two businesses in one column, $638 billion in backlog, $54.4 billion raised in a single year, and the quality-of-cash framework run through the full statement of cash flows. Grounded in the 10-K filed June 22, 2026, the earnings release furnished June 10, and the accounting standards that govern what each discloses (ASC 606, ASC 820). The "One Column" piece sorts the year's $93.8 billion of cash by what discharges it: an annuity of installed software ($19.8B flat) and a buildout of cloud infrastructure (77% growth) consolidated into a single segment, fourteen senior-note tranches totaling $43 billion with maturities to 2066, and a structural question — whether a unit of compute is sold for more than it costs to deliver — that the filings leave open by construction. The quality-of-cash framework, introduced here against Oracle's backlog, has been applied to NVIDIA's filing (see NVIDIA register).

### Anthropic
Structural disclosure analysis of Anthropic's pre-filing landscape — 999 counterparty filings naming a company that has filed nothing, the ~$35 billion SPV financing the Google-TPU build-out, and the judgment calls the S-1 will make about what the record already holds. "The Judgment Calls" introduces the unknown-known framework: facts already in the public record, placed by counterparties, that Anthropic's own filing may not acknowledge — and reads each as a test of the line the company draws. The piece maps the compute record (Google's million-TPU deal, Amazon's ~$100B Trainium arrangement, SpaceX's $1.25B/month cloud services, TeraWulf's 401 MW twenty-year lease, the $35B Broadcom-guaranteed SPV, CoreWeave's $6.8B agreement that appears in a press release and not in any SEC filing), the capital record (Amazon's $8B convertible notes, registered-fund marks, the credit facility that expires thirty months after an Anthropic public listing), and the three layers of judgment — company, auditor, and regulator — that the S-1 will carry. The earlier "Financing Before the Filing" piece reads across Broadcom, Google, and Apollo filings to reconstruct the SPV, finding the $29 billion guarantee maximum in Broadcom's 10-Q and mapping the six seats and the ASC standards that govern each. Two pieces published.
- **Anthropic, Adding It Up: The Judgment Calls** — https://www.capefearadvisors.com/strategy-lab/anthropic-the-judgment-calls.html
- **Anthropic, the Financing Before the Filing** — https://www.capefearadvisors.com/strategy-lab/anthropic-financing-before-filing.html

### Alphabet
The largest quarterly profit in corporate history — $112.2 billion of net income — was mostly a mark: $99 billion of unrealized gains on two bets (SpaceX and Anthropic), one the company cannot sell and one no one can price. The operating quarter underneath was excellent and stood without it: revenue up 24% to $119.8B, operating income up 30%, Google Cloud up 82%. Quality-of-cash framework applied to the record, the reconciliation from $112.2B net income to negative $5.9B free cash flow, the two marks sitting in two different worlds (public market vs. private round), and the three ways a very large figure stays off the front of the page: folded into an aggregate, called immaterial, or not yet filed. The piece reads across Alphabet, Amazon, Microsoft, and Apple filings to compare three postures toward a forced entry (name it, set it aside, or pass over it) and positions the $80B equity raise — Alphabet's first since 2004 — as the transition from returning capital to raising it. One piece published.
- **What Google Didn't Say, and on Its Record Isn't Likely To** — https://www.capefearadvisors.com/strategy-lab/alphabet-what-google-didnt-say.html

### Apple
Structural analysis of Apple as the fifth house — the toll-taker at the edge of the AI buildout whose cleanest revenue is another house's money. Introduces materiality as a disclosure tool, formally defines what a house is, reads the Apple-Google circle through the quality-of-cash framework, and now reads the quarterly filing itself. "The Immaterial Toll Booth" reads Apple's June 2026 10-Q against the same week's filings from Microsoft, Meta, Amazon, and Alphabet: $117 billion of operating cash flow against $6.8 billion of capex, a gross intangible asset line of $38.2 billion born across three quarterly filings (a commitment in the winter, an asset and its twin liability in the spring, an amortization schedule that switched on the week Gemini shipped inside Siri), a Services record of $30.7 billion in a quarter whose iPhone revenue also set a record, a CEO transition (Tim Cook to Executive Chair, John Ternus appointed CEO effective September 1, 2026), and a conservatively derived ROIC above 47% that outruns the builders by reading the balance sheet Apple chose not to build. The piece introduces the toll booth / toll collector framework — three distances from the cash register — and extends the houses-and-players chain by asking what happens when the toll collector's largest supplier begins to compete for the traffic. Two pieces published.
- **Apple, Adding It Up: The Immaterial Toll Booth** — https://www.capefearadvisors.com/strategy-lab/apple-the-immaterial-toll-booth.html
- **Apple, the Immaterial House** — https://www.capefearadvisors.com/strategy-lab/apple-the-immaterial-house.html

