Dominion Energy

Phase A — Understand the business

Lens 1 · Company Overview

Dominion Energy is a state-regulated electric and gas utility holding company headquartered in Richmond, Virginia. It is one of the largest US utilities by rate base, and — critically — it sits on top of the densest concentration of data-center load in the world (Loudoun County, "Data Center Alley"). The 10-K is a combined filing by Dominion Energy, Inc. (NYSE: D) and its largest subsidiary, Virginia Electric and Power Company ("VEPCO" / "Virginia Power").

The business is now an almost-pure regulated utility after a multi-year strategic repositioning (see Lens 8). It runs in three reportable segments:

All figures.

How it makes money: the regulated model. Dominion invests capital in generation, transmission, and distribution; the Virginia State Corporation Commission (SCC) and South Carolina PSC set the "rate base" (the asset value on which it earns) and the allowed return on equity. Earnings grow when the rate base grows — and the rate base is growing because data centers force massive grid investment. Data centers represented 28% of Virginia Power's electricity sales in 2025 (26% in 2024).

Key customers: ~2.8M VA/NC retail customers, but the marginal driver is a handful of hyperscalers (Amazon/AWS, Microsoft, Google, Meta, and colocation developers) concentrated in Loudoun County. The interconnection queue is ~70,000 MW of data-center requests — roughly 3x the system's all-time peak load — with ~10 new requests/month; ~25,000 MW expected to connect by 2031. customers.csv is empty on the shelf — concentration is therefore disclosed qualitatively, not as a `` percentage table.

Contract structure / payment terms: classic cost-of-service regulation (not take-or-pay, not recurring SaaS — it's a rate-regulated return). The important recent structural change: to protect existing ratepayers from data-center risk, Virginia Power secured a new large-load rate class (GS-5, approved Nov 2025) that makes data-center developers directly fund the infrastructure they require, and under the 2025 Biennial Review order, from January 2027 it will require a 14-year minimum contract from large-load customers. This is the single most thesis-relevant fact in the file — it converts speculative data-center demand into contracted, recoverable capital.

Lens 2 · Supply Chain

A utility's "supply chain" is fuel + equipment in → regulated delivery → end customer, plus the capital-project supply chain (which is where CVOW lives). Named stakeholders along the chain:

Upstream (fuel & inputs):

• Natural gas — purchased gas $297M (2025); supplied via interstate pipelines. Note Dominion sold its own gas-transmission/storage business to Berkshire Hathaway Energy in 2020 (Lens 8), so it is now a gas buyer, not a midstream owner.
• Nuclear fuel — for Millstone (CT, ~2 GW) and the VA fleet (North Anna, Surry). Enriched uranium supply chain (Westinghouse/Toshiba legacy relationships appear in the glossary).
• Electric fuel & energy purchases — $4,489M (2025), the largest single cost line; coal, gas, and PJM market purchases.

The CVOW capital supply chain (the chokepoint that matters):

• Monopiles / foundations, transition pieces, undersea export cables — fully-fabricated components being delivered and installed; "over 90% of costs fixed".
• Turbines — 176 turbines (Siemens Gamesa nacelles in prior disclosures); installation is the critical-path item entering the slow final stretch.
• Charybdis — Dominion's own Jones Act-compliant wind-turbine installation vessel (a single-source dependency it chose to vertically integrate to avoid the foreign-vessel bottleneck).
• Stonepeak — 50% capital partner / noncontrolling owner of CVOW (see Lens 5/9).

Tariff exposure: the 10-K flags ~$0.6B of equipment subject to tariffs on goods from Mexico/Canada/EU/other delivered Mar-2025→Mar-2026, and a ~$0.3B contingency increase for remaining construction.

Downstream: PJM Interconnection (the wholesale market/grid operator for the DOM Zone). PJM projects 5.4% average peak annual load growth over the next ten years for the PJM DOM Zone — extraordinary for a mature US utility (most are flat-to-1%).

Chokepoints / single-source dependencies: (1) the offshore-wind installation vessel and the European foundation/cable fabricators — concentrated, weather-and-logistics-exposed; (2) PJM transmission build-out approvals; (3) the SCC as the single regulatory counterparty that determines whether all this capex earns a return.

