Supply Chain Analysis — Vertiv Holdings Co. (VRT)

§ 01Executive View

Vertiv's supply chain is structurally more resilient than its $15B backlog headline implies, but converts the cohort's calendar-mismatch risk into a more acute form than ETN: where ETN is gated by GOES and transformer winding capacity, VRT is gated by its own manufacturing ramp (new South Carolina sites, Ironton Ohio expansion, Mexicali expansion — none fully operational until 2H 2026-Q2 2027), by copper and aluminum commodity pricing (copper is the single largest input for UPS busbars and busway, now subject to Section 232 25% tariff), and by lithium battery cell supply for UPS BBUs (concentrated in CATL/Samsung SDI/LG Energy with China-origin Section 301 exposure). The pass-through story is strong and improving — management confirmed "pricing exceeded inflation in 2025" and expects "price/cost positive for the year inclusive of tariff impact" in 2026 — but this is a demonstrated pricing-power claim, not a contractual-escalator mechanism like ETN's CPI tool, leaving legacy-backlog margin exposure as the key unquantified risk. VRT does NOT make large power transformers or MV switchgear at scale in-house, reducing its direct GOES exposure relative to ETN, but it buys package substations that depend on the same GOES-constrained transformer supply chain — making it an indirect sufferer of the cohort's foundry-analog chokepoint.

§ 02Chain Map: Raw Materials to Customer

§ 03Input Map

Tier 1

Tier 2 Chokepoints

1. HFO refrigerant supply concentration (hidden, underpriced) Honeywell and Chemours collectively produce the majority of R1234ze (GWP <1) globally. The EU F-Gas Regulation (effective 2025 for new chillers) and US AIM Act HFC phasedown are forcing the entire HVAC/cooling industry to transition from R134a and R410a to HFO refrigerants simultaneously. VRT's chiller and CRAC/CRAH product lines are transitioning to R1234ze and R513A. The production capacity for HFO refrigerants, while growing, is being absorbed by a global demand spike from cooling equipment manufacturers (VRT, Carrier, Trane, Lennox, Daikin) all transitioning at the same time. A spot tightness event in 2026-2027 — plausible given the regulatory step-change in 2025 — would constrain VRT's chiller and CRAC/CRAH production throughput and force order deferrals. This is the analog of GOES for VRT's thermal product line: a regulatory-transition-driven scarcity affecting a concentrated supplier base.

2. CoolIT Systems CDU sub-supply (moderate concentration) VRT's liquid cooling portfolio — specifically the CoolChip CDU product family and some direct-to-chip subsystems — uses CoolIT Systems as a supplier of CDU sub-components (manifold loops, quick-disconnect fittings, pump assemblies for high-density CDU solutions). VRT made an early investment in CoolIT, establishing a partnership that predates the AI-DC cooling inflection. CoolIT is a private Canadian company (Calgary) with limited disclosed capacity. As VRT scales its liquid-cooling backlog — potentially the fastest-growing segment in 2026-2027 — a production bottleneck at CoolIT would directly limit throughput on VRT's most strategically important product line. VRT is likely dual-sourcing some CDU elements in-house, but CoolIT concentration is a named single-point risk not disclosed in VRT's 10-K.

3. Compressor supply (Copeland / Emerson dominant) Copeland compressors are the default choice for VRT's CRAC/CRAH thermal management units globally. Emerson spun off Copeland in mid-2023 as an independent company; Copeland's Sidney, Ohio facility produces scroll and reciprocating compressors for HVAC/datacenter cooling at scale. As a dominant OEM supplier to VRT (and to Schneider/Stulz/Airedale/Rittal), Copeland carries meaningful single-supplier concentration risk for VRT's air-side cooling product lines. An unplanned production disruption at the Sidney plant, or a capacity-tightness event as the entire industry transitions to high-density liquid/air hybrid systems, would constrain VRT's CRAC/CRAH manufacturing throughput.

