Scarcity Alpha — Part 6 — February 2026

Transformers & Grid Equipment

You can build a data center in 12 months, but the transformer it needs takes 3 years. The physical grid is the invisible bottleneck of the 21st century. Lead times for large power transformers have exploded from 30 weeks to 150+ weeks.

Severity 9/10 Industrials Infrastructure Highest Investability
Scarcity Alpha6/13
Severity The Crisis Lead Times GOES Steel Backlogs Companies Trade Setups Risks

Shortage Severity Assessment

Severity: 9/10
Expected Duration
Medium (3-5y)

Capacity expansion is slow due to specialized labor and GOES steel constraints. New factories take 3+ years to build and qualify.

Investability
Highest in Series

Publicly traded pure-plays with record backlogs, pricing power, and multi-year earnings visibility. Book-to-bill >1.2.

Severity Breakdown

FactorScoreRationale
Substitutability10/10No alternative to transformers for voltage conversion; physics demands them at every grid node
Supply Response Time9/10New manufacturing capacity takes 3-4 years; specialized workforce takes 5+ years to train
Demand Visibility10/10Utility 10-year capex plans, data center interconnection queues, and IRA/IIJA funding are locked
Input Constraints9/10GOES (grain-oriented electrical steel) has only 5 major global producers; 2-year lead times for steel
Pricing Power9/10Transformer prices up 60-80% since 2020; customers accept because alternative is no power
Workforce Bottleneck8/10Average transformer factory worker is 55+; apprenticeship pipeline is empty

The Grid Crisis: Three Decades of Underinvestment

The US electrical grid was largely built between 1960 and 1980. For the next four decades, investment was minimal — utilities optimized for shareholder returns by sweating existing assets. Today, the average age of a US large power transformer is 40+ years, approaching or exceeding the 40-year design life. The result is a grid that is simultaneously aging out and being asked to carry exponentially more load from EVs, data centers, heat pumps, and renewable interconnection.

Avg US Transformer Age
40+
years (design life: 40)
Grid Investment Needed
$2T+
by 2035 (globally)
LPT Manufacturers
~5
globally for large units
US Interconnection Queue
2,600 GW
waiting for grid connection

Demand Drivers Converging Simultaneously

DriverGrid Equipment DemandTimelineScale
AI Data CentersSubstation + transformer per 50-100 MW campus2024-203030+ GW of new capacity announced in the US alone
Renewable InterconnectionStep-up transformers at every solar/wind farmOngoing2,600 GW in US interconnection queue (5x installed capacity)
EV Charging InfrastructureDistribution transformers for fast-charging hubs2025-20351.2M public chargers needed in the US by 2030
Grid ReplacementLike-for-like replacement of aging fleetAlready overdue70% of US LPTs are 25+ years old
Re-shoring / IRA ManufacturingNew factories need grid connections2024-2030$200B+ in announced US manufacturing investment
Grid Hardening (Climate)Flood/fire-resistant equipment upgradesOngoing$50B+ in utility climate resilience capex

Why Just-in-Time Died for Transformers

For 30 years, utilities bought transformers with 6-12 month lead times and kept zero inventory. They treated these multi-million-dollar, custom-engineered pieces of equipment like off-the-shelf commodities. When demand surged simultaneously from data centers, renewables, and replacement needs, the supply chain collapsed. Today, if a large power transformer fails catastrophically, there is no spare. The utility must wait 150+ weeks for a replacement — nearly 3 years. This "inventory panic" is now driving a massive capex supercycle where utilities are double-ordering (ordering from two manufacturers to secure at least one delivery), pre-ordering 5 years ahead, and even paying 50%+ premiums for expedited delivery. The just-in-time model is dead. What has replaced it is a scramble for physical equipment that resembles the semiconductor shortage of 2021-2022.