Lens 3 · Competitive Advantages (moats)

Utilities don't compete for customers — the moat is the regulated monopoly franchise plus the regulatory relationship. Dominion's specific edges:

• Geographic monopoly over the best demand zone in the country. Virginia Power is the franchised provider in the territory that is Data Center Alley. You cannot replicate the location. This is the durable moat — a structural, government-granted local monopoly sitting under exponential demand. No competitor can underprice a regulated monopoly in its own service territory.
• Scale & rate-base growth runway. A ~$65B 2026–2030 capex plan that grows earnings by construction — every dollar of approved capex becomes rate base earning an allowed ROE. Few utilities have this much sanctioned growth.
• Nuclear baseload + license extensions. Millstone, North Anna, Surry — carbon-free, dispatchable, increasingly valuable as data centers demand 24/7 firm power. ~$1.7B nuclear-life-extension capital (Rider SNA) recoverable.
• Bargaining power — who needs whom: post-GS-5, Dominion has gained bargaining power over data-center developers (they must contract 14 years and fund their own grid). Over fuel suppliers it has cost pass-through. Over the SCC it has less power — the regulator and Virginia politics are the binding constraint, not a competitor.

Where the moat is thin: it is a regulated moat, so the ROE is capped and the regulator can disallow costs (CVOW overruns, see Lens 12/13). The moat protects against competition but not against political/regulatory risk or balance-sheet risk — the two things that have actually hurt this stock. positioning.md and bottlenecks.md are missing on the shelf, so the moat read is filing- + first-principles-grounded, not commercial-layer-grounded.

Lens 4 · Segments

Revenue, operating income, and earnings by segment — every number `` (FY2025 vs FY2024 segment tables):

Note: the FY2024 segment table extract captured revenue + expense lines but not each segment's net-income-attributable row in the read window; the consolidated FY2024 figure ($2,034M) and FY2023 ($1,962M) are confirmed. Do not treat the per-segment 2024 net income as sourced — it is n/a here.

Trend & cause: The Virginia segment is accelerating (+15.7% revenue), driven by (a) data-center load growth and (b) the 2026 rate increase + new riders flowing through. Consolidated revenue +14.2% YoY, and net income attributable jumped from $2,034M → $2,998M (+47%), but beware: 2024 was depressed by one-offs and the 2025 number is flattered by ~$933M of "Other income" in Corporate (gains, tax items) and lower NCI. Diluted EPS went $2.25 (2023) → $2.33 (2024) → $3.45 (2025). Geographically the business is ~95%+ Virginia + South Carolina + Connecticut (Millstone) — domestic, single-currency (USD), no foreign exposure.

---

Phase B — Measure performance

Lens 5 · Earnings Result (latest print: Q1 2026, 10-Q filed 2026-05-01)

The headline tension of the latest print: revenue and operations are booming, but bottom-line EPS fell — because a partner is now taking its share.

All.

Read: The revenue surge (+23%) is real demand + rate relief. But two lines eat it: (1) interest expense +16.6% (the balance sheet is funding the capex build), and (2) noncontrolling interests +257% — this is Stonepeak's 50% of CVOW earnings/AFUDC now flowing out as the project nears service. So reported EPS down YoY is structurally expected, not a deterioration. On an operating-earnings basis, management framed Q1 2026 as the highest first-quarter operating EPS in recent years and reaffirmed full-year 2026 operating EPS guidance of $3.45–$3.69 (midpoint $3.57) and all 4Q25 guidance (credit, dividend, 5–7% LT growth).

Margins: electric-fuel pass-throughs inflate both revenue and expense; the cleaner signal is income-from-operations margin ~27.7% (Q1'26) vs ~30.0% (Q1'25) — compressed by higher purchased capacity and fuel, but these are largely recoverable.

Balance-sheet flags (from 10-K, FY2025): Cash just $250M (down from $310M); short-term debt $2,457M + current maturities $2,409M; total long-term debt $44,075M; gross debt ≈ $48.9B, net debt ≈ $48.7B. Operating cash flow $5,361M vs capex $12,641M → FCF ≈ −$7.3B for FY2025; Q1'26 OCF $882M vs capex $3,023M → ≈ −$2.1B. This is a utility in a heavy-build phase funded by debt + equity, not internally. Market reaction: stock ~$68 in June 2026 — roughly flat-to-up modestly post-print; the market is in "show me CVOW finishes" mode (Lens 8).

Lens 6 · Earnings Calls (sentiment trend)

transcripts/ is empty on the shelf (transcripts=0), so this lens is ``-grounded from the earnings releases and coverage of the last several calls.

What management is focused on (consistent across 4Q25 → 1Q26 calls):

1. CVOW execution — "on time, on budget, over 90% of costs fixed," first power targeted, completion end-2026. The drumbeat is deliberately reassuring because this project is the stock's overhang.
2. Data-center demand discipline — pivoting the narrative from "demand is huge" to "demand is huge and we've structured it so existing customers and our balance sheet are protected" (GS-5, 14-year contracts, ~70 GW queue with firm-queue deposit requirements).
3. Credit & dividend stability — repeatedly reaffirming credit metrics and the (flat) dividend; this is management speaking to a bruised, income-oriented shareholder base.

