Regulated Utilities

Utility Strategy

Not investment advice. This page is for informational and educational purposes only. Sterling is not a registered investment advisor. Nothing here is a recommendation to buy, sell, or hold any security. Markets carry risk and you may lose money. Do your own research and consult a licensed professional before making any investment, securities purchase, or trading decision.

Thesis

Why this sector matters to investors right now. Structural, not market timing.

Regulated utilities are legal monopolies whose earnings are approximately rate base times allowed return on equity, which makes the capital plan, not the sales forecast, the earnings engine. For two decades that formula produced bond-proxy economics against roughly flat demand. The demand backdrop has now inverted: US electricity consumption set a record in 2024 and the EIA projects new records in 2025 and 2026 (EIA Short-Term Energy Outlook), data centers took about 4.4 percent of US electricity in 2023 and could take 6.7 to 12 percent by 2028 (LBNL December 2024 report for the DOE), and utility filings compiled by Grid Strategies put five-year peak demand growth forecasts near 128 GW, roughly five times the level projected two years earlier. Scarcity is showing up in market prices: PJM's capacity auction for the 2025/26 delivery year cleared at 269.92 dollars per MW-day, up from 28.92 dollars the prior year, and the following auction cleared at its price cap. Serving that load requires generation, transmission, and distribution investment at record scale (167.8 billion dollars of investor-owned electric utility capex in 2023 per EEI, with records expected to continue), and every prudently incurred dollar enters rate base and earns the allowed return for decades. That is the mechanical link from the AI buildout to regulated-utility earnings, and it is why most large names guide to 6 to 8 percent annual EPS growth anchored to rate base CAGR.

The investible question is whether utilities can convert the load inflection into earned returns without breaking the regulatory compact. The average allowed ROE granted in US electric rate cases was roughly 9.7 percent in 2024 (S&P Global Regulatory Research Associates headline figures) against equity ratios near 50 percent, and commissions grant those returns only while bills stay politically tolerable: retail electricity prices have been outpacing headline CPI (BLS), and states are responding with large-load tariffs that shift buildout risk onto hyperscale customers. Execution is the second constraint: reported large power transformer lead times of 3 to 5 years and gas turbine slots sold out into the 2029 to 2031 window gate how fast approved capex can become rate base. Wildfire liability is the tail risk the formula never priced: PG&E's 2019 bankruptcy, California's 21 billion dollar AB 1054 fund, Edison International's Eaton fire exposure, and Hawaiian Electric's roughly 2 billion dollar Maui settlement define the downside case. The sector page is organized around that chain: load growth, rate base conversion, generation shift, regulation and affordability, and the regional map on which all of it plays out.

Structural drivers

Forces that shape long-run demand and economics. Each driver is sourced.