4. Samsung SDI / CATL lithium cell sourcing (China-Korea geographic concentration) All major Li-ion cell producers for UPS BBU applications are in Northeast Asia (Korea: Samsung SDI, LG ES; China: CATL, EVE, Gotion; Japan: Panasonic). There is no US domestic large-format prismatic or pouch cell production for UPS applications at scale. Section 301 tariffs on CATL China-origin cells add meaningful cost pressure if VRT sources from CATL for US-bound UPS. VRT's named Samsung SDI partnership mitigates the CATL tariff risk for US applications, but Samsung SDI's Cheonan and Hongseong (Korea) plants represent a geographic concentration (Korean peninsula geopolitical risk, though lower than Taiwan). A Korea supply disruption — or an escalation of Section 301 to Korean cells — would leave VRT without a qualified domestic alternative for 6-12+ months.

5. Indirect GOES / transformer exposure via SmartRun / integrated power modules VRT's fastest-growing product category is integrated power modules (SmartRun, OneCore) that combine busway, cooling piping, networking, and containment with bought-in transformer/switchgear. These products incorporate distribution transformers procured from third-party transformer manufacturers who themselves face GOES-constrained lead times of 18-36 months for larger units. As VRT scales SmartRun — which is described as reducing on-site deployment time by up to 85% vs. traditional builds — the embedded transformer supply chain becomes a rate-limiter even though VRT is not the transformer manufacturer. This is the same calendar-mismatch mechanism as ETN's GOES constraint, one step removed.

§ 04Risk Scoring

Synthesis. VRT's supply-chain profile mirrors ETN's in structure (own-manufacturing-capacity is the binding constraint; tier-1 commodities are multi-sourced), but with two VRT-specific differences: (1) VRT has less GOES exposure (no large power transformer manufacturing) but more HFO refrigerant and CoolIT concentration; (2) VRT's backlog is more purely AI-DC concentrated than ETN's, so the calendar-mismatch risk is higher-beta — there is less diversified-industrial cushion if the ramp slips. Pass-through is strong; the structural input concern is the pace at which new manufacturing capacity converts the $15B backlog to revenue.

§ 05Manufacturing Footprint

Geographic exposure summary vs. cohort:

• Taiwan: near-zero (no VRT manufacturing in Taiwan; electronic components carry indirect Taiwan-tail via TSMC foundry for some IFX/TI chips)
• China: deliberately declining for US-bound product; residual Shenzhen serves China domestic
• Mexico: significant — Mexicali, Monterrey, Reynosa, Tijuana; 4 sites; USMCA-compliant for most output but Section 232 on metals applies regardless; management actively shifting China inputs to Mexico and pursuing USMCA compliance to minimize tariff exposure
• US domestic: growing rapidly with South Carolina x2, Ohio Ironton expansion, Pennsylvania facility
• VRT described its tariff mitigation as: "only single-digit percentage of US factory inputs from China" — implying the US factories are substantially de-risked from China tariffs; Mexico factories face Section 232 on metals but USMCA for finished goods

§ 06Pass-Through Power

VRT's pass-through posture is strong and has been demonstrated through multiple input-cost cycles:

1. Pricing vs. inflation track record. "2025 pricing exceeded inflation" (Q4 2025 management). "We expect the same in 2026" (Q4 2025). "Price/cost positive for the year inclusive of tariff impact and the countermeasures" (Q1 2026).
2. Tariff offset speed. Management stated VRT was "already there at the end of Q1" [2026] in materially offsetting tariff impact on a margin-exit-rate basis — a faster offset than ETN achieved in its 2022-2024 commodity cycles, reflecting higher price agility in VRT's shorter-cycle, more purely AI-DC customer mix.
3. No formal CPI mechanism. Unlike ETN, VRT does not publish a Commodity Price Index or have disclosed formal contractual cost-escalator clauses. Pass-through is achieved via pricing discipline (price increases at order/renewal) rather than contractual pass-through. This means legacy-backlog exposure is potentially larger than ETN's — orders placed in 2023-2024 at then-prevailing input costs are delivered in 2026-2027 at current (higher) costs, with no contractual adjustment mechanism unless VRT has added escalator language for recent orders.
4. Oligopolistic market structure. VRT competes in an effective three-incumbent market for data-center power/cooling infrastructure: VRT, Schneider Electric, and ETN (with Delta Electronics growing). This market structure enables sustained pricing discipline. Data-center operators facing 12-18 month delivery lead times cannot easily re-source to competitors; VRT's design-wins on Blackwell/Rubin deployments are sticky.
5. Service revenue moat. Service revenue at 30-35% incremental margins provides a pricing floor. Day-1 service contracts bundled with equipment purchases create a multi-year recurring revenue stream that partially hedges the cyclical product margin. Service engineer headcount of ~4,500 (Q4 2025) is a scaling asset that VRT is hiring to geography.