Utility Capex: The $2 Trillion Supercycle

US utilities alone have announced $600B+ in grid investment through 2030, up from $350B in similar plans just two years ago. Globally, the figure exceeds $2 trillion. This is not aspirational — these are board-approved capital budgets driven by regulatory mandates, reliability requirements, and customer demand from data centers and manufacturing re-shoring.

Utility5-Year Grid Capex PlanAnnual Growth RateKey Driver
Southern Company (SO)$48B (2024-2028)+12% YoYData centers in Georgia (Meta, Google); nuclear expansion (Vogtle)
NextEra Energy (NEE)$40B (2024-2027)+10% YoYRenewable interconnection; Florida grid hardening
Duke Energy (DUK)$73B (2024-2028)+8% YoYGrid modernization; coal retirement replacement
AEP (AEP)$43B (2024-2028)+9% YoYData center demand in Virginia/Ohio; transmission buildout
Dominion Energy (D)$43B (2024-2028)+11% YoYVirginia data center alley; offshore wind interconnection
Eversource (ES)$23B (2024-2028)+7% YoYGrid reliability; storm hardening in New England

The Interconnection Queue Crisis

As of 2025, approximately 2,600 GW of generation projects are waiting in US interconnection queues — more than twice the total installed US generation capacity of ~1,200 GW. The average wait time has stretched from 2 years (2010) to 5+ years (2025). The bottleneck is not regulatory approval alone — it is the physical availability of transformers, switchgear, and substations needed to connect new generation to the grid. FERC Order 2023 attempts to address this by requiring "first-ready, first-served" processing, but the physical equipment constraint remains. Every GW of new solar or wind capacity requires step-up transformers, protection equipment, and grid infrastructure that the same overwhelmed manufacturers must produce.

Why Grid Equipment Demand Is "Non-Discretionary"

Unlike many industrial products, grid equipment demand is not cyclically sensitive in the traditional sense. Utilities are regulated monopolies whose capex is approved by state public utility commissions and recoverable through rate base increases. Once a utility commits to a grid investment, it is building whether the economy is booming or in recession — because reliability requirements are legally mandated. Similarly, data center operators must have power, period. A $5 billion data center campus that cannot operate because it lacks a $10 million transformer is not going to cancel the transformer order because of a recession. The "must-have" nature of this equipment creates a demand floor that is largely immune to economic cycles.

Lead Time Explosion

Source: Wood Mackenzie, Rystad Energy, DOE Grid Deployment Office. Large Power Transformer (LPT) = 345kV+ class.

Lead Times by Equipment Type

Equipment Category2020 Lead Time2025 Lead TimeChangeImpact
Large Power Transformers (345kV+)30-40 weeks150-200 weeks+400%Delays every data center, solar farm, and substation project
Medium Power Transformers (69-230kV)20-30 weeks80-120 weeks+300%Distribution grid upgrades stalled
Distribution Transformers8-12 weeks40-60 weeks+400%Residential solar, EV charger installations delayed
Switchgear (HV)16-20 weeks52-80 weeks+300%Substations incomplete without switchgear
Power Circuit Breakers12-16 weeks40-60 weeks+275%Safety and grid protection delays

GOES Steel: The Hidden Constraint

Even if you could build a new transformer factory overnight, you still need Grain-Oriented Electrical Steel (GOES) for the core. GOES is a specialty steel with precisely aligned crystal structures that minimize energy loss during magnetic flux transformation. It is the critical input for every transformer ever built, and there are only five major producers globally.

GOES Steel: The Hidden Constraint

What is GOES? Grain-Oriented Electrical Steel is a silicon-iron alloy where crystal grains are aligned in a single direction through a complex cold-rolling and annealing process. This alignment reduces core losses (energy wasted as heat during voltage transformation) by 75% compared to non-oriented steel. A large power transformer core requires 50-200 tonnes of GOES.

