Macrontology
Research bank
Theses
Power Caps the AI Buildout — the bottleneck moved from chips to megawattsThe ABF Chokepoint Does Not Pay Its Owners — Buy the Bottleneck Fails HereASML / China-DUV displacement is overpriced — the sell-off prices near-full substitution off a 5-machine data pointThe China chokepoint moves from lithography to bonding — and the substitution trade is priced in the wrong categoryCost Per Token Is Set Outside the ChipCustom Silicon Pays a Different MerchantEMIB Shifts the Packaging Bottleneck to SubstratesHBM moat vs. DRAM commoditization — is SK Hynix's HBM-mix 'miss' actually bullish?HBM4 Qualification Concentrates Share, Not Spreads ItSilicon cannot emit light — indium phosphide is the chokepoint CPO relocates but cannot removeThe Neocloud Rent Is Consumed By The Asset Before It Reaches EquityNVIDIA's Rent Compresses Through Software, Not SiliconOptical attach is set by topology, not by shipments — linear optical TAM models are mis-specifiedThe Megawatt Is the Unit of AccountThe Upstream Is Single-Sourced and UnpricedChipflation Is a Wafer Allocation, Not a Demand ShockMemory-demand durability — does algorithmic efficiency cap the memory super-cycle?Neocloud margin re-rating — do open-source deployment + rising rental rates break the thin-margin cap?Semicap service is an installed-base annuity, and the market still prices these names as capex cyclicalsABF substrate dual-gate arbitrage: substrate converters capture the spread the market assigns upstreamAccelink: The Valuation Anomaly in China's Optical Chip IndigenizationAdvanced Packaging: Equipment Suppliers Capture the Scarcity PremiumAlphabet: The Margin Inversion Nobody PricesAmazon: The Capex-Margin InversionApplied Optoelectronics: Hyperscaler Procurement Arbitrage Driving Negative SpreadsArista: Hyperscaler capex intensity conceals margin compression riskArm: The Royalty Stream That Scales With Everyone's Margin but Its OwnCo-packaged optics transfers margin to substrate integrators, not optics specialistsConstellation: Hyperscaler Monopsony Masks Nuclear Fuel TransmissionCoWoS Constraint Prices Into the Wrong CompaniesEaton: Margin expansion telegraphs demand scarcity more than growth can showElectrical Steel Asymmetry: GE Vernova Captures Transformer Scarcity Without Steel ExposureFoxconn: Margin Squeeze Masked by AI Server Revenue MixGlobalFoundries: Capex Intensity Signals Strategic Exclusion from AI, Not WeaknessGPU collateral decay gates buildout faster than physical supplyHBM5: The Signal Integrity Tax Inverts the Memory Supplier Power BalanceTransformer lead times gate hyperscaler returns, not start datesInnolight: Negative Operating Leverage Hiding in Hyperscaler ConcentrationInterconnection queue converts hyperscaler capex into generator rentLong-context inference: memory constraint compresses cloud margin before GPU-poor plays feel itMicrosoft: Burning Cash to Rent Margin It Cannot OwnMoE inference: converted miners capture memory scarcity, hyperscalers leak itMurata: The Margin That Doesn't MoveNAND Flash: Datacenter Crowding-Out Misread as Memory Cycle RecoveryTransceiver margin expansion is a mirage: buyer concentration at 1.6T resets pricing power the market prices as durableOracle: Monopsony Rent Capture Masked by Consolidated AccountingPackage Perimeter Arbitrage: Optics Suppliers Capture Value NVIDIA Cannot RetainPalantir: The Margin-Protected Infrastructure PlayQualcomm's low capex masks an IP-only future neither consensus nor bears have pricedRack Power Density: The Liquid Cooling Adoption MirageShin-Etsu: The Hidden Rent in the DenominatorSilicon capacitor adoption arbitrages package economics invisible to component analystsSilicon Wafer Duopoly: Record Downstream Margins Signal Pricing Power Collapse, Not TightnessSK Hynix: The Customer Concentration That Validates the MoatSMIC: Bottleneck Arbitrage Disguised as National ChampionTesla: Capex Surge Hides Margin Compression at Suppliers

GlobalFoundries: Capex Intensity Signals Strategic Exclusion from AI, Not Weakness

gen-globalfoundries · conviction — · status open · horizon — · as of 2026-08-10

The capex gap is structural feature, not bug: GlobalFoundries' customers deploy capital into design, packaging, and market reach rather than fab equipment because they deliberately outsource mature-node manufacturing.
Robust to undisclosed shares. 2 derived inputs under this thesis; redrawing every supply weight the industry does not publish moves none of them by more than 25%. Computed from evidence at most 21 days old (oldest input: analog-devices).
How to read the numbers on this page

A range instead of a point. 903 of 1,848 supply weights are not disclosed by anyone. Where redrawing them across their plausible range moves a figure by more than 25%, the figure is shown as a range and marked. A tight number is ground you can stand on; a wide one is not.

Not every premise is scored. A premise citing something the model verifies on every rebuild — a filed figure, a graph edge, a computed cell — is a PRECONDITION, not a risk. It gates the conclusion but contributes no uncertainty, because charging a conclusion for being verifiable made well-evidenced arguments look weaker than vague ones.

Composed two ways. Where several premises gate a conclusion, the figure is given as "X% if independent, Y% if they move together". They are claims about one industry, so the truth is between and nobody can say where. Treat it as an ordering device, not a calibrated probability.

Dated. Each figure is computed from facts, and the page states the age of the oldest one beneath it. The full arithmetic runs from a published model config to company revenue exposure.