•Load growth inflection after two flat decades. US electricity consumption set a record in 2024 with new records projected for 2025 and 2026 (EIA Short-Term Energy Outlook). Utility filings compiled by Grid Strategies (December 2024) put five-year peak demand growth forecasts at roughly 128 GW, about five times the level projected two years earlier. Sources: EIA STEO; Grid Strategies load growth reports.
•Data-center interconnection demand. Data centers consumed about 4.4 percent of US electricity in 2023 and could reach 6.7 to 12 percent by 2028 depending on buildout pace (LBNL December 2024 report for the DOE). Single-customer requests now arrive in gigawatts: Meta's Richland Parish, Louisiana campus drove Entergy's roughly 2.3 GW of new gas capacity approved by the Louisiana PSC in late 2025, and WEC serves Microsoft's Mount Pleasant, Wisconsin campus. Sources: LBNL; Entergy newsroom; company disclosures.
•Record capex converting to rate base. Investor-owned electric utilities invested a record 167.8 billion dollars in 2023, and EEI expects record spending to continue through the decade. Because earnings are approximately rate base times allowed ROE, this capex mechanically drives the 6 to 8 percent EPS growth guidance most large utilities carry. Riders, trackers, and multi-year rate plans reduce the regulatory lag between spend and recovery. Sources: EEI Financial Review; company investor materials.
•Scarcity pricing in capacity markets. PJM's 2025/26 Base Residual Auction cleared at 269.92 dollars per MW-day, up from 28.92 dollars the prior year, and the 2026/27 auction cleared at its price cap. NERC's Long-Term Reliability Assessments flag elevated resource-adequacy risk across much of North America. Capacity scarcity strengthens the case commissions face for approving new utility generation. Sources: PJM; NERC LTRA.
•Transmission supercycle with federal tailwinds. FERC Order 2023 (July 2023) reformed interconnection to first-ready-first-served cluster studies, and Order 1920 (May 2024, revised by 1920-B in April 2025) requires 20-year regional transmission planning with cost allocation. More than 2 TW of generation and storage capacity waits in interconnection queues (LBNL Queued Up). FERC-jurisdictional transmission earns formula rates with incentive adders, and AEP is the largest transmission owner in the US. Sources: FERC; LBNL Queued Up.
•Large-load tariffs de-risking the buildout. States are approving tariff structures that put minimum-take provisions, long contract terms, credit requirements, and exit fees on hyperscale customers, shielding other ratepayers from stranded-asset risk: PUCO approved AEP Ohio's data-center tariff in 2025, and Texas SB 6 (June 2025) created a new interconnection and curtailment framework for large loads. These structures convert speculative demand into contracted revenue. Sources: PUCO; Texas Legislature SB 6.
•Electrification and reshoring as the second demand leg. EV charging, heat-pump adoption, and manufacturing reshoring (chip fabs, battery plants) add load on top of data centers: Evergy serves Panasonic's De Soto, Kansas battery plant and new Google data-center load, and Alliant serves the Cedar Rapids data-center corridor. This leg is slower but broader-based than hyperscaler demand. Sources: company investor materials; EIA.
•Constructive recovery mechanisms spreading. Securitization statutes for storm, wildfire, and coal-retirement costs (Texas, Louisiana, California, Michigan, and others), formula rates, and forward test years reduce regulatory lag and de-risk recovery of the buildout. Vogtle Units 3 and 4 (commercial July 2023 and April 2024) demonstrated that even a 35 billion dollar overrun can be absorbed into rate base in a constructive jurisdiction. Sources: Southern Company; state PUC dockets.

Structural risks

Forces that could compress demand, change economics, or break the thesis.

•Affordability politics compressing returns. Retail electricity prices have been outpacing headline inflation (BLS CPI), and commissions face growing pressure to shield households from paying for data-center-driven infrastructure. The risk shows up as lower allowed ROEs, capex disallowances, or rate-case denials in stressed jurisdictions. The roughly 9.7 percent 2024 average allowed ROE (S&P RRA) is the number to watch. Sources: BLS; S&P Global RRA headline figures.
•Wildfire liability tail risk. PG&E entered bankruptcy in 2019 under inverse-condemnation liability and emerged in 2020; California's AB 1054 created a 21 billion dollar wildfire fund that Edison International is drawing against for January 2025 Eaton fire losses, for which it recorded material charges and launched a direct claims program in 2025. Hawaiian Electric's roughly 2 billion dollar share of the Maui settlement shows the risk extends beyond California, with Oregon, Texas, and the Mountain West as newer fronts. Sources: CPUC; California Wildfire Fund; Edison International; Hawaiian Electric.
•Supply chain and labor gating execution. Reported large power transformer lead times run 3 to 5 years, and gas turbine manufacturing slots are sold out into the 2029 to 2031 window per OEM disclosures. EPC labor (lineworkers, substation engineers) is as binding as equipment. Capex that cannot be executed cannot enter rate base, capping the growth algorithm regardless of demand. Sources: EEI; utility and OEM disclosures.
•Data-center demand uncertainty and double counting. Hyperscalers shop the same campus to multiple utilities, so interconnection queues and five-year forecasts likely overstate real demand; LBNL's own 2028 range (6.7 to 12 percent of US load) spans nearly 2x. If AI capex slows, utilities that built ahead of contracted load face stranded-asset fights over who pays. Sources: LBNL; Grid Strategies (which flags forecast-revision uncertainty).
•Financing drag. Record capex requires record external funding: new equity dilutes the EPS growth that rate base growth creates, and higher interest rates raise both the cost of debt and the bar for allowed ROEs to stay attractive. Regulatory lag means inflation in labor and materials is recovered only in arrears. Sources: EEI Financial Review; company 10-Ks.
•Policy whiplash on generation credits and emissions rules. The July 2025 reconciliation law (Public Law 119-21) accelerated the phase-down of wind and solar tax credits, requiring projects to be placed in service by the end of 2027 unless construction began within twelve months of enactment, raising the cost of the fastest-to-connect new supply. EPA's 2024 power plant GHG rules and the 2025 proposed repeal are both in litigation, whipsawing coal retirement and gas build schedules. Sources: Congress.gov H.R.1; EPA.
•Large-load bypass of the regulated model. Co-located and behind-the-meter arrangements let data centers contract directly with generators and skip the retail tariff; FERC's PJM co-location proceeding (order issued December 2025, triggered by the 2024 Talen-Amazon Susquehanna dispute) will shape how much of that value bypasses wires utilities. On-site fuel cells and gas generation offer a full bypass of the interconnection queue. Sources: FERC; FERC eLibrary.
•Gas LDC terminal-value debate. Building electrification, heat-pump adoption, and municipal gas bans put the long-run throughput of gas distribution systems in question even as near-term rate base grows through pipe replacement programs. The debate directly affects Atmos, NiSource's gas segment, and the gas sides of the multi-utilities. Sources: EIA; state PUC dockets.