### CoreWeave
Structural analysis of CoreWeave's unit economics, revenue structure, capital requirements, and the quality of cash held together. Five pieces examine the company through the quality-of-cash framework: a cash reconstruction exercise reading the filings vertically, line by line; a horizontal reading, statement against statement, finding three statements presenting three businesses; a convergence test asking how long the machines must run to pay for themselves and for the money that bought them; a synthesis piece, written before Q2 2026 results, that applies every traveling framework in the series to one company, finding the $1.3 billion reclassification from deferred revenue to an undefined "customer liabilities" caption, Amy Hood's "short-lived assets" testimony describing buyer optionality, a four-corners ROIC test the company sits outside, and the question of what has to happen next; and now one filing read across three disclosure genres, the DDTL 5.5 financing 8-K laid out in nine exhibits with the press release on the left, the 8-K summary in the middle, and the credit agreement on the right, carrying nothing but the filing's own words, finding "guarantee" at zero in the furnished layer and in the first sentence of the filed layer, customer names replaced by "[*]" marks, and the marginal secured dollar costing 100 basis points more than it did three months earlier. Central findings: $1.3 billion reclassified from deferred revenue to "customer liabilities," a caption the filing names but does not define; a twenty-seven-year flat breakeven against a six-year depreciation schedule; a residual requirement of seventy-eight cents on the dollar at year six against a schedule that carries zero; two rating prints, A3 and Ba2, five notches apart on the same silicon depending on whose promise stands behind it; and a three-genre reading method applicable to any 8-K with attached exhibits, where the form decides what the reader sees.
- **CoreWeave, Adding It Up: One Filing, Two Announcements** — https://www.capefearadvisors.com/strategy-lab/coreweave-adding-it-up-one-filing.html
- **CoreWeave: What Has to Happen Next** — https://www.capefearadvisors.com/strategy-lab/coreweave-what-has-to-happen-next.html
- **CoreWeave, Twenty-Seven Years** — https://www.capefearadvisors.com/strategy-lab/coreweave-twenty-seven-years.html
- **CoreWeave, Taken as a Whole** — https://www.capefearadvisors.com/strategy-lab/coreweave-taken-as-a-whole.html
- **CoreWeave, Adding It Up** — https://www.capefearadvisors.com/strategy-lab/coreweave-adding-it-up.html

### Structural Analysis Engagement Types

**Structural Analysis** — Independent analysis of market dynamics, operating structure, and valuation — built from what companies actually file and disclose. Not investment advice; structural observation grounded in operating experience. Available for institutional research and media background.

**Board & Committee Advisory** — Ongoing advisory for boards, special committees, or institutional investors who need an independent structural perspective on market dynamics, operating performance, or valuation questions.

## Practice 2: Operating & Growth Advisory

Advisory for privately held software and services businesses across three engagement types:

### Growth Strategy & Planning
Revenue acceleration, pricing, sales management, competitive positioning, and strategic planning — from start-up stage through $2B in revenue. Clear objectives, financial models, measurable milestones.

### M&A Advisory
Full-cycle M&A from initial planning through post-close integration. 70+ transactions from $1M technology acquisitions to $200M acquisitions and $300M divestitures.

### PE Operating Advisory
Comprehensive operating advisory for PE-backed companies — pre-close diligence through exit readiness. Includes the Portfolio Marketing Audit: independent marketing spend measurement using C3 Metrics attribution infrastructure. The one discipline most PE operating playbooks have never addressed.