Tone shift over time: the multi-year arc is from "fix the balance sheet / strategic review" (2020–2023, defensive) → "execute the build / harvest the data-center decade" (2024–2026, cautiously offensive). The phrase they stopped saying is the language of strategic review, asset sales, and "rebasing." The phrase they added: "bias to the upper half of the 5–7% growth range in 2028–2030" — i.e. they are now selling an acceleration story.

Lens 7 · Comps

Peer table — US regulated electric utilities. Multiples are `` with date; where a multiple isn't sourced it is marked n/a. None are fabricated.

Sources: marketbeat/investing.com/companiesmarketcap, Nasdaq, Morningstar, all June 2026; Duke/AEP/NEE guidance from FY2026 8-Ks.

Read: Dominion trades roughly in line to slightly cheap vs the group on forward P/E (~17.9x vs ~18–19x peers) and at the high end on dividend yield (~3.9% vs 3.0–3.6%). The yield premium + the flat dividend is the market's verdict: it prices Dominion as a higher-risk, lower-quality utility (CVOW execution risk, BBB+ with a Moody's Negative outlook, a decade of dead TSR) rather than a NextEra-style premium grower. EV/EBITDA and 5-yr ROE columns are n/a (not reliably available from the searches run; do not fabricate). Dominion's implied EV ≈ ~$109B (≈$60.2B equity + ~$48.9B net debt) + balance-sheet debt ].

Lens 8 · Stock-Price Catalysts (what moves D >5%)

The five-year tape is dominated by strategic, balance-sheet, and project events — not quarterly beats. A regulated utility's earnings rarely surprise; the stock moves on the structure of the business and the credibility of the plan.

• Jul 2020 — Sale of gas transmission/storage to Berkshire Hathaway Energy (~$9.7B EV) + dividend rebase. The defining repositioning to a "pure-play state-regulated utility." Stock and yield reset hard.
• Nov 2020 — Dividend cut −33% (final cut larger than the 28% initially signaled), to ~$2.50/share, ~65% target payout. The trust-breaking event still shadowing the stock.
• 2022–2023 — Strategic / business review under CEO Robert Blue. Period of "value trap" framing; Morningstar called it a value trap with few near-term catalysts.
• Oct 2024 — Sale of 50% of CVOW to Stonepeak (de-risking the offshore-wind capital + cost overruns). The reason NCI now drags reported EPS.
• 2024–2025 — CVOW cost increase (~9%, $9.8B → $10.7B) + SCC cost-sharing settlement. Each construction update is a 5%-class catalyst because the project is the overhang.
• 2025–2026 — Data-center rate-case wins (GS-5 rate class, 2026 rate increase, 14-year contract rule). The emerging positive catalyst — converting demand into contracted rate base.

Pattern / what the market actually reacts to: capital-allocation credibility, the dividend, regulatory outcomes in Virginia, and CVOW milestones. It does not trade on EPS beats. The re-rate trigger is binary and identifiable: CVOW finishing on time/on budget + the first dividend increase in years.

---

Phase C — Judge people & books

Lens 9 · Management

Robert M. ("Bob") Blue, 57 — Chair (since Apr 2021), President & CEO (since Oct 2020). Joined Dominion in 2005; earlier Counselor/Director of Policy for the Governor of Virginia — i.e. a regulatory/political operator, not an engineer or financier, which fits a company whose binding constraint is the SCC and Richmond politics.

1. Track record (quantified): Blue owns the 2020 repositioning — sold gas transmission to Berkshire, cut the dividend, de-levered by ~$21B, narrowed to a pure regulated utility, and de-risked CVOW via the Stonepeak 50% sale. Operationally credible (CVOW is tracking on time/budget; 387+ consecutive quarterly dividends as a franchise). But the shareholder scorecard is brutal: 5-year TSR through 2025 was low single digits while the S&P 500 returned ~100%. He fixed the balance sheet by destroying the dividend and the multiple.
2. Tenure & skin in the game: ~5.5 years as CEO. Insider ownership is modest (typical professional-manager utility CEO, not a founder); insider-transactions.csv not on the shelf, so ownership % is n/a.
3. Capital allocation: mixed-to-poor on a returns basis. The SCANA acquisition (2018, inherited the failed V.C. Summer nuclear project) and the original ACP/Atlantic Coast Pipeline (cancelled 2020) were value-destructive. The recovery moves (Berkshire sale, Stonepeak, data-center discipline) are sound risk management but came at the cost of the dividend and growth optionality.
4. Red flags: CEO pay rose to $16.0M in 2025 (+19.5%) even as Virginians' bills rose and TSR lagged — and there is an active shareholder proposal to split the Chair/CEO roles and install an independent chair (NLPC-led). Combined-chair-and-CEO + lagging returns + rising pay is the classic governance-pressure setup.
5. Archetype: professional manager / regulatory technocrat. Right archetype for this stage (executing a regulated mega-build under intense political scrutiny). Wrong archetype if you wanted a capital-allocator who maximizes per-share value — the record says ratepayer-and-regulator balance has been prioritized over shareholders.

