Supply Chain Analysis — Eaton Corporation plc (ETN)

§ 01Executive View

Eaton's supply chain is structurally sound at the tier-1 level — broadly multi-sourced on copper, aluminum, and castings — but faces two binding chokepoints that are genuinely underpriced: (1) grain-oriented electrical steel (GOES) for power transformers, where US domestic supply is effectively a monopoly (Cleveland-Cliffs/AK Steel), and where the global market is running lead times of 128-144 weeks for power transformers as of Q2 2025; and (2) its own manufacturing capacity, where aggressive multi-site expansion is converting backlog to revenue 12-30 months after order, creating a calendar mismatch between a $19.6B backlog (Q4 2025) and the capacity to ship it. The pass-through story is the strongest in the cohort — Eaton publishes a Commodity Price Index, is explicitly committed to dollar-for-dollar tariff recovery, and has demonstrated sustained pricing discipline since 2021. The core supply-chain risk is not that ETN runs out of materials; it is that the transformer and switchgear capacity additions scheduled to come online in 2H 2026-2027 arrive later than the backlog implies, creating a 2026-2027 margin trough from ramp costs before the 2027-2028 revenue inflection that the synthesis calls "Schneider's 2028-2030 real impact."

§ 02Chain Map: Raw Material to Customer

§ 03Input Map

Tier 1

Tier 2 chokepoints

The hidden dependencies are where ETN's input-side picture is more fragile than the tier-1 table suggests:

1. GOES — the transformer manufacturing chokepoint (cohort's foundry analog) GOES production globally is concentrated in 6-7 producers. The US domestic market has a single producer (Cleveland-Cliffs / AK Steel, which acquired all domestic GOES production in 2018). US antidumping and countervailing duties on GOES from China, Germany, Japan, Korea, and Czech Republic have historically constrained the import option; duty rates range from 3-30%+ depending on origin. The Cleveland-Cliffs Weirton WV plant expansion adds capacity, but even with it, US GOES supply satisfies only 50-60% of domestic transformer demand at 2025 demand levels. The resulting import dependency routes through Korea and Japan — both viable geopolitically but exposed to trans-Pacific shipping disruptions and currency movements. A transformer maker cannot switch GOES grades quickly: core-loss specifications require pre-qualification on each GOES lot, adding 3-6 months qualification lag per new source.

Consequence for ETN: Transformer lead times of 12-24 months are the direct output of this bottleneck. GOES availability plus winding-labor capacity are the two rate-limiters for ETN's new South Carolina transformer facility (announced 2025) and its other transformer capacity additions. The $19.6B backlog is real — but its conversion to 2026 revenue is gated by steel availability and winding capacity, not by customer demand.

2. Skilled transformer winding / MV switchgear assembly labor Transformer winding is a specialized craft skill that takes 2-3 years to train. The US transformer workforce contracted severely during the 2010s offshoring era. ETN and peers (ABB, Hitachi Energy, GE Vernova) are all competing for the same limited pool of experienced winders and switchgear assemblers. ETN's new South Carolina site and expanded Mexico and Dominican Republic facilities face the same constraint: the building can be built faster than the trained workforce can be hired and certified. This is the labor analog of the silicon fab qualification cycle — not a raw-material bottleneck but a human-capital one that is equally rate-limiting.

3. SiC MOSFET availability for UPS / drives Eaton's 9395XC UPS uses SiC-based topology. SiC supply has been tight post-2022 due to substrate (SiC boule) constraints at Wolfspeed and competition from EV OBC applications. The SiC market is entering oversupply in 2025-2026 as Wolfspeed, Onsemi (Bucheon), STM (Catania), and Chinese producers (TanKeBlue, SICC) add capacity. For ETN, this is transitioning from a supply constraint to a price tailwind — SiC ASPs are declining, improving UPS margin structure from the semiconductor-input side. However, qualification cycles for new SiC vendors are 12-18 months, so ETN cannot immediately capture the cost benefit of the Chinese SiC price collapse.