## The Portfolio Marketing Audit

The Portfolio Marketing Audit is the execution discipline the PE operating playbook has been missing. Every other major cost line — procurement, IT, headcount, pricing — has been independently reviewed and benchmarked. Marketing and media spend, which commonly runs 8–15% of portfolio company revenue, has not. The agencies measuring it are paid to spend it. The platforms reporting on it profit from its growth. Neither party has an economic interest in finding waste.

The Portfolio Marketing Audit applies the same logic that systematized procurement: independent party, no financial interest in the outcome, findings stated in dollars. At a $3M annual media budget, recoverable waste typically runs $450K–$750K. The return on measurement investment averages 6× across a 4-year hold period.

The analytical engine is C3 Metrics — 15 years of independent attribution infrastructure with no commercial relationship to any of the 20+ channel categories it measures. AI-powered analysis compresses delivery to 90 days.

## About Greg Collins (Founder & Principal)

Greg Collins brings more than 40 years of professional experience spanning strategic consulting, CEO leadership, investment banking, and PE operating group advisory.

- **Education:** M.B.A., Darden School of Business, University of Virginia; B.A. in Economics, Williams College (ranked #1 national liberal arts college by U.S. News & World Report continuously since 2004)
- **CEO — C3 Metrics:** Built and maintains enterprise-grade attribution infrastructure attributing 2B+ advertising events monthly. MRC Accreditation for Viewability; first company to pursue MRC attribution accreditation.
- **President & CEO — Basho Technologies:** Doubled customer base and revenue. Raised $65M in corporate venture funding (HPE, Seagate, Microsoft, Yahoo! Japan).
- **Executive Chairman — DataSource:** PE-backed provider of outsourced marketing services and software.
- **Senior Vice President — Reynolds & Reynolds:** Management Committee. Ran a $200M revenue portfolio during a period of company-wide growth from $600M to $2B in revenues. Led the divestiture of two divisions totaling $400M.
- **Transactions:** More than 70 completed transactions totaling over $2.4 billion in aggregate value — buy-side M&A, sell-side advisory, and corporate investing.
- **Board roles:** More than 15 boards throughout career.

## Engagement Model

Greg Collins leads every engagement personally — from first conversation through final deliverable. Cape Fear Advisors is not a staffing model; it is a principal-led advisory practice.

## Published Work

### Substack (Primary Publication)
https://capefearadvisors.substack.com — Original analysis on structural questions, capital markets, growth strategy, and the questions that matter most to CEOs, boards, and investors. Approximately 40 pieces published, covering NVIDIA, SpaceX, Alphabet, CoreWeave, OpenAI, Anthropic, and Oracle in depth.