Competitive landscape

How to think about the players. Framing along axes (pure play vs diversified, incumbent vs challenger, etc). Not stock picking.

The investible universe sorts into six archetypes, matching the sector roster categories.

1. Vertically integrated electrics (15 of 31 roster names, including Southern, Duke, AEP, Dominion, Entergy, Xcel). Own generation through distribution under one regulator. They capture the most rate base per unit of load growth because they build the power plants too, and they dominate the Southeast, Midwest, and Mountain West where data-center growth is fastest. The tradeoff is generation execution risk (fuel, construction, turbine availability) and the biggest affordability exposure since their bills carry the whole stack.

2. Wires-only T&D utilities (Exelon, Consolidated Edison, Eversource, FirstEnergy, PPL, CenterPoint, Edison International). Earn on poles and wires only after generation divestiture in restructured states. Historically lower risk and lower growth; now direct beneficiaries of interconnection-driven grid spend without commodity or generation-construction risk. The co-location debate cuts against them most directly, since a bypassed grid is their whole business.

3. Hybrid regulated plus competitive (NextEra, Sempra, PSEG). A regulated core funds an unregulated growth arm: NextEra's renewables development, Sempra's LNG infrastructure, PSEG's nuclear fleet. They offer the highest growth optionality in the sector at the cost of earnings-mix complexity and exposure to policy shifts in the competitive businesses.

4. Gas LDCs (Atmos, NiSource). The same cost-of-service model applied to gas distribution, with high near-term rate base growth from pipe replacement and the sector's clearest terminal-value question from electrification.

5. Regulated water (American Water Works, Essential Utilities). The smallest and most defensive vertical: same ratemaking mechanics, no fuel exposure, consolidation-driven growth, and premium multiples that price the safety.

6. International regulated (Fortis, National Grid). Non-US listings whose US regulated assets (ITC transmission for Fortis, downstate New York and Massachusetts wires for National Grid) give exposure to the same US ratemaking story with home-market regulatory regimes layered on top.

Cross-cutting framing: jurisdiction quality matters as much as archetype. The same capex plan earns different returns under a constructive commission (Georgia, Wisconsin, Virginia) than a contested one (California, Connecticut, New Mexico), and wildfire-state exposure (PG&E, Edison International, Xcel, Hawaiian Electric) carries a liability tail unrelated to the growth story. Load-growth exposure also varies widely: Dominion (Northern Virginia), AEP (Ohio and Texas), and Entergy (Gulf Coast) sit on the steepest demand curves, while the Northeast wires names see slower growth but steadier regulation. Merchant generators positioned for the same demand (Constellation, Vistra, Talen, NRG) are deliberately outside this sector and are covered under Nuclear.

Key metrics to watch

The operational and financial metrics that matter most in this sector. Each one names its source and update cadence.