4. Electrical copper tariff pass-through lag The April 6, 2026 Section 232 expansion added copper to the tariff regime. Electrical-grade copper in rod and bar form used for busbar and winding applications is now subject to a 25% base tariff (15% transitional rate for grid equipment per the Apr 2026 proclamation). Eaton publishes a Commodity Price Index (CPI) and has contractual price-adjustment clauses in most project contracts — but fixed-price legacy backlog orders (booked at 2022-2023 pricing before the tariff) have no pass-through mechanism. ETN has guided to recover tariff costs "on a dollar-for-dollar basis over the year" — the lag is the risk, not the permanent burden.

5. Li-ion battery cell supply chain (China concentration) UPS systems are shipped with batteries. Lithium-ion battery cells for large UPS are primarily produced by CATL, Samsung SDI, and LG Energy; CATL's Chinese origin makes it subject to Section 301 tariffs. ETN can dual-source to Korean cells (Samsung SDI, LG), but Korean cells carry a premium. In the context of a $27B revenue base, battery cell tariffs are a manageable ~50-100 bps margin impact — real but not structural.

§ 04Risk Scoring

Synthesis. ETN's supply-chain risk profile is dominated by the self-inflicted capacity constraint (own manufacturing ramp) and the GOES chokepoint, not by classical single-source commodity risk. The pass-through posture is the strongest among the cohort's infrastructure names — ETN has tools (CPI, contractual adjustors, pricing discipline) that NVTS and even VRT lack. The structural input-side read is: medium-resilient on raw materials, tight on GOES and capacity, and exposed to a 12-18 month revenue-conversion lag that the backlog growth disguises. The calendar mismatch risk (synthesis Open Question §2, Schneider "2028-2030 real impact") is real and is the central question for the 2026-2027 investment horizon.

§ 05Manufacturing Footprint

Eaton operates approximately 193 manufacturing facilities in 34 countries (per 2024 10-K, Item 2). Key geographic concentrations:

Geographic concentration summary vs. cohort:

• Taiwan exposure: near-zero at the manufacturing level (ETN manufactures hardware, not chips; its chip buys from IFX/TI/ON are indirect Taiwan-tail via those companies' TSMC dependencies)
• Mexico exposure: significant (~10-15% cost base), mitigated by USMCA for qualifying content, but tariff risk is active and not fully resolved
• China exposure: deliberately declining; residual 5-10% mostly serves China domestic market
• Versus TXN (the cohort's Taiwan-tail hedge): ETN has even less Taiwan manufacturing exposure but more Mexico and China tariff exposure than TXN's predominantly US/Malaysia/Japan/Philippines network

§ 06Pass-Through Power

Eaton's pass-through posture is the strongest in this cohort's electrical/infrastructure segment and is one of the company's defining competitive characteristics.

Evidence:

1. Commodity Price Index (CPI) mechanism. Eaton publishes a formal CPI and has embedded cost-escalation clauses in most project contracts. This is standard practice for large industrial equipment but ETN executes it more transparently than peers.
2. Tariff recovery commitment. Management explicitly guided in Q4 2025 that it intends to recover Section 232 and other tariff costs "on a dollar-for-dollar basis over the course of the year." The April 2026 reduction to 15% transitional rate for grid equipment provides meaningful relief from the prior 25-50% exposure.
3. Margin trajectory under input pressure. Electrical Americas operating margin was 29.8% in Q4 2025 despite 180 bps of capacity-ramp drag — a performance that implies underlying (tariff- and ramp-adjusted) margin well above 31%, already at or above management's 30%+ 2026 target. This is structural pricing power, not cyclical.
4. Backlog re-pricing dynamics. Order duration has shortened from multi-year advance bookings to 12-18 month delivery windows — meaning new orders in 2025-2026 are priced at current copper/steel/GOES levels. The risk is legacy backlog booked in 2022-2024 at pre-inflation pricing; management has not quantified what fraction of the $19.6B backlog carries fixed pricing vs. adjustable pricing. This is the most significant undisclosed input-side risk.
5. Spec lock-in and switching costs. Medium-voltage switchgear and power transformers for large data centers are engineered-to-order products. Once ETN designs a facility's power distribution architecture, replacement requires re-engineering by a new vendor over 18-36 months. This gives ETN real pricing leverage at renewal. The switchgear market has concentrated to ETN, Siemens, Schneider, and ABB/Hitachi — an oligopoly that limits price erosion.