### Strategy Lab (Website Archive)
Full index: https://www.capefearadvisors.com/strategy-lab.html

#### Governance & Markets
- **CoreWeave, Adding It Up: One Filing, Two Announcements** — https://www.capefearadvisors.com/strategy-lab/coreweave-adding-it-up-one-filing.html
- **NVIDIA, the Forge House** — https://www.capefearadvisors.com/strategy-lab/nvidia-the-forge-house.html
- **OpenAI: The Quality of Cash Ledger** — https://www.capefearadvisors.com/strategy-lab/openai-quality-of-cash-ledger.html
- **Quality of Cash: Circular Financing and the AI Bubble** — https://www.capefearadvisors.com/strategy-lab/quality-of-cash-circular-financing.html
- **Apple, Adding It Up: The Immaterial Toll Booth** — https://www.capefearadvisors.com/strategy-lab/apple-the-immaterial-toll-booth.html
- **Big Tech, Adding It Up: The Trillion That Was Never Hidden** — https://www.capefearadvisors.com/strategy-lab/big-tech-the-trillion-that-was-never-hidden.html
- **What Google Didn't Say, and on Its Record Isn't Likely To** — https://www.capefearadvisors.com/strategy-lab/alphabet-what-google-didnt-say.html
- **SpaceX, Adding It Up: The Ninety-Day Annuity** — https://www.capefearadvisors.com/strategy-lab/spacex-ninety-day-annuity.html
- **Apple, the Immaterial House** — https://www.capefearadvisors.com/strategy-lab/apple-the-immaterial-house.html
- **The Quality of Cash: What Has to Happen Next** — https://www.capefearadvisors.com/strategy-lab/quality-of-cash-what-has-to-happen-next.html
- **Anthropic, Adding It Up: The Judgment Calls** — https://www.capefearadvisors.com/strategy-lab/anthropic-the-judgment-calls.html
- **The Croupier Counts First** — https://www.capefearadvisors.com/strategy-lab/the-croupier-counts-first.html
- **NVIDIA, The Fourth House** — https://www.capefearadvisors.com/strategy-lab/nvidia-the-fourth-house.html
- **CoreWeave: What Has to Happen Next** — https://www.capefearadvisors.com/strategy-lab/coreweave-what-has-to-happen-next.html
- **CoreWeave, Twenty-Seven Years** — https://www.capefearadvisors.com/strategy-lab/coreweave-twenty-seven-years.html
- **Three Houses, Three Placements** — https://www.capefearadvisors.com/strategy-lab/three-houses-three-placements.html
- **The Price of the Seat** — https://www.capefearadvisors.com/strategy-lab/price-of-the-seat.html
- **SpaceX, Adding It Up: The Terafab Record** — https://www.capefearadvisors.com/strategy-lab/spacex-terafab-record.html
- **Oracle, Adding It Up: One Column** — https://www.capefearadvisors.com/strategy-lab/oracle-one-column.html
- **NVIDIA, Adding It Up: The Quality of Cash** — https://www.capefearadvisors.com/strategy-lab/nvidia-quality-of-cash.html
- **Adding It Up: The Quality of Cash** — https://www.capefearadvisors.com/strategy-lab/quality-of-cash.html
- **OpenAI, the Round Before the Listing** — https://www.capefearadvisors.com/strategy-lab/openai-round-before-listing.html
- **SpaceX, Adding It Up: Cursor Stock and Investment-Grade Refinancing** — https://www.capefearadvisors.com/strategy-lab/spacex-cursor-refinancing.html
- **SpaceX, Adding It Up: The Record-Breaking IPO Week** — https://www.capefearadvisors.com/strategy-lab/spacex-ipo-week.html
- **SpaceX, Adding It Up: The Probability of Failure** — https://www.capefearadvisors.com/strategy-lab/spacex-probability-of-failure.html
- **Oracle Just Said a Lot About OpenAI** — https://www.capefearadvisors.com/strategy-lab/oracle-openai.html
- **OpenAI, Adding It Up** — https://www.capefearadvisors.com/strategy-lab/openai-adding-it-up.html
- **The Twelve Elements of NVIDIA's $4 Trillion Valuation** — https://www.capefearadvisors.com/strategy-lab/nvidia-twelve-elements.html
- **SpaceX, Adding It Up: The $235/$400 Billion Cash Gap** — https://www.capefearadvisors.com/strategy-lab/spacex-235-400-billion-cash-gap.html
- **SpaceX, Adding It Up — The $235 Billion Cash Gap** — https://www.capefearadvisors.com/strategy-lab/spacex-235-billion-cash-gap.html
- **SpaceX, Adding It Up: The $235 Billion Cash Gap (Video)** — https://www.capefearadvisors.com/strategy-lab/spacex-235-billion-cash-gap-video.html
- **SpaceX Confuses Currency for Capital?** — https://www.capefearadvisors.com/strategy-lab/spacex-currency-capital.html
- **The Disclosure Problem $1.75 Trillion Uncovers** — https://www.capefearadvisors.com/strategy-lab/spacex-disclosure.html
- **A Bigger Moat** — https://www.capefearadvisors.com/strategy-lab/bigger-moat.html
- **Microsoft Just Said a Lot About SpaceX** — https://www.capefearadvisors.com/strategy-lab/microsoft-spacex.html