Score: 2 — strong pass-through, limited by absence of contractual escalator mechanism and legacy-backlog repricing opacity.

§ 07Backlog Conversion Analysis

Conversion bottleneck: $15B backlog at 12-18 month conversion implies approximately $10-12.5B theoretically shippable in 2026. Against $13.25-13.75B 2026 revenue guidance, this appears feasible — but only if manufacturing capacity ramps as guided. The South Carolina sites (7x regional capacity increase), Ironton Ohio (+45% thermal, Q2 2027), and Mexicali (+45% power) are the critical path. South Carolina is the most consequential: SmartRun integrated power modules and infrastructure solutions are the highest-growth product lines, and 7x capacity increase implies these sites are early-stage in 2026 rather than at full run-rate. The ramp risk is real and unquantified in terms of exact production-line milestone dates.

Backlog quality: Management confirmed all backlog is binding purchase orders (not soft LOI or framework agreements). Cancellation provisions are standard in complex engineered-to-order infrastructure (typically ≥20% cancellation penalty) but not disclosed by VRT. AI-DC capex is sticky — hyperscalers and colos that have issued POs are not canceling; the risk is deferral at the customer's request (push-right), not cancellation.

§ 08Stress Scenarios

Scenario 1: Mexico tariff escalation — USMCA benefits narrowed for metals-content goods

Probability: low-mid, 20-30%. The April 2026 Section 232 restructuring already imposes metals tariffs on copper/aluminum/steel derivatives regardless of USMCA status. If a further escalation (25% IEEPA on all Mexico-origin goods, or narrowing of the "15% transitional rate for electrical equipment") takes effect in 2H 2026 or 2027, VRT's four Mexico manufacturing sites face higher COGS on US-bound product.

Mechanism. Mexico accounts for a significant fraction of VRT's Americas manufacturing base (Mexicali, Monterrey, Reynosa, Tijuana). Under current USMCA, finished goods enter duty-free for non-metals content; under a new IEEPA 25% scenario on all Mexico goods, VRT's UPS and power distribution products manufactured in Mexico would face a 25% import duty on the full customs value entering the US.

Financial impact. Mexico manufacturing estimated at 15-20% of VRT's Americas COGS base. A 25% tariff on all Mexico-origin goods would be a 250-400 bps gross-margin headwind before pass-through. VRT has demonstrated ability to pass through tariff costs — but the speed of pass-through on a step-function tariff would create a 1-2 quarter margin compression. 2026 adjusted operating margin guidance (22-23%) would compress to 19-21% in a full-escalation scenario before offset.

Response options. Accelerate re-shoring to US domestic facilities (South Carolina, Ohio, Pennsylvania — already in progress); qualify additional USMCA-compliant steel/aluminum inputs at Mexico sites to ensure metals meet "melted and poured" rules-of-origin; pricing actions (VRT demonstrated Q1 2026 tariff-offset capability).

Scenario 2: HFO refrigerant supply tightness — Honeywell / Chemours production lags F-Gas demand transition

Probability: low-mid, 20-30%. The global cooling industry is transitioning from R134a, R410a, and other HFCs to HFO refrigerants (R1234ze, R1234yf, R513A) under EU F-Gas and US AIM Act mandates. Honeywell and Chemours are the dominant global HFO producers. If the demand-side transition (driven by regulatory deadlines) outpaces HFO capacity additions in 2026-2027, spot tightness for R1234ze emerges — constraining VRT's chiller and CRAC/CRAH production throughput.

Mechanism. VRT's thermal product lines require refrigerant as a production input. Without refrigerant, units cannot be completed, tested, and shipped. Spot-buy of HFO refrigerant at a premium is the mitigant, but if physical supply is constrained (not just priced), VRT faces production deferrals on its thermal segment.

Financial impact. Thermal management is an increasingly large fraction of VRT's revenue as liquid cooling scales. A 3-6 month production disruption in thermal due to refrigerant availability could defer $200-500M of revenue into 2H 2027. Margin impact: revenue deferral plus spot-buy cost premium if refrigerant is available at all.