Score: 2 — strong pass-through, with legacy-backlog repricing as the binding uncertainty.

§ 07Backlog Conversion Analysis

The $19.6B backlog (Q4 2025) is real but not uniformly convertible.

The conversion bottleneck: Management described 24 capacity-expansion projects. Approximately half were complete by mid-2025; the remaining half are "largely done" by 1H 2026, with ramp costs persisting as a 130 bps headwind through 2026. The implication is that significant new capacity is not shipping at full rate until 2H 2026 — meaning a non-trivial portion of the $13.2B Electrical Americas backlog converts to revenue in 2027-2028, not 2026.

Legacy-backlog margin dilution risk: Orders booked in 2022-2024 at then-current copper and steel prices are being delivered in 2026-2027 at current (higher) input costs. The CPI mechanism covers most of this, but the fraction of fixed-price contracts is unknown. Even a 15-20% fixed-price share of the legacy backlog could create a $300-400M revenue base delivering at below-current-cost margins in 2026-2027.

§ 08Stress Scenarios

Scenario 1: GOES supply crunch extends — Cleveland-Cliffs Weirton ramp slips 12 months

Probability: mid, 25-35%. Cleveland-Cliffs announced the Weirton facility would add 30-40% more GOES tonnage from 2H 2025 onward. Industrial project timelines at Cleveland-Cliffs have a history of slippage (the facility was originally a steel mill being repurposed). If the Weirton ramp is delayed 12 months, US GOES supply tightens further into 2026-2027.

Mechanism. ETN's transformer factories (including the new South Carolina site) are constrained by GOES availability and core-winding capacity. A GOES supply shortfall forces ETN to either (a) import at cost-and-tariff penalty, (b) prioritize higher-margin power transformers over distribution transformers, or (c) slip transformer delivery commitments, damaging customer relationships in a market where ETN is trying to take share.

Financial impact. Transformer segment revenue 5-10% below plan in 2026; margin compressed 100-150 bps on mix shift (distribution transformers sold at lower margin than power transformers). The backlog number doesn't change — the conversion timeline does. 2026 Electrical Americas segment margin tracking closer to 28-29% vs the guided 30%+.

Response options. Accelerate imports from Posco / Nippon Steel at tariff cost; sign long-term supply agreements with Korean producers; increase share of liquid-filled transformer designs that use less GOES per kVA than dry-type.

Scenario 2: USMCA renegotiation imposes 25% tariff on Mexico-manufactured electrical equipment

Probability: low-mid, 15-25% over 24 months. USMCA is up for a formal review in 2026. If Eaton's Mexico-assembled switchgear fails USMCA rules-of-origin because of non-USMCA steel or aluminum content, it could face IEEPA tariffs of 25% on entry to the US.

Mechanism. ETN assembles switchgear and circuit breakers in Juarez and Queretaro, Mexico. Under USMCA, goods with sufficient North American content enter duty-free. The risk is twofold: (a) steel inputs in Mexico may not satisfy USMCA steel melted-and-poured requirements if ETN sources from Asian steel distributors in Mexico; (b) the political negotiation could rewrite rules-of-origin percentages for electrical equipment as leverage in broader USMCA renegotiation.

Financial impact. Mexico operations estimated at ~10-15% of Electrical Americas manufacturing cost base; a 25% tariff on Mexico-origin goods would be a ~200-350 bps gross-margin headwind on the affected product lines before any recovery actions. ETN has an extensive US domestic footprint that can absorb some volume, but full re-rouring would take 12-18 months.

Response options. Accelerate US domestic assembly capacity (ETN already investing); qualify Mexican content to USMCA steel-melted-and-poured; lobby as "critical infrastructure manufacturer" for tariff exclusion (precedent set by the April 2026 15% grid-equipment transitional rate).