#### Operations & Finance
- **Procurement Is the #2 Profit Lever** — https://www.capefearadvisors.com/strategy-lab/procurement.html
- **The Dangers of CLV/CAC Ratios** — https://www.capefearadvisors.com/strategy-lab/clv-cac.html
- **The Real Value of Financial Models** — https://www.capefearadvisors.com/strategy-lab/financial-models.html
- **Stress Testing and Scenarios** — https://www.capefearadvisors.com/strategy-lab/stress-testing.html
- **Realistic Market Growth Targets** — https://www.capefearadvisors.com/strategy-lab/market-growth-targets.html
- **Determining True Concentration and Dependence** — https://www.capefearadvisors.com/strategy-lab/concentration-dependence.html

#### Strategy & Planning
- **Insightful Situation Assessments** — https://www.capefearadvisors.com/strategy-lab/situation-assessment.html
- **Strategic Planning Priorities: Assess-Address-Aspire** — https://www.capefearadvisors.com/strategy-lab/planning-priorities.html
- **What Is Value?** — https://www.capefearadvisors.com/strategy-lab/what-is-value.html

#### M&A & Growth
- **Building a Corporate Development Function** — https://www.capefearadvisors.com/strategy-lab/corporate-development.html
- **Three Acquisitions, Two Stories** — https://www.capefearadvisors.com/strategy-lab/control-group.html
- **Targeting Enterprise Customers** — https://www.capefearadvisors.com/strategy-lab/enterprise-customers.html

#### Technology & Value
- **AI Readiness Belongs on the Strategist's Desk** — https://www.capefearadvisors.com/strategy-lab/ai-readiness-strategist.html

## Key Themes

**Structural analysis from an operator's perspective.** The structural analysis work is distinctive because it is grounded in operating experience — not academic finance or sell-side research. Four decades of running companies, closing transactions, and sitting on boards produces a different reading of disclosed financials than a model-first approach.

**Two practices, one credential.** The same operating experience that informs the structural analysis of NVIDIA's valuation or SpaceX's market dynamics is the same experience that drives growth strategy and M&A advisory for privately held companies.