Why is it scarce? GOES production is capital-intensive ($500M+ for a new rolling mill), technically demanding (only ~5 companies globally have mastered the process), and faces competing demand from EV motor cores and renewable energy generators. Lead times for GOES have stretched to 12-18 months, up from 4-6 months pre-pandemic. US anti-dumping tariffs on Chinese and Russian GOES further constrain Western supply. The result: even manufacturers with open factory capacity cannot get enough steel to fill orders.

Major GOES Producers

ProducerCountryEst. Capacity (kt/yr)ShareStatus
Nippon SteelJapan~500~25%Full capacity; expanding slowly
JFE SteelJapan~300~15%Full capacity
POSCOSouth Korea~300~15%Adding capacity for EV motors
Baowu SteelChina~400~20%Subject to anti-dumping tariffs in US/EU
ThyssenKrupp / StalproduktGermany / Poland~250~12%European supply; tight allocation
Others (NLMK, Cleveland-Cliffs)Russia / USA~250~13%NLMK sanctioned; Cliffs expanding

Order Backlogs & Book-to-Bill

Source: Company quarterly filings, earnings calls. Backlog figures represent total order book in billions USD.

What Book-to-Bill Tells You

Book-to-bill ratio measures new orders ("bookings") divided by revenue shipped ("billings"). A ratio of 1.0 means the company ships as fast as it receives orders — the backlog stays flat. A ratio above 1.2 means orders are growing 20%+ faster than the company can produce, so the backlog is expanding. For grid equipment makers, sustained book-to-bill above 1.2 means: (a) demand is outstripping capacity, (b) the company has pricing power (customers can't go elsewhere), and (c) revenue visibility extends years into the future. This is the single most important metric for these stocks. As long as book-to-bill stays elevated, the earnings upgrades keep coming.

The Double-Ordering Problem

Utilities have started double-ordering — placing orders with two or three manufacturers simultaneously to hedge against delivery delays, then canceling the extras once one arrives. This inflates backlogs and creates a risk of "backlog cliff" if cancellations accelerate. However, the evidence suggests double-ordering accounts for only 10-15% of current backlogs, based on manufacturer commentary. The remaining 85-90% represents genuine demand. Even adjusting for double-ordering, the supply-demand imbalance remains severe and extends through at least 2028.

The Secondary Market: Scarcity Premium in Action

A parallel market has emerged for used and refurbished transformers. Utilities that cannot wait 3 years for a new unit are buying second-hand equipment at extraordinary premiums. A used large power transformer that would have sold for 20-30% of new cost five years ago now sells for 70-90% of new replacement cost. In some cases, utilities have paid more than the price of a new unit simply because the used one is available immediately.

Used LPT Premium (2020)
20-30%
of new replacement cost
Used LPT Premium (2025)
70-90%
of new replacement cost
New LPT Price Increase
+60-80%
since 2020

This secondary market behavior is a clear signal of scarcity. When buyers are willing to pay near-new prices for decades-old equipment, it confirms that: (a) demand is genuinely desperate, (b) there is no alternative source, and (c) the shortage is structural, not a temporary supply chain hiccup. For equipment manufacturers, this also means that every unit produced generates maximum margin — there is zero pricing pressure from competition because there is no spare capacity anywhere in the system.

Grid Equipment Makers: Fundamental Comparison

Company Ticker Market Cap Backlog Book/Bill Operating Margin Revenue Growth (YoY) Key Exposure
Eaton Corp ETN ~$130B $13.6B 1.2x 22% +11% Electrical components, switchgear, UPS, data center power
Powell Industries POWL ~$5B $1.3B 1.4x 19% +42% Custom switchgear, bus duct, LNG/petrochemical
Hubbell Inc HUBB ~$22B $2.1B 1.1x 24% +8% Utility T&D, pole-mount transformers, connectors
GE Vernova GEV ~$90B $45B+ 1.3x 12% +15% Grid solutions, onshore/offshore wind, gas turbines
Siemens Energy ENR.DE ~$55B $120B+ 1.4x 8% +18% Grid tech, HV transformers, Gamesa (wind)
Hitachi Energy Hitachi sub Part of 6501.T $25B+ 1.3x 10% +20% HVDC, transformers, grid automation (ex-ABB grid)