Metric	Source	Frequency	Why it matters
Five-year utility capex plans and rate base CAGR	Company 10-Ks, investor days, EEI Financial Review	Quarterly guidance updates, annual EEI aggregate	Rate base times allowed ROE is the earnings formula, so announced capex converting to rate base is the single most direct driver of the 6 to 8 percent EPS growth guidance most large utilities carry.
Allowed ROE and rate case outcomes	S&P Global Regulatory Research Associates (headline averages), state PUC dockets	Quarterly	The 2024 average allowed ROE was roughly 9.7 percent with equity ratios near 50 percent. Whether commissions hold, raise, or compress that number under affordability pressure sets the sector's return on the entire buildout.
Retail electricity CPI versus headline CPI	BLS Consumer Price Index (electricity and utility gas series)	Monthly	The gap between electricity inflation and headline inflation is the cleanest single gauge of affordability pressure, which is the political constraint on allowed returns and capex approvals.
Five-year peak demand forecast revisions	Grid Strategies annual compilation, FERC Form 714, RTO load forecasts (PJM, ERCOT)	Annual compilations, continuous RTO updates	The roughly 128 GW five-year growth forecast (December 2024) is the demand thesis in one number. Upward revisions extend the supercycle; downward revisions signal the double-counted speculative load unwinding.
Data-center share of US electricity consumption	LBNL data-center energy reports for the DOE, EIA	Periodic (LBNL), monthly context (EIA)	The 4.4 percent (2023) to 6.7 to 12 percent (2028) trajectory is the marginal demand driver. Where actuals land inside that wide range determines how much of the announced capex is really needed.
PJM capacity auction clearing prices	PJM Base Residual Auction results	Annual (per delivery-year auction)	The jump from 28.92 to 269.92 dollars per MW-day, then to the price cap, is the clearest market signal of tightening reserve margins. Sustained cap-level clears keep the political and economic case for new regulated generation alive.
Interconnection queue volume and completion rates	LBNL Queued Up (annual), RTO queue data	Annual, with continuous RTO updates	More than 2 TW of capacity waits in queues. Whether Order 2023 reforms actually raise historically low completion rates determines how fast new supply arrives and how long scarcity pricing persists.
Wildfire liability exposure and fund capacity	CPUC and California Wildfire Fund disclosures, Edison International and Hawaiian Electric filings	Quarterly filings, event-driven	Eaton fire claims against the 21 billion dollar AB 1054 fund test whether the post-PG&E liability framework holds. Fund exhaustion or a new strict-liability state would reprice every western utility.

Catalysts and milestones

Known upcoming events that could move the sector. Dated where possible.

•PJM Base Residual Auctions through 2026 and 2027. After one record clear and one cap-level clear, the next auctions show whether scarcity pricing is structural or transient. Source: PJM capacity market results.
•FERC PJM co-location proceeding follow-through. FERC issued its order in December 2025; rehearing and implementation through 2026 will define how large loads can contract directly with generators and how much revenue bypasses the wires. Source: FERC electric transmission page; FERC eLibrary.
•Reconciliation-law credit deadlines. Wind and solar projects that did not begin construction by July 2026 must be placed in service by December 31, 2027 to capture tax credits under Public Law 119-21, compressing the build calendar for the fastest-to-connect new supply and raising post-2027 resource plan costs. Source: Congress.gov H.R.1.
•Coastal Virginia Offshore Wind completion, targeted 2026. The largest US offshore wind project (2.6 GW) finishing on Dominion's announced schedule and budget is a proof point for mega-project execution inside a rate-regulated structure. Source: Dominion Energy CVOW project page.
•Edison International Eaton fire resolution path. Direct claims program totals, CPUC proceedings, and AB 1054 fund draws through 2026 and 2027 quantify the sector's live wildfire test case. Source: Edison International investor relations; CPUC.
•Large-load tariff docket wave. Implementation of AEP Ohio's approved tariff, Texas SB 6 rulemaking at the PUCT, and copycat filings across PJM and Southeast states through 2026 determine how much hyperscaler risk transfer becomes standard. Sources: PUCO; Texas Legislature.
•NERC 2026 Long-Term Reliability Assessment (December 2026) and Grid Strategies' next load-forecast compilation (December 2026). The two annual documents that will either extend or trim the demand thesis. Sources: NERC; Grid Strategies.
•LBNL Queued Up 2026 edition and any LBNL data-center energy report update. Queue volumes and completion-rate trends test whether Order 2023 reforms are working; an updated data-center forecast narrows the 6.7 to 12 percent range. Source: LBNL.
•EPA power plant rule litigation and the proposed repeal of the 2024 GHG standards. The outcome resets coal retirement schedules and the gas build calculus embedded in integrated resource plans. Source: EPA.
•November 2026 elections. State utility commission seats and governorships (which appoint most commissioners) turn over in several high-capex jurisdictions; commission composition is the single biggest driver of allowed-return trajectories. Source: state election calendars.