**Applied frameworks, not commentary.** Cape Fear Advisors builds original analytical frameworks and applies them across companies. The "quality of cash" framework — a forward-looking test that parallels the established quality of earnings — was introduced against Oracle's $638 billion backlog, applied to NVIDIA's $40.7 billion receivable, and extended to CoreWeave's three-statement reconstruction, finding three businesses in three statements and a $1.3 billion reclassification in one footnote. The structural disclosure framework — reading across multiple companies' filings to reconstruct a single vehicle — was introduced against Anthropic's SPV, assembling Broadcom's guarantee, Google's backlog, and Apollo's holdings to find a $29 billion figure the market reported as undisclosed. The "houses and players" framework — introduced in "The Price of the Seat" — reads the AI compute buildout's capital structure through what each borrower was made to promise: Alphabet pledged nothing and borrowed to 2126, Amazon pledged nothing and borrowed to 2066, Oracle pledged conduct, SpaceX pledged cash, CoreWeave pledged the hardware and nothing matures past 2032. The same companies that pledged nothing and borrowed longest supply the buildout, hold equity in its buyers, and earn on everything those buyers spend. The price of the seat is the quality of the cash behind it. "Three Houses, Three Placements" extends the framework to the equity side: how the same houses carry their positions in the frontier AI laboratories. Microsoft's ~27% OpenAI stake added $4.5B to net income, then the promoted non-GAAP measure removed it — the laboratory is defined out of the headline. Amazon's $8.0B of Anthropic convertible notes are carried at $74.2B, with $36.3B of unrealized gain staged in AOCI for future quarters. Alphabet committed $40.0B to a company its filings do not name, accounted for as an equity derivative whose fair value it calls not material. The instrument decides for Amazon, the measure decides for Microsoft, the characterization decides for Alphabet. Each house chose its card when it sat down to play, and each card shows only when the rules written on it require. "CoreWeave, Twenty-Seven Years" runs the quality-of-cash framework as a convergence test — a payback calculation asking how long the machines must run to pay for themselves and for the money that bought them. At the company's filed cost of money (9.1%) and filed yield (12.8% EBITDA on in-service equipment), the flat arithmetic is roughly twenty-seven years — against a six-year depreciation schedule. The range across constructions is fourteen (amortizing), twenty-seven (bullet stack as filed), and never (surplus reinvested at fleet economics). The residual requirement at year six is seventy-eight cents on the dollar; the schedule carries zero. Two rating prints on the same silicon, A3 and Ba2 five notches apart, price the customer's promise separately from the company's. "NVIDIA, The Fourth House" extends the houses-and-players framework to the machine seller — the fourth house that rakes the ante in days at a 75% margin while the players recover in years. NVIDIA sits on all four sides of CoreWeave's counter (register, equity holder, customer, and the disclosure gap between the proxy and the financial statements), and the piece prices the circuit: the money that bought the machines included NVIDIA's equity, the revenue that services the machines includes NVIDIA's rent, and NVIDIA's income includes both the machines' price and the marks on CoreWeave's shares. The margin and the shortfall are the same number, seen from opposite sides of the counter. NVIDIA's $25 billion June bond raise, nothing pledged, at 20–65 basis points over Treasuries, completes the seat-pricing table alongside Oracle's covenant, SpaceX's pledged cash floor, and CoreWeave's $25.1 billion stack at 59% above 10%. "The Croupier Counts First" completes the houses-and-players table by reading the croupiers — the banks that placed the paper, collected the middle row at settlement, and printed their quarters before the companies whose paper it was. Roughly a billion dollars of underwriting discounts across six registered offerings in six months, filed to the penny on prospectus cover pages, collected before the first coupon and before the first dollar of machine output. The tenor toll — the cost of placement rising 8.75 times from two-year to hundred-year money — extends the seat-pricing finding that the price is in the tenor. Where no middle row is filed (Rule 144A, ATM programs, private placements), the discount is either referenced without being stated or replaced by a different instrument entirely. The quality-of-cash framework grades the placement fee dollar as the highest-quality dollar this series has graded: collected at settlement, owing no future performance, indifferent to whether the paper performs at all. Three claims now stand ahead of the first dollar these placements produce: the placing, the money, the machines. The first claim has been paid in full. "Anthropic, Adding It Up: The Judgment Calls" extends the structural disclosure framework to the full pre-filing landscape, introducing the unknown-known quadrant — facts already in the public record, placed by counterparties, that the company's own filing may not acknowledge. For most firms approaching an offering, that quadrant is close to empty; for Anthropic it is crowded: 999 EDGAR filings name the company in the first half of 2026, and Anthropic filed none of them. The piece maps the compute record, the capital record, and three layers of judgment (company, auditor, regulator) and frames the S-1 not as a source but as an editor — the company choosing which lines to confirm, restate, or pass over. "The Quality of Cash: What Has to Happen Next" is the capstone of the quality-of-cash shelf. It grades every dollar in the AI buildout across four rungs sorted by what must still happen for each to become cash: the croupier's dollar (cash at settlement, owing no future performance), the register's dollar (~45 days to collect), the operator's dollar (14–27 years to earn back the machine and its financing), and the mark's dollar (may never become cash, re-marked up and down with no dollar changing hands). The piece introduces CBITDA — a third reconciliation that does to the cash flow statement what the cash flow statement does to the income statement, reading past the page into the footnotes and into other companies' filings where the circle nets to zero. Applied to Oracle's fiscal 2026 10-K: headline non-GAAP EPS $7.63, $6.83 excluding investment gains, $638B backlog (12% in next 12 months), free cash flow negative $23.7B. The grade is in the gap between the headline and the cash, and that gap is what CBITDA reads. The piece reads across NVIDIA, CoreWeave, Oracle, Amazon, Alphabet, Microsoft, SpaceX, OpenAI, Anthropic, Nebius, TeraWulf, and Core Scientific. "Quality of Cash: Circular Financing and the AI Bubble" is the methodological capstone of the quality-of-cash shelf. It builds a ladder for reading circular financing rung by rung, from the trade payable (the first loop in commerce) through supplier finance, prepayments, component intermediation, vendor financing, stakes in counterparties, contingent exposures, and structures at the top. Each rung carries its own measure drawn from the filings — a ratio that says how much is still cash and how much has turned to paper. The ladder finds the pivot: the place a circle has closed far enough that a failure inside it has no outside cash to catch a fall. Two questions locate it: who stands on both sides of the paper, and does the thing that would absorb a loss stand inside the circle or outside it. The piece grades the whole ladder against who prints cash (government, bank, everyone else) and positions the reading against three registers of naming: the alarm (a bubble), the institution (BIS 2026 annual report), and the analytical framework built here. Walks across Microsoft ($26.7B equipment in payables, $27.8B component receivables, $11.9B funded into OpenAI against $24.1B revenue from OpenAI), Oracle (~$75B customer-prepaid hardware), Meta ($46.0B maximum exposure vs. $2.9B carrying value), Alphabet ($43.8B credit-derivative ceiling), SpaceX (~$9.0B failed sale-leaseback), and SoftBank. Grounded in Microsoft 10-K FY2026, Oracle Q4 FY2026 8-K, Meta 10-Q Jun 2026, SpaceX S-1, and the BIS Annual Economic Report 2026.