Why Only ~5 Companies Can Make Large Power Transformers

A large power transformer (LPT) rated at 345kV or above is not a commodity product — it is a custom-engineered, one-of-a-kind machine weighing 200-400 tonnes that must operate for 40+ years without failure. Manufacturing one requires: (1) specialized winding machines that cost $10M+ each, (2) vacuum-pressure impregnation chambers for insulating oil treatment, (3) GOES steel cores hand-assembled by workers with 10+ years of training, and (4) testing facilities that can simulate full grid voltage. Building a new LPT factory costs $200-500M and takes 3-4 years. Training the workforce takes another 5+ years. There is no shortcut, no "fabless" model, and no software disruption coming. This is why the oligopoly of Siemens Energy, Hitachi Energy, GE Vernova, and a handful of others will persist for decades.

Trade Setups & Positioning

ETN (Eaton Corp) — The Compounder

Core Holding

Thesis: Dominant player in electrical components benefiting from three mega-trends: AI data centers, re-shoring, and energy transition. Backlog at record $13.6B (+19% YoY). Electrical Americas segment growing 15%+ with 22% margins. Management has guided for "above-market growth for the remainder of the decade." Quality compounder with visibility rarely seen in industrials.

Entry Zone
$310-340
Stop Loss
$280
Target 1
$420
Target 2
$500
R:R
1:2.3 / 1:4.7

POWL (Powell Industries) — The Specialist

High Growth / Volatile

Thesis: Niche player in custom-engineered switchgear and power control solutions. Massive exposure to LNG terminals, petrochemical plants, and data centers. Revenue up 42% YoY with 19% operating margins (up from 8% two years ago). Small-cap ($5B) with small float = volatile, but earnings trajectory is extraordinary. Book-to-bill of 1.4x means the backlog is still expanding.

Entry Zone
$200-240
Stop Loss
$175
Target 1
$320
Target 2
$400
R:R
1:2.3 / 1:4.0

GEV (GE Vernova) — The Grid + Gas Turnaround

Turnaround / Scale

Thesis: Spun out of GE in April 2024, GE Vernova combines the grid solutions business (transformers, HVDC, substations) with the world's largest gas turbine franchise. Grid segment backlog exceeds $45B. Margins are still recovering from the wind turbine losses (Offshore Wind was a drag), but the core Grid + Gas businesses are highly profitable and growing. As Offshore Wind losses shrink and Grid margins expand, EPS has significant upside.

Entry Zone
$350-380
Stop Loss
$310
Target 1
$480
Target 2
$550
R:R
1:2.2 / 1:3.6

Trade Thesis Summary

Grid equipment stocks are the highest-conviction trade in the Scarcity Alpha series. Unlike commodity plays where price volatility creates drawdowns, grid equipment makers have multi-year backlog visibility, pricing power through long-term contracts, and margin expansion as fixed costs are spread over higher volumes. The setup: buy on quarterly pullbacks (earnings "sell-the-news" or broader market corrections), hold for 12-24 months, and let backlog conversion drive earnings beats. The key risk is valuation — these stocks have re-rated significantly. But as long as book-to-bill stays above 1.0 and backlogs grow, the multiple is justified.