Scenario 3: Calendar mismatch materializes — 2026-2027 revenue conversion lags consensus

Probability: mid-high, 40-55%. This is the synthesis Open Question §2 / Schneider "2028-2030 real impact" thesis made operational for ETN specifically. If the 24 capacity-expansion projects experience even a 6-month average delivery slip, and if legacy backlog carries 15-20% fixed-price exposure, ETN's 2026-2027 revenue ramps below the $19.6B backlog-implied pace and margins are compressed by both ramp costs (known: 130 bps headwind) and legacy-backlog pricing drag (unknown quantum).

Mechanism. Markets are pricing ETN for ~7-9% organic growth in 2026 (company guidance). If new capacity (transformers, switchgear) does not ship on schedule, the 7-9% range becomes 5-6%, and the $13.13-$13.50 adjusted EPS guidance range is the top, not the midpoint.

Financial impact. 5-7% miss to consensus revenue in 2026; EPS $12.00-$12.50 vs guided $13.13-$13.50; multiple compression as the "execution risk" narrative gains traction. The backlog is not impaired — the delivery schedule is.

Response options. Accelerate third-party contract manufacturing for certain switchgear; deploy expedited freight to compress lead times (adds cost but demonstrates delivery); communicate the capacity-ramp timeline with more granularity to close the gap between backlog narrative and investor expectation.

§ 09Bull Points — Supply-Chain Dimension

• GOES chokepoint benefits ETN vs. new entrants more than it harms ETN vs. incumbents. The same GOES scarcity that limits ETN's output equally limits ABB/Hitachi Energy and any new transformer entrant. ETN's long-standing supplier relationships with Cleveland-Cliffs, Posco, and Nippon Steel give it allocation priority. A new hyperscaler-backed transformer venture (absent a GOES supply agreement) cannot enter at meaningful scale. The chokepoint is a moat reinforcer, not a threat, in relative terms.
• Pass-through power is structurally durable. CPI mechanism, oligopolistic market structure (ETN / Siemens / Schneider / ABB/Hitachi for MV switchgear and power transformers), engineering-to-order lock-in, and 12-18 month delivery commitments all support pricing discipline through input cycles. This is the inverse of NVTS's squeezed-middle pass-through failure.
• Geographic diversification relative to cohort is superior. Near-zero Taiwan manufacturing exposure (vs. TXN at <5%, NVTS at ~60%). Even Mexico exposure is modest and partially hedged by USMCA and the Apr 2026 15% transitional tariff rate for grid equipment. ETN is a US-manufacturing-heavy business in a world rewarding US manufacturing.
• Capacity investment cycle peaks in 2026, not 2028. The $1.5B incremental capex program is being absorbed now. By 2027, the ramp cost headwind rolls off, new capacity is productive, and the high-margin data-center revenue mix flows through fully. The investment peak = margin trough = the buy point for supply-chain believers.
• SiC UPS cost tailwind emerging. As SiC MOSFET ASPs fall 15-25% over 2025-2027 (per the WOLF / onsemi capacity additions tracked in the cohort), ETN's 9395XC UPS and SiC-based power electronics see input cost declining on the semiconductor side. ETN is a beneficiary of SiC oversupply, not a victim.