Response options. Strategic refrigerant inventory build (refrigerant is storable; pre-buying 6 months of supply is standard practice for equipment OEMs); dual-source to Chemours (Opteon), Arkema (Forane), and Mexichem/Orbia as alternatives to Honeywell; accelerate product line transition to coolants not dependent on HFO (CDU/liquid-cooling systems use glycol-water, not refrigerant — the mix shift to liquid cooling is a natural hedge to refrigerant supply risk).

Scenario 3: Calendar mismatch materializes — 2H 2026 capacity ramp slips 6+ months

Probability: mid-high, 35-50%. The South Carolina sites (7x capacity expansion), Ironton Ohio (+45% thermal, Q2 2027 target), and Mexicali (+45% power) are the critical path for VRT's 2026-2027 revenue ramp. Industrial construction and manufacturing ramp projects have a consistent history of 3-9 month delays. If the South Carolina sites — which are the most ambitious in footprint expansion — take until Q4 2026 rather than Q2 2026 to reach production-level output, VRT's H2 2026 revenue falls 5-8% below guidance.

Mechanism. $13.25-13.75B 2026 revenue guidance implies ~$3.3-3.4B per quarter at steady state, with back-half loading (management typically guides conservatively for H1 and loads the ramp into H2). If H2 capacity is not available, the guidance range shifts down by $400-700M. The backlog does not shrink — the delivery schedule pushes right.

Financial impact. Revenue miss of 3-5% for full year 2026 ($400-700M vs. $13.5B midpoint); adjusted EPS tracking toward $5.50-5.70 vs. guided $5.97-6.07. Operating margin at 21-22% vs. guided 22-23% due to ramp costs on underutilized new capacity. Stock reaction: VRT at 53x EV/EBITDA (vs ETN at 28x) is more sensitive to execution risk — a 5% revenue miss could compress the multiple by 5-8 turns, creating significant downside vs. ETN.

Response options. Accelerate third-party contract manufacturing for lower-complexity products (cabinets, busway segments) to shift simpler assembly out and free factory capacity for high-complexity items (UPS, CDU); prioritize high-margin AI-DC accounts over industrial/colo for limited capacity; expedited freight to compress lead times (adds cost but captures revenue earlier); communicate ramp cadence granularly to close the investor-expectation gap.

§ 09Bull Points — Supply-Chain Dimension

• No large-transformer / GOES exposure. VRT does not manufacture power transformers or distribution transformers at scale. The GOES chokepoint that constrains ETN's backlog conversion is a second-degree (bought-in substation) issue for VRT, not a first-degree constraint. VRT's core products (UPS, CDU, busway, PDU) are not GOES-dependent, giving it a cleaner path to backlog conversion than ETN's transformer-heavy Electrical Americas segment.
• Fastest demonstrated pass-through in the cohort. Exited Q1 2026 having "materially offset" tariff impact — a faster offset cadence than ETN and far faster than NVTS (which has essentially no pass-through given its customer mix). Data-center infrastructure oligopoly supports this discipline durably.
• Mexico USMCA-compliant manufacturing reduces effective tariff burden. VRT described US factory inputs from China at "single-digit percentage" — implying the US manufacturing base is substantially de-risked from China tariffs. Mexico factories serve as a tariff-efficient Americas production platform for non-metals-intensive assembly under USMCA.
• Liquid cooling mix shift is a natural refrigerant-risk hedge. The more VRT's revenue shifts from air-side cooling (refrigerant-dependent CRAC/CRAH) to liquid-side cooling (CDU, direct-to-chip, chilled water — primarily glycol-water loop), the less exposure VRT has to the HFO refrigerant transition bottleneck. The AI-DC architecture trend is VRT's own hedge against its cooling-product supply risk.
• Service revenue flywheel reduces supply-chain fragility of product revenue. 4,500 field engineers, Day-1 service contracts, and services incremental margins of 30-35% represent a recurring revenue stream that is supply-chain-independent (people, not parts). As service grows as a % of total, VRT's top-line resilience to a product-side supply disruption increases.