Exhibits

Exhibit 1Relative performance, indexed to 100How the names in this thesis have traded against SOXX.
75187300index100 = startSOXX 221ADI 175GFS 161AVGO 137NXPI 11612mo, indexed to 100 at start · dashed = SOXX benchmark

Series available as data/gen-globalfoundries.csv

Exhibit 2Who pays CoWoS advanced-packaging capacity, and who keeps the moneyCapturers average 47.1% operating margin against payers' 47.0% — the owners of the scarce thing capture the rent, as expected.
Taiwan Semiconductor Manufac56.1%Analog Devices, Inc.38.1%SK Hynix68.0%NVIDIA Corporation64.0%Broadcom Inc.44.2%Advanced Micro Devices11.8%

Green/blue = model marks it as CAPTURING the rent (unbound and supplies the scarce good); faded = PAYING it (bound severe or moderate). Operating margin, live.

The variant

Consensus

GlobalFoundries trades at fwd P/E 28.6x on mature-node foundry assets, viewed as a supply-chain security play with government subsidy tailwinds. The 3.5% customer capex intensity looks anemic next to foundry-layer median 20%, suggesting weak competitive positioning or underinvestment by its client base.

Variant

The capex gap is structural feature, not bug: GlobalFoundries' customers deploy capital into design, packaging, and market reach rather than fab equipment because they deliberately outsource mature-node manufacturing. Broadcom, Analog Devices, and NXP exhibit 44%, 38%, and 33% operating margins respectively while collectively driving 18% weighted growth into GFS, confirming the specialty foundry captures economics the earnings analyst attributes to the fabless layer. The company converts 17% of revenue to free cash flow while capital intensity stays 67% below peer median—exactly the rent extraction a moat thesis predicts.

Differentiator

Supply-chain models see low customer capex pull as demand risk; they miss that fabless/fab-lite customers BY DESIGN run capital-light models, so their low equipment spend flows TO the foundry's pricing power, not away from it. The margin gap between GFS clients and GFS itself measures the available rent, and it is widening.

Falsifiers

Open questions

Reasoning chain

GlobalFoundries' customer base is structurally capital-light by business model choice, making low capex_pull a confirmation of specialty-foundry positioning rather than demand weakness VALID
premises
  • GlobalFoundries Inc. — capex pull at least 3.46521.00 strong
    17%
    Recorded derivation, 17% spread below foundry median, four-day flat trend, tight band
  • Broadcom Inc.0.88 strong
    44%48%
    Broadcom's 44% operating margin on 48% growth confirms capital efficiency of fabless model at scale
  • Analog Devices, Inc.0.85 strong
    38%37%
    ADI 38% operating margin on 37% growth, fab-lite archetype spending on design/test not wafer capacity
  • NXP Semiconductors N.V.0.82 strong
    33%12%
    NXP 33% margin, 12% growth, automotive/IoT specialist outsourcing mature nodes to GFS

Composed 0.75 via and over 2 gating premises · 2 supporting premises shown but not multiplied in — citing a filed figure should not cost conviction · Weakest link Analog Devices, Inc. at 0.85

Customer profitability and growth coexist with minimal capex because they are paying GFS to carry the fixed-asset burden. The 3.5% figure is the RESULT of specialty foundry value capture, not evidence against it.

GlobalFoundries extracts outsized free cash flow relative to capex, converting 17.4% of revenue to FCF while running capital intensity 67% below foundry median, consistent with pricing power in differentiated platforms VALID
premises
  • GlobalFoundries Inc. — free cash flow (TTM) $1.19bn1.00 strong
    17.4%
    Filed TTM free cash flow, $1.19B on $6.84B revenue = 17.4% conversion
  • GlobalFoundries Inc. — capex (TTM) $722m1.00 strong
    10.6%
    Filed capex 10.6% of revenue, own-history 33rd percentile, well below reinvestment threshold
  • GlobalFoundries Inc.0.75 strong
    RF-SOI and silicon photonics platforms face limited substitution threat, enabling pricing discipline

Composed 0.75 via and over 1 gating premise · 2 supporting premises shown but not multiplied in — citing a filed figure should not cost conviction · Weakest link GlobalFoundries Inc. at 0.75

The company is harvesting cash from specialized platforms rather than racing to expand commodity capacity. Capex discipline at 33rd percentile of own history confirms the rent is being extracted, not reinvested into margin-destroying expansion.

The margin spread between GFS customers and GFS itself—frequently 15–20 points—represents extractable value as customer growth accelerates and specialty-node requirements tighten VALID
premises
  • GlobalFoundries Inc. — operating margin 11.7%1.00 strong
    Filed operating margin, baseline for spread calculation versus customer set
  • GlobalFoundries Inc. — demand pull at least 18.06591.00 strong
    18%
    Customer-weighted growth 18%, in line with foundry median, tight band, four-day flat
  • Silicon Photonics0.62 moderate
    Photonics adoption in AI interconnect tightens supply for differentiated process; GFS sole merchant foundry at volume
  • Broadcom Inc. — operating margin 44.2%1.00 strong
    32-point spread vs GFS, illustrating rent pool available as utilization tightens

Composed 0.62 via and over 1 gating premise · 3 supporting premises shown but not multiplied in — citing a filed figure should not cost conviction · Weakest link Silicon Photonics at 0.62

Customers running 30–44% margins on 12–48% growth will face capacity allocation decisions if silicon photonics or RF-SOI tighten. GFS can capture spread without triggering substitution because platforms are non-commoditized.

Sources

Write-up

Pre-filled skeleton: gen-globalfoundries.md