§ 10Bear Points — Supply-Chain Dimension

• GOES scarcity is a real constraint on transformer revenue conversion. The $13.2B Electrical Americas backlog sounds impressive; the 128-144 week (Q2 2025) power-transformer lead time reveals that a large fraction of it does not become cash flow in 2026. US power-sector trade groups have formally flagged a "critical electrical steel crunch" (Power Magazine). Cleveland-Cliffs Weirton expansion is necessary but not sufficient.
• Legacy backlog pricing risk is unquantified. Eaton has not disclosed what fraction of the $19.6B backlog is fixed-price vs. CPI-adjustable. At 2022-2023 booking levels for large data center contracts, a fixed-price tranche at prior-cycle copper and GOES prices delivered in 2026-2027 at current input costs is a structural margin dilution of unknown size. This is the single disclosure that would most change this read.
• Mexico tariff exposure is not fully resolved. The IEEPA 25% tariff on Mexico goods (non-USMCA qualifying) is the risk. The Section 232 15% grid-equipment transitional rate through 2027 provides partial relief, but it is explicitly transitional and subject to political reconfiguration. A re-escalation would compress ~10-15% of ETN's Electrical Americas cost base by 25 points in unit economics.
• Manufacturing ramp costs will create a 2026 margin trough. Management confirmed 130 bps headwind in 2026, front-loaded in Q1-Q2. This is self-inflicted and temporary — but "temporary" margin compression at a 35x PE company creates real multiple risk if the ramp lasts longer than guided.
• Semiconductor chips (bought-in) carry Taiwan tail indirectly. ETN buys SiC, IGBT, GaN, and MCU components from IFX, ON, TI, STM — all of which have some Taiwan-located foundry dependency. In a Taiwan disruption event, ETN's power electronics supply chains would be impaired even though ETN itself has no Taiwan operations. The impact is 1-2 orders of magnitude smaller than for TSMC or NVDA, but it is not zero.

§ 11Chip-to-Grid Position: ETN's Relationship to the Semiconductor Cohort

ETN sits at G1-G3 in the chip-to-grid stack: utility interface, datacenter power backbone, and 800V transition layer. Its relationship to the semiconductor side of the cohort is that of a customer, not a supplier.

What ETN buys from the chip cohort:

• From TI (TXN): Gate driver ICs, MCUs, analog power management chips for relay protection, switchgear control electronics
• From Infineon (IFX): IGBT modules for large UPS; SiC MOSFETs for the 9395XC
• From ON Semi: SiC MOSFETs and IGBTs for power conversion stages
• From STMicro: IGBTs and gate drivers
• From MPWR: Control ICs for power-factor correction and DC-DC conversion

BOM share for power semis:

• Large UPS (9395XC, 1125-2250 kVA): power electronics (IGBT/SiC + gate drivers + MCUs) estimated at 5-12% of equipment BOM cost — significant in aggregate given ETN's UPS revenue, but not the dominant cost driver (copper transformer/inductor windings, cabinets, and batteries are larger)
• Switchgear (MV): power electronics content much lower — protection relays, measurement ICs, comms modules — estimated <2% of BOM
• Transformers (power and distribution): essentially zero active power semis in the core product; only in monitoring/control electronics

ETN spec control over semiconductor vendor choice: For the UPS product line, ETN engineers the topology and qualifies specific die families — it is not component-agnostic. For the 9395XC, the SiC MOSFET selection is ETN's design choice (not the customer's). For switchgear protection relays, ETN specifies the MCU/relay IC; the end customer cannot substitute. This is partial spec lock-in on the buy side — ETN can negotiate multi-vendor contracts with IFX and ON for SiC supply but must re-qualify each new device.

§ 12Conviction (1-5)

3 / 5 on the supply-chain dimension specifically.

The score reflects: (a) a structurally sound tier-1 commodity picture — copper, aluminum, steel are multi-sourced, and the commodity tailwinds from the April 2026 15% grid-equipment tariff rate and falling SiC ASPs are real; (b) the GOES chokepoint is real but affects all transformer incumbents equally, making it a moat enforcer as much as a constraint; (c) pass-through power is the strongest in the cohort's infrastructure segment — the CPI mechanism, oligopolistic structure, and demonstrated margin discipline are genuine; (d) the geography is better than the semiconductor names — near-zero Taiwan exposure, US-heavy, Mexico manageable under USMCA.

Not higher because: (e) own-manufacturing capacity is the binding constraint, and the ramp timeline risk is material and unresolved — 130 bps margin headwind confirmed for 2026; (f) legacy-backlog fixed-price exposure is unquantified and could be the single largest unknown in the supply-chain picture; (g) GOES scarcity makes transformer delivery commitments structurally less certain than the backlog headline implies; (h) the calendar mismatch thesis (synthesis Open Question §2) has a 40-55% probability of materializing — the 2027-2028 revenue inflection may be later than the 2026 investor narrative suggests.