Reinforcement Signals

  • Book-to-bill ratio sustained above 1.2 for 2+ consecutive quarters
  • Utilities announcing 10-year capex plans with 8%+ annual grid investment growth
  • Secondary market transformer prices continuing to rise (scarcity premium)
  • FERC grid reform and interconnection queue acceleration

Invalidation Signals

  • Book-to-bill dropping below 1.0 for 2+ quarters (demand normalization)
  • Chinese transformer imports bypassing tariffs at scale
  • GOES steel prices collapsing (would signal demand destruction)
  • Major data center project cancellations or deferrals

Timing & Position Sizing

Horizon: Medium-term (12-24 months). Catalysts: Eaton Investor Day (May 2026), Infrastructure Bill funding tranches (rolling), FERC interconnection reform orders (Q2-Q3 2026), quarterly earnings (book-to-bill data). Sizing: ETN 4-5% (core), POWL 2-3% (satellite, higher volatility), GEV 3-4% (turnaround leverage). Total grid exposure: 8-12% of portfolio. Beta: ETN ~1.1, POWL ~1.8, GEV ~1.3. Scaling: Build 60% at entry, add 20% on earnings pullback, final 20% on backlog confirmation. These are "hold through volatility" positions — backlog visibility means quarterly misses are buying opportunities, not exit signals.

Risk Framework

RiskProbabilityImpactMitigation
Demand Normalization Medium High If data center buildout slows or interest rates kill utility capex. Monitor book-to-bill as early warning.
Chinese Import Competition Medium Medium Chinese manufacturers (TBEA, XD Group) expanding globally. US tariffs provide buffer, but allies may not maintain barriers.
Margin Compression from GOES Costs Medium Medium Rising input costs could squeeze margins. Mitigated by cost-plus contracts and pricing power from backlogs.
Valuation Multiple Compression Medium Medium ETN trades at 32x forward P/E. If growth decelerates, multiple could compress 20-30%. Buy on dips.
Backlog Cancellations (Double-Ordering Unwind) Low Medium Estimated at 10-15% of backlogs. Real demand absorbs any cancellations. Watch net orders closely.
Project Delays (Permitting, NIMBY) Medium Low Delays push out revenue but do not destroy demand. Backlog stays elevated longer = more visibility.

The Valuation Debate

The most common pushback on grid equipment stocks is: "They're too expensive." ETN at 32x, GEV at 45x, POWL at 25x. These are industrial multiples that look more like tech. But the comparison misses the point. These companies have multi-year revenue visibility that most tech companies would envy. When your backlog is 2-3x annual revenue and growing, when your book-to-bill is >1.2, and when your margins are expanding because demand exceeds supply, a premium multiple is rational — it reflects certainty, not euphoria. The real risk is not overpaying today; it is waiting for a pullback that may not come until the cycle turns, by which point you've missed 50%+ of the move. The better strategy is to size appropriately and build positions gradually.

The Global Dimension: Not Just a US Problem

While this analysis focuses primarily on US-listed equities, the grid equipment shortage is a global phenomenon. Every major economy is simultaneously facing the same convergence of grid aging, electrification demand, and renewable interconnection needs.

RegionGrid Investment (2024-2035)Key ChallengeMajor Equipment Buyer
United States$600B+ (utility plans)Aging fleet, data centers, renewable queue (2,600 GW)Southern Co, Duke, AEP, NextEra
Europe (EU)$580B+ (REPowerEU)Cross-border interconnection, offshore wind grid, Russian gas replacementTenneT, Amprion, RTE, National Grid
China$500B+ (Five-Year Plan)Ultra-high voltage (UHV) transmission from western renewables to eastern citiesState Grid Corp, China Southern Power Grid
India$200B+ (Green Energy Corridor)500 GW renewable target by 2030; rural electrification; grid losses (20%+)PowerGrid Corp, state utilities
Middle East$150B+ (Vision 2030+)Industrial diversification, NEOM, data centers, desalination powerSaudi Electricity, ADWEA, DEWA
Southeast Asia$100B+ (ASEAN Grid)Demand growth 5-7%/yr; cross-border power trade; coal replacementPLN (Indonesia), TNB (Malaysia), EGAT (Thailand)

The global nature of this shortage is actually bullish for Western grid equipment makers. When Chinese manufacturers (TBEA, XD Group) are fully booked serving domestic demand, they have less capacity to export to the West at discount prices. The global demand surge creates a rising tide that lifts all manufacturers — even the less competitive ones cannot lose, because there is simply not enough capacity worldwide to meet demand.