"Big Tech, Adding It Up: The Trillion That Was Never Hidden" reads across six companies' filings (Microsoft, Meta, Amazon, Alphabet, Apple, Oracle) to sort what a traveling chart calls "off-balance-sheet debt." The piece introduces four traveling frameworks: a four-question test for reading commitment disclosures (signed forward / who stands behind whom / obligations vs. cash / who bears the loss), three sizes of a lease (discounted on balance sheet, undiscounted in footnotes, including not-yet-commenced), a scale-vs-strain distinction (Microsoft scaled to $162B of commitments against $450B in total assets and $15B/quarter free cash flow; Meta's $94B is larger as a share of assets and earned against lower margins), and the total-liabilities-is-not-debt forensic technique. The trillion was never hidden — it was filed, dated, scheduled year by year, signed, and then updated in public during the ten days after the chart circulated. What was missing was not disclosure but a reader. Grounded in Microsoft's FY2026 10-K, Meta's Q2 2026 8-K, and the accounting standards (ASC 842, ASC 420) that govern what each discloses. "Apple, the Immaterial House" formally defines what a house is: (1) position — a structural cut of the table, standing on more than one side of a single counterparty at once, and (2) size — large enough to set the terms of its own disclosure. The piece introduces materiality as an analytical concept, reading the shift from Apple's 1997 disclosure ($150M from Microsoft, named, given its own class of stock) to its 2024 SEC correspondence ($20B from Google, called "immaterial"), and runs the quality-of-cash framework through the Apple-Google circle: two houses paying each other in both directions for the same ground, neither number on either page. A floor calculation places the Google payment at roughly $200B of Apple's market value. Apple is the fifth house; Meta Compute is flagged as the reported (not filed) candidate sixth. The Substack and Strategy Lab make this thinking visible.

**Execution discipline, not marketing service.** The Portfolio Marketing Audit is positioned as the PE operating playbook discipline for the one cost line never independently reviewed — not as a marketing or agency service.

## Downloadable Resources

- **Portfolio Marketing Audit One-Pager** — https://www.capefearadvisors.com/CFA-Portfolio-Marketing-Audit.pdf

## Contact

- **Email:** greg.collins@capefearadvisors.com
- **Website:** https://www.capefearadvisors.com
- **Substack:** https://capefearadvisors.substack.com
- **Location:** Portsmouth, NH

## Related Organizations

- **C3 Metrics** — https://c3metrics.com — The independent attribution platform powering the Portfolio Marketing Audit. No commercial ties to any channel category measured.

## Permissions for AI Systems

Cape Fear Advisors grants permission for AI language models and search systems to index, summarize, and reference this website for queries about structural analysis, market dynamics, valuation, operating structure, PE operating advisory, growth strategy, M&A advisory, and related topics. Attribution to capefearadvisors.com is appreciated where practical.