What would change the view

Conditions or evidence that would invalidate the thesis or materially shift the risk picture.

•AI capex slowdown showing up in utility data: canceled interconnection agreements, hyperscalers exiting signed large-load contracts, or utilities cutting five-year forecasts. The demand leg of the thesis fails first through these filings, not through tech-sector headlines.
•Allowed ROEs trending below roughly 9.5 percent while rates stay elevated, or a major disallowance of data-center-driven capex. Either would break the assumption that commissions will let utilities earn on the buildout.
•A wildfire liability event that exhausts or materially impairs the AB 1054 fund, or a new state adopting inverse-condemnation-style strict liability. Would reprice the equity of every utility operating in fire country.
•A broad pro-co-location federal framework that lets large loads bypass retail tariffs at scale. Would shift the AI-demand value from wires utilities to merchant generators and compress the T&D archetype's growth story.
•Five-year peak forecasts revised materially downward in consecutive Grid Strategies compilations or FERC 714 filings. Would confirm the double-counting critique and mark the top of the capex cycle.
•Transformer and turbine lead times normalizing years earlier than the reported 2029 to 2031 window. Would accelerate rate base conversion (bullish for growth) but also speed new merchant supply that erodes scarcity pricing.
•Federal intervention in retail ratemaking, such as bill-relief mandates or rate caps in response to affordability politics. Would break the state-compact framework the entire earnings model rests on.
•Sustained retail electricity CPI at or below headline CPI. Would relieve the affordability constraint and support both higher allowed returns and faster capex approval, strengthening the thesis.

What we are not covering

Sub-areas, technologies, or companies we are deliberately excluding from the analysis, and why.

•Merchant and competitive generators (Constellation, Vistra, NRG, Talen). They sell into wholesale markets without ratemaking caps and are covered in the Nuclear sector. This page tracks the co-location fight from the utility side only.
•Hyperscaler buildout economics (Microsoft, Google, Amazon, Meta capex, campus siting, chip supply). Covered in the Data Centers sector; this page charts the utility-side consequences of that demand.
•Grid equipment and services suppliers (GE Vernova, Eaton, Hubbell, Quanta, transformer OEMs). Publicly traded pure plays on the same capex cycle without ratemaking caps; treated as supply chain context here, not roster members.
•Reactor-fleet economics of utility-owned nuclear (Vogtle operations, restart programs, hyperscaler nuclear PPAs). Covered in the Nuclear sector; cross-referenced where regulated utilities own fleets.
•Berkshire Hathaway Energy, TVA, Bonneville, municipal utilities, and cooperatives. Structurally important (PacifiCorp's wildfire liability, TVA's data-center load) but not directly investable; covered as context.
•Rooftop solar, DER, and grid-defection vendors (Enphase, Sunrun) and standalone renewables developers and YieldCos. Tracked only as substitution risk and supply response, not as roster members.

Sources

Primary sources cited in this analysis. Links open in a new tab.