The Workforce Crisis Nobody Talks About

Even if GOES steel appeared overnight and factory space were free, the grid equipment industry would still face a binding constraint: people. Winding a large power transformer core is a highly specialized skill that takes 5-10 years to master. The average age of these workers in Western factories is 55+. Apprenticeship pipelines were emptied during the decades of underinvestment. New training programs are being launched (Eaton, Siemens Energy, Hitachi all announced workforce initiatives in 2024-2025), but the graduates won't be production-ready until 2028-2030. This labor constraint is the least-discussed but potentially longest-lasting bottleneck in the entire grid equipment supply chain.

Catalyst Calendar

Eaton Investor Day
Multi-year backlog guidance, data center segment detail
May 2026
FERC Interconnection Reform Order
Queue acceleration = more transformer orders
Q2 2026
Infrastructure Bill Fund Disbursements
IIJA + IRA grid funding tranches rolling out
Throughout 2026
GE Vernova Analyst Day
First full-year guidance as standalone; margin trajectory
Q3 2026
DOE Grid Modernization Announcements
$20B+ in federal grid spending programs
Ongoing

The China Question: Threat or Non-Factor?

The most common bear argument against grid equipment stocks is: "Won't Chinese manufacturers flood the market with cheap transformers?" It's a legitimate question that deserves a nuanced answer.

The Bull Case for Chinese Competition

  • Chinese manufacturers (TBEA, XD Electric) produce 40%+ of global transformer output
  • Costs are 30-50% below Western manufacturers for comparable units
  • Some non-US allies (India, Southeast Asia, Middle East) are buying Chinese
  • Quality has improved significantly over the past decade

The Bear Case Against Chinese Competition

  • US anti-dumping duties of 60-75% on Chinese transformers (since 2022)
  • "Buy America" provisions in IIJA/IRA require domestic content for federal funds
  • Utilities avoid Chinese equipment for grid security (CFIUS scrutiny)
  • Chinese domestic demand is consuming all available capacity
  • Certification and testing requirements (IEEE, ANSI) take 12-18 months per model

Our assessment: Chinese competition is a real factor in the global market (non-US), but is effectively neutralized in the US and most Western allies through tariffs, domestic content requirements, and security concerns. For ETN, POWL, and GEV — which derive the majority of grid revenue from US and allied markets — Chinese competition is a minor risk. The bigger risk is actually capacity expansion by existing Western competitors (new factory announcements from Siemens Energy, Hitachi Energy), which could moderate pricing power in 2028-2030. But even then, the sheer volume of demand means all manufacturers remain fully utilized.

Key Takeaways

150+ Week Lead Times

Large power transformer lead times have exploded 5x since 2020. There is no quick fix: new factories take 3-4 years to build.

5-Company Oligopoly

Only ~5 global manufacturers can produce large power transformers. Barriers to entry are immense: $500M+ factories, decade-long workforce training.

Highest Investability

Record backlogs, book-to-bill >1.2, pricing power, and multi-year earnings visibility make grid stocks the best risk/reward in the series.

GOES Steel Bottleneck

The specialty steel needed for transformer cores has only 5 producers globally. Even with factory capacity, manufacturers can't get enough steel.

Disclaimer: This analysis is for informational and educational purposes only. It does not constitute financial advice, investment recommendation, or solicitation to buy or sell any securities. All data sourced from Wood Mackenzie, Rystad Energy, DOE Grid Deployment Office, and company filings as of February 2026. Past performance is not indicative of future results. Industrial equities carry risks including cyclical demand fluctuations, input cost volatility, and geopolitical factors. Consult a licensed financial advisor before making investment decisions.

Part 5: Copper Series Index Part 7: Natural Gas

Back to Market Watch  ·  Scarcity Alpha Series  ·  February 2026

Scarcity Alpha6/13