This is a 3 for supply-chain paired with a structurally higher conviction on the demand thesis. The supply-chain constraint is a timing risk, not a structural failure. If the capacity ramp executes on plan and the legacy-backlog pricing gap is disclosed to be small, the supply-chain conviction moves to 4.

§ 13Key Risks to This Read

• Assumption: legacy backlog is predominantly CPI-adjustable, not fixed-price. If >20% of the $19.6B backlog is fixed-price (2022-2023 era pricing), the margin dilution on 2026-2027 deliveries is $200-500M of below-normal profitability. ETN's 10-K does not disclose this split. This is the single disclosure that would most change the read.
• Assumption: Cleveland-Cliffs Weirton GOES expansion adds 30-40% additional tonnage by 2H 2025 as announced. A 12-month slip in this expansion creates a supply crunch that directly limits ETN transformer output in 2026. I am taking management's press release at face value; there is no third-party confirmation of actual ramp timing.
• Assumption: USMCA rules-of-origin compliance holds for ETN's Mexico-assembled goods. If steel content in ETN's Mexico products is sourced from non-USMCA countries and fails the "melted and poured" test, the 25% IEEPA tariff applies. This is a legal/compliance question I cannot resolve from public disclosures.
• Assumption: The 24 capacity-expansion projects ramp on plan with the half completing in 1H 2026 as guided. Hurricane Helene disrupted one ETN facility in 2024; construction and equipment installation timelines have consistently lagged in the industrial sector. A 6-month average delay on the remaining 12 projects is the central downside case.
• Assumption: SiC MOSFET supply for UPS continues to normalize. If a Taiwan disruption event affects Infineon's SiC foundry partners or ON Semi's Bucheon site, ETN's 9395XC UPS production could be disrupted. This is a 5-10% probability tail, not a base-case scenario.
• Disclosure that would most change this read: (1) Disclosure of fixed-price vs. CPI-adjustable split in backlog; (2) GOES supply agreements by source (domestic / import / tons committed) in ETN's 2025 10-K; (3) Named factory expansion timelines and specific capacity additions (ETN describes "two dozen projects" without identifying most of them); (4) Section 232/USMCA tariff impact quantification (dollar headwind) for 2026; (5) SiC MOSFET vendor mix for 9395XC and stated requalification status of alternative suppliers.