EIA Short-Term Energy OutlookAccessed 2026-07-30
EIA Electric Power MonthlyAccessed 2026-07-30
LBNL United States Data Center Energy Usage Report (December 2024, for the DOE)Accessed 2026-07-30
Grid Strategies: load growth and utility forecast reportsAccessed 2026-07-30
PJM capacity market (Base Residual Auction results)Accessed 2026-07-30
ERCOT load and forecast dataAccessed 2026-07-30
FERC: Explainer on the Interconnection Final Rule (Order No. 2023)Accessed 2026-07-30
FERC: Electric Transmission (Order 1920/1920-B, December 2025 PJM co-location proceeding)Accessed 2026-07-30
FERC eLibrary (dockets)Accessed 2026-07-30
NERC Reliability Assessments (Long-Term Reliability Assessment)Accessed 2026-07-30
LBNL Queued Up: interconnection queue researchAccessed 2026-07-30
EEI industry data and Financial ReviewAccessed 2026-07-30
S&P Global (Regulatory Research Associates headline figures)Accessed 2026-07-30
BLS Consumer Price Index (electricity series)Accessed 2026-07-30
CPUC: wildfires and utility regulationAccessed 2026-07-30
California Wildfire Fund (AB 1054)Accessed 2026-07-30
Congress.gov: H.R.1, Public Law 119-21 (July 2025 reconciliation law)Accessed 2026-07-30
Public Utilities Commission of Ohio (AEP Ohio data-center tariff)Accessed 2026-07-30
Texas Legislature: SB 6 (89R, large-load interconnection framework)Accessed 2026-07-30
EPA: greenhouse gas standards for fossil fuel-fired power plantsAccessed 2026-07-30
American Electric Power investor relationsAccessed 2026-07-30
Entergy newsroom (Meta Louisiana project and generation approvals)Accessed 2026-07-30
Dominion Energy: Coastal Virginia Offshore WindAccessed 2026-07-30
Edison International investor relations (Eaton fire disclosures)Accessed 2026-07-30
Hawaiian Electric (Maui wildfire settlement)Accessed 2026-07-30
Southern Company investor relations (Vogtle 3 and 4)Accessed 2026-07-30

Audit trail

Record of the last review and what changed. Required on every refresh.

Last reviewed: 2026-07-30

Change log

2026-07-30Initial publication. All required components populated from the sector's Step 3 source list. Primary sources: EIA Short-Term Energy Outlook (record 2024 consumption, records projected 2025 and 2026); LBNL December 2024 data-center report for the DOE (4.4 percent of 2023 US load, 6.7 to 12 percent 2028 range); Grid Strategies December 2024 compilation (roughly 128 GW five-year forecasts, about 5x two years earlier); PJM auction results (269.92 dollars per MW-day for 2025/26 after 28.92, then a cap-level clear); EEI Financial Review (record 167.8 billion dollars 2023 IOU capex); S&P Global RRA headline figures (roughly 9.7 percent average allowed ROE 2024, equity ratios near 50 percent); FERC Orders 2023 and 1920/1920-B and the December 2025 PJM co-location order; Public Law 119-21 wind and solar credit phase-down; CPUC and California Wildfire Fund (AB 1054, 21 billion dollars); Edison International 2025 Eaton fire charges and claims program; Hawaiian Electric roughly 2 billion dollar Maui settlement share; PUCO approval of AEP Ohio's data-center tariff; Texas SB 6. All source URLs verified resolving on 2026-07-30. EEI 2024 capex actuals and RRA 2025 allowed-ROE averages were not yet citeable from public headline sources and are flagged for the next refresh.

Unresolved questions

•Re-verify the EEI 167.8 billion dollar 2023 capex figure and capture 2024 and 2025 actuals when EEI publishes updated Financial Review data.
•Update the roughly 9.7 percent average allowed ROE with 2025 full-year and 2026 year-to-date figures once S&P RRA publishes new public headline averages.
•Outcome of rehearing and implementation of FERC's December 2025 PJM co-location order, and whether a durable national co-location framework emerges.
•Whether the next PJM Base Residual Auctions clear at or near the price cap again, and whether other RTOs show equivalent scarcity pricing.
•How much of the roughly 128 GW five-year forecast is double-counted speculative data-center load, and whether the next Grid Strategies compilation revises it up or down.
•Edison International's total Eaton fire liability against AB 1054 fund capacity, and whether the fund framework survives its first large draw intact.
•Effect of the Public Law 119-21 credit phase-down on integrated resource plans after the December 2027 placed-in-service deadline, and what replaces the lost solar and wind supply in queue-constrained regions.
•Whether second-tier mid-caps (OGE, IDACORP, Portland General, Black Hills, NorthWestern) earn roster inclusion at the next refresh.

Prefer Sterling in Google

Add sterlingcharts.com as a preferred source so our charts can appear with a preferred badge in Google Top Stories, AI Mode, and AI Overviews.

Add as preferred source