§ 10Bear Points — Supply-Chain Dimension

• Own-manufacturing capacity is the binding constraint and the ramp is ambitious. 7x South Carolina capacity, +45% Ohio Ironton (Q2 2027), +45% Mexicali — none at full rate in 2026. The $13.5B 2026 revenue midpoint requires a production ramp with essentially no margin for delay. At VRT's 53x EV/EBITDA multiple (vs ETN's 28x), any slip in the ramp creates disproportionate multiple compression.
• No formal CPI escalator mechanism exposes legacy-backlog orders to input-cost drag. The $15B backlog was accumulated rapidly in 2024-2025 when copper, lithium cells, and refrigerants were all at elevated prices. Orders booked in mid-2024 at then-prevailing input costs are delivered in 2026-2027 at current (still elevated) costs, with no contractual protection unless VRT has added ad hoc escalator language. VRT has not disclosed the fixed-price vs. adjustable-price split in its backlog. This is the single most important undisclosed supply-chain risk.
• Lithium cell supply has no domestic fallback. All major Li-ion cell producers for UPS BBU applications are in Northeast Asia. Samsung SDI (Korea) is VRT's confirmed partner but a Korea-peninsula disruption or US trade-policy escalation to Korean cells would leave VRT without a 12-month-qualified alternative. This is a low-probability but high-impact tail risk as VRT's Li-ion UPS penetration grows.
• HFO refrigerant oligopoly (Honeywell/Chemours) creates a quiet supply-side fragility. The EU F-Gas and US AIM Act mandate simultaneous industry-wide transition to low-GWP refrigerants in a window (2025-2028) that coincides with VRT's maximum production ramp. Honeywell and Chemours control the majority of R1234ze supply globally. A production shortfall at either produces a constraint that VRT cannot easily substitute around given the 6-12 month product re-certification requirement for switching refrigerant grades.
• CoolIT Systems CDU concentration — unquantified, private, growing. As VRT scales liquid cooling faster than any other product line, the CoolIT dependency for CDU sub-components scales with it. CoolIT is private, does not disclose capacity, and its production ceiling is opaque to investors. VRT has not quantified CoolIT as a % of its liquid-cooling BOM or disclosed any multi-sourcing progress on CDU manifold/pump assemblies.

§ 11Conviction (1-5)

3 / 5 on the supply-chain dimension specifically.

The supply-chain picture is structurally similar to ETN's at the same conviction level, but for different reasons. ETN's binding constraint is GOES / transformer winding capacity; VRT's is own manufacturing ramp pace. Both have strong pass-through. Both have the calendar-mismatch risk (Schneider's "2028-2030 real impact" framing as the counterweight to the backlog excitement). VRT's GOES insulation is a genuine differentiator vs. ETN. VRT's higher multiple (53x vs. ETN's 28x EV/EBITDA) makes execution risk more consequential per dollar invested. The supply-chain analyst's conviction of 3 reflects: (a) structurally sound tier-1 input picture; (b) strong pass-through demonstrated in real-time; (c) manageable geographic/tariff exposure vs. the cohort's Taiwan-heavy names; (d) penalized by own-manufacturing ramp risk (the binding constraint is self-imposed capacity), undisclosed legacy-backlog pricing split, and the CoolIT/HFO refrigerant concentration risks that are real but not quantified in filings.