§ 14Sources

• Eaton Corporation 10-K (filed February 2025, fiscal year ending December 31, 2024) — Item 1 raw materials, Item 2 properties, risk factors. https://www.stocktitan.net/sec-filings/ETN/10-k-eaton-corp-plc-files-annual-report-52453dbea12c.html
• Eaton Q4 2025 Earnings Call Transcript (February 3, 2026). https://www.fool.com/earnings/call-transcripts/2026/02/03/eaton-etn-q4-2025-earnings-call-transcript/
• Eaton Q4 2025 earnings press release — Record $19.6B backlog, 200% data center order growth, $1.5B capex. https://www.eaton.com/us/en-us/company/news-insights/news-releases/2026/eaton-reports-record-fourth-quarter-2025-results.html
• Eaton Q3 2025 Earnings Call Transcript — backlog growth, capacity ramp commentary. https://www.fool.com/earnings/call-transcripts/2025/11/05/eaton-etn-q3-2025-earnings-call-transcript/
• Eaton Q3 2025 Analyst Presentation — segment margins, book-to-bill, mega-project backlog. https://www.eaton.com/content/dam/eaton/company/investor-relations/quarterly-earnings/filings/2025/q3/3Q-2025-analyst-presentation.pdf
• Eaton Q1 2025 earnings press release — record results, raised organic growth guidance. https://www.eaton.com/us/en-us/company/news-insights/news-releases/2025/eaton-reports-record-first-quarter-2025-results.html
• Eaton press release — New Dominican Republic manufacturing site (fuses for EV, data center). https://www.eaton.com/us/en-us/company/news-insights/news-releases/2024/eaton-cuts-ribbon-on-new-dominican-republic-manufacturing-site.html
• Eaton press release — $85M North American manufacturing investment (2023). https://www.eaton.com/us/en-us/company/news-insights/news-releases/2023/eaton-investing-85-million-to-increase-north-american-manufactur.html
• Eaton press release — $150M manufacturing investment for electrical infrastructure (2023). https://www.eaton.com/us/en-us/company/news-insights/news-releases/2023/eaton-invests-150-million-to-increase-manufacturing.html
• Eaton Smart Factory announcement — Juarez (Mexico) and Brazil modernization, 2024. https://www.eaton.com/us/en-us/company/news-insights/news-releases/2024/eaton-modernizes-its-manufacturing-footprint-with-two-new-smart.html
• Eaton Commodity Price Index tool. https://www.eaton.com/us/en-us/products/utility-grid-solutions/commodity-price-index--cpi-.html
• Eaton 9395XC UPS — SiC-based topology, 97.5% efficiency, 1125-2250 kVA. https://www.eaton.com/us/en-us/catalog/backup-power-ups-surge-it-power-distribution/eaton-9395xc-ups.html
• Eaton Supercapacitor backup power solution brochure. https://www.eaton.com/content/dam/eaton/products/backup-power-ups-surge-it-power-distribution/backup-power-ups/power-xpert-9395/eaton-supercapacitors-back-up-solution-brochure-br153059en.pdf
• Cleveland-Cliffs — New GOES transformer plant in Weirton WV, +30-40% tonnage. https://www.clevelandcliffs.com/news/news-releases/detail/644/cleveland-cliffs-announces-its-new-state-of-the-art
• Power Magazine — U.S. Power Sector Trade Groups Flag Critical Electrical Steel Crunch. https://www.powermag.com/u-s-power-sector-trade-groups-flag-critical-electrical-steel-crunch/
• Electrical Trader — Transformer Shortages: Supply Chain Impact on Pricing (128-144 week lead times for power transformers). https://electricaltrader.com/blogs/news/transformer-shortages-supply-chain-impact-pricing
• Taish Transformer / Industry data — Global Transformer Price Trends and Market Insights 2025. https://taishantransformer.com/global-transformer-price-trends-market-insights/
• BeyondSPX — Eaton: Powering the Electrification Megatrend with Record Backlogs. https://beyondspx.com/quote/ETN/analysis/eaton-powering-the-electrification-megatrend-with-record-backlogs-etn
• Electrical Trends — Eaton and Atkore: The Year That Was and What's Next (2026 outlook). https://electricaltrends.com/2026/02/10/eaton-and-atkore-they-year-that-was-and-whats-next/
• GEODIS / Perkins Coie / C.H. Robinson — Section 232 tariff expansion on aluminum, steel, and copper effective April 6, 2026. https://geodis.com/us-en/resources/customs-corner/united-states-announces-changes-section-232-steel-aluminum-and-copper
• Tariff headwind and EPS guidance commentary — Ainvest. https://www.ainvest.com/news/eaton-profit-forecast-cut-navigating-tariff-headwinds-volatile-market-2505/
• TIKR / Seeking Alpha — Eaton 7-9% organic growth guidance 2026, 30%+ Electrical Americas margin target. https://www.tikr.com/blog/eaton-stock-200-data-center-order-growth-and-a-7-trillion-market-puts-590-target-in-sight
• Cohort synthesis — Section 2 (G1-G3 in chip-to-grid stack), Section 3.2 (AI capex pass-through to power infra), Section 5 (tailwinds/headwinds table), Section 6 contested claim #16 (ETN corpus thin). ../synthesis.md
• Cohort corpus — "The AI Power Crisis Part 2" (NuttyCLD / Crucible Capital): switchgear, transformer, copper busbars, 800V DC data center architecture. ../corpus/corpus.md
• Cross-reference NVTS/supply-chain.md — foundry-counterparty risk analog; GOES analog framing. ../NVTS/supply-chain.md
• Cross-reference TXN/supply-chain.md — Taiwan-tail hedge posture; cohort geographic concentration frame; Section 232/USMCA tariff analysis. ../TXN/supply-chain.md