§ 12Key Risks to This Read

• Assumption: legacy backlog is predominantly priced with current-cost pricing or adjustable terms. If a material fraction (>20%) of the $15B backlog carries 2023-2024-era fixed pricing delivered at 2026-2027 input costs, the margin impact on deliveries is larger than management's "price/cost positive" commentary implies. VRT has not disclosed this split. This is the single disclosure that would most change this read.
• Assumption: South Carolina and Ironton ramp on announced timelines. The two South Carolina sites are described as providing "7x regional capacity when fully ramped" — an aggressive expansion for industrial manufacturing. If ramp is delayed by 6+ months (40-50% probability of at least some delay in complex industrial build-outs), 2026 revenue tracking below the $13.25-13.75B guidance range.
• Assumption: CoolIT Systems has adequate capacity to support VRT's liquid-cooling backlog ramp. This cannot be confirmed from public disclosures; CoolIT is private. If CoolIT is the rate-limiting sub-supplier for CDU assembly, VRT's liquid-cooling segment — its fastest-growing and strategically most important product line — could be capacity-constrained from a supplier-side, not a VRT-factory-side, bottleneck.
• Assumption: USMCA compliance holds for VRT's Mexico-manufactured goods. If steel/aluminum inputs in VRT's Mexico facilities fail the "melted and poured" rules-of-origin test (common if sourced from Asian steel distributors in Mexico), a 25% IEEPA tariff on the full customs value of Mexico-origin goods to the US becomes operative. Management commentary (Q1 2026) suggests they are actively managing this — but the compliance status is not disclosed in filings.
• Assumption: HFO refrigerant (R1234ze) supply normalizes as Honeywell/Chemours add capacity. Both companies have announced capacity expansions for HFO production. If those expansions are delayed or insufficient to meet the simultaneous industry-wide demand transition in 2025-2027, VRT's chiller and CRAC/CRAH production faces an input bottleneck not visible in current supply-chain disclosures.
• Disclosure that would most change this read: (1) Fixed-price vs. adjustable-price split in the $15B backlog; (2) Named CDU sub-supplier concentration by % of liquid-cooling BOM; (3) South Carolina site operational-readiness timeline with milestone granularity (not just "fully ramped eventually"); (4) Tariff dollar headwind quantified for 2026 (currently managed but not sized for investors); (5) Li-ion cell supplier mix by volume (Samsung SDI vs. CATL vs. LG ES) for US-bound UPS shipments.

§ 13Sources

• Vertiv Holdings Q4 2025 Earnings Press Release — Record Q4 results, $15B backlog, 252% organic order growth, 2026 guidance ($13.25-13.75B revenue, $5.97-6.07 EPS, 22-23% operating margin). https://investors.vertiv.com/news/news-details/2026/Vertiv-Reports-Strong-Fourth-Quarter-with-Organic-Orders-Growth-of-252-and-Diluted-EPS-Growth-of-200-Adjusted-Diluted-EPS-37/default.aspx
• Vertiv Q4 2025 Earnings Call Transcript (February 11, 2026) — backlog conversion window (12-18 months), binding purchase orders, pricing discipline, manufacturing capacity commentary. https://www.fool.com/earnings/call-transcripts/2026/02/11/vertiv-vrt-q4-2025-earnings-call-transcript/
• Vertiv Q1 2026 Earnings Call Transcript (April 22, 2026) — tariff offset "materially achieved at end of Q1," price/cost positive, backlog elongated into 2027, service scaling, 800V DC product launch 2H26. https://www.fool.com/earnings/call-transcripts/2026/04/22/vertiv-vrt-q1-2026-earnings-transcript/
• Vertiv Q1 2026 Press Release — Q1 organic sales +23% YoY, raised 2026 guidance. https://investors.vertiv.com/news/news-details/2026/Vertiv-Reports-Strong-First-Quarter-with-Diluted-EPS-Growth-of-136-Adjusted-Diluted-EPS-Growth-of-83-Raises-Full-Year-Guidance/default.aspx
• Vertiv Annual Report 2025 (Fiscal Year 2024 10-K) — supply chain risk factors, raw materials sourcing strategy ("more than one supplier"), single-source risk disclosure, geographic manufacturing footprint. https://s205.q4cdn.com/554782763/files/doc_financials/2024/ar/Vertiv-Annual-Report_10K_2025.pdf
• Vertiv 2025 Annual Report (Fiscal Year 2025 10-K, filed February 2026) — full-year 2025 net sales $10,229.9M, backlog $15.0B. https://s205.q4cdn.com/554782763/files/doc_financials/2025/ar/Vertiv-2025-Annual-Report.pdf
• Vertiv Manufacturing Expansion Announcement (March 24, 2026) — South Carolina (2 sites, 7x capacity), Mexicali Mexico (+45% power capacity), Pennsylvania (AI rack cabinets). https://www.vertiv.com/en-us/about/news-and-events/corporate-news/2026/vertiv-announces-expansion-of-manufacturing-capacity-spanning-infrastructure-solutions-power-and-rack-systems-to-meet-rising-demand/
• Vertiv Ohio Manufacturing Expansion — Ironton OH $50M investment, +45% liquid cooling / chilled water capacity, operational Q2 2027. https://www.vertiv.com/en-us/about/news-and-insights/corporate-news/2026/vertiv-to-expand-ohio-manufacturing-to-boost-u.s.-production-of-critical-thermal-management-technologies-for-ai-data-centers/
• Vertiv 2017 Li-ion UPS announcement (Samsung SDI batteries confirmed for large UPS systems). https://www.vertiv.com/en-us/about/news-and-insights/corporate-news/2017/vertiv-to-offer-lithium-ion-battery-option-with-large-size-ups-systems/
• Vertiv Samsung Lithium-Ion product page — Samsung SDI confirmed as UPS energy storage partner. https://www.vertiv.com/en-us/products-catalog/critical-power/uninterruptible-power-supplies-ups/samsung-lithium-ion-batteries/
• Vertiv CoolChip CDU product — liquid cooling portfolio including CoolChip and CoolPhase CDUs. https://www.vertiv.com/en-us/products-catalog/thermal-management/high-density-solutions/vertiv-coolchip-cdu/
• Vertiv and Compass Datacenters partnership — multi-billion-dollar liquid-cooling supply agreement; CoolPhase Flex first deployment Q1 2025. https://www.vertiv.com/en-emea/about/news-and-insights/news-releases/vertiv-and-compass-datacenters-collaborate-on-combination-liquid-and-air-cooling-system-to-accelerate-deployment-of-liquid-cooling-for-ai-applications/
• Vertiv Refrigerants of the Future white paper — R1234ze, regulatory phaseout, HFO transition. https://www.vertiv.com/4ae9e9/globalassets/documents/white-papers/vertiv-refrigerants_of_the_future-wp-uk-low-rev.1-09.2018_246049_0.pdf
• Vertiv Low-GWP Refrigerants white paper — US AIM Act, EU F-Gas, R513A transition, 56% lower GWP vs R134a. https://www.vertiv.com/48f7e3/contentassets/8eee3437fd734143a36d0395e65e8de7/vertiv-low-gwp-refrigerants-wp-en-na-sl-71228-web.pdf
• Section 232 Tariff Expansion (April 2026) — copper derivatives taxed at full customs value, not just metal content; 25% base rate, 15% transitional for electrical/industrial equipment. https://mohawkglobal.com/word-from-the-white-house/new-section-232-tariff-changes-on-steel-aluminum-copper-what-importers-need-to-know/
• USMCA / Section 232 interaction — Section 232 on metals applies even to USMCA-qualifying goods; USMCA confers MFN duty relief only. https://www.cello-square.com/en/marketupdates/view-2213.do
• Seeking Alpha — Vertiv $15B backlog, liquid cooling dominance, 2026 AI infrastructure analysis. https://seekingalpha.com/article/4890719-vertiv-holdings-the-15-billion-backlog-liquid-cooling-dominance-and-the-ai-infrastructure-trade-wall-street-is-still-underpricing
• Seeking Alpha — Vertiv 2026 guidance ($6.02 EPS, $13.5B sales), capacity expansion strategy. https://seekingalpha.com/news/4550824-vertiv-projects-6_02-eps-and-13_5b-sales-for-2026-while-advancing-capacity-expansion-strategy
• Cooling Post — R134a chiller ban from 2025 (EU); HFO transition timeline. https://www.coolingpost.com/world-news/r134a-faces-chiller-ban-from-2025/
• Vertiv Wikipedia — manufacturing locations (Delaware OH, Ironton OH, Lincoln NE, Anderson SC, Pelzer SC, Tijuana, Mexicali, Monterrey, Reynosa — Americas; Italy, Czech, Croatia, India, Brazil, China-Shenzhen internationally). https://en.wikipedia.org/wiki/Vertiv
• Cohort synthesis — Section 2 (VRT at G1-G4 in chip-to-grid stack), Section 3.2 (VRT $15B backlog / 2.9x book-to-bill as primary box-builder anchor), Section 3.5 (cooling transition / CDU pricing power), corpus AI Power Crisis Part 2. ../synthesis.md
• ETN supply-chain memo — GOES cohort analog, calendar-mismatch framing, ETN-VRT pair structure, cross-ticker tariff / Mexico analysis. ../ETN/supply-chain.md
• Refinement log — C-ETN-1 findings on VRT NVIDIA GTC partnership, cooling as a value-chain layer, backlog-conversion as the binding question. ../refinement-log.md