Lens 10 · Forensic Red Flags

Forensic read of the FY2025 10-K + Q1 2026 10-Q.

• ★ Restatement flag (the one to know): The 10-K cover checks "financial statements reflect the correction of an error" AND "error corrections are restatements that required a clawback recovery analysis" (§240.10D-1(b)). The substance: a deferred-income-tax misstatement on unrealized gains/losses (Corporate & Other / Contracted Energy / VA), which increased deferred income taxes by $723M in the revised 2024 balance sheet. Management assessed it as "not material" to prior periods and corrected via revision (a "little-r" revision, not a "Big-R" restatement). Verdict: not an integrity bombshell, but a control signal — a tax-accounting error large enough to trigger the clawback checkbox deserves a watch on internal-control quality.
• Cash flow vs earnings divergence: Net income $2,998M but FCF ≈ −$7.3B (FY2025), driven by $12.6B capex. Normal for a building utility, but it means earnings are non-self-funding — the story depends on continuous access to debt + equity markets at reasonable cost. With cash at $250M, liquidity rests on credit facilities ($7.0B revolver extended to 2030; $1.0B + $1.3B term facilities).
• Receivables/inventory: customer receivables $2,531M (up from $2,169M) — broadly tracks revenue growth, no obvious outrunning. Materials & supplies inventory $1,541M (rising with the build). No red flag.
• Regulatory assets: $1,380M current + $8,276M non-current = ~$9.7B of costs deferred to future recovery. Standard utility accounting, but a large balance the SCC must keep blessing — a soft risk if Virginia politics turn against cost recovery.
• Goodwill/intangibles: Goodwill $4,143M (flat, no impairment); no long-lived-asset impairment tests triggered 2023–2025.
• AFUDC / capitalized interest: with $48.9B debt and a huge construction book, a meaningful slice of earnings is non-cash AFUDC/equity-return-on-construction — quality-of-earnings caveat for CVOW.

Regulatory findings (required sub-section):

• SEC Litigation Releases / AAERs: None. Verified via SEC EDGAR EFTS (LR + AAER), search period 2021-06-22 → 2026-06-22, total_sec_findings: 0.
• 10-K Item 3 (Legal Proceedings): Primarily legacy SCANA / V.C. Summer matters (events before the 2019 SCANA combination); Dominion carries an "inconsequential" reserve and states it cannot estimate a range for some matters but flags potential materiality. No new material litigation disclosed.
• Non-SEC enforcement: No material federal enforcement action (FTC/DOJ/FDA/CFPB) against Dominion surfaced. The relevant "regulatory" pressure is rate-case / political (Virginia SCC, data-center cost allocation, the independent-chair shareholder proposal) — contested proceedings, not enforcement actions.
• Conclusion: No material regulatory or legal enforcement findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-22. The genuine forensic items are the deferred-tax revision/clawback flag and the structurally negative FCF, both disclosed.

---

Phase D — Project & stress-test

Lens 11 · Forward Projection (operating EPS, FY2026 → FY2028)

Built bottom-up from the latest actuals + management guidance. Output ``; every input labeled.

Anchors:

• FY2025 reported diluted EPS $3.45; FY2025 operating EPS midpoint guidance was ~$3.30 ex-RNG-45Z.
• FY2026 operating EPS guidance $3.45–$3.69, midpoint $3.57 (incl. $0.07 RNG 45Z).
• LT growth 5–7% off the $3.30 (2025) base, biased to upper half 2028–2030.
• Share count ~879M, with $1.8B equity issuance planned in 2026 (+3% dilution).

**

The base call (for the calibration tracker, NOT logged here per --watchlist rules): D FY2026 non-GAAP operating EPS ≥ $3.57 — p ≈ 0.60. (forecast.ts create is skipped in the watchlist loop; recorded here as the would-be forecast for Connor to log if he promotes this to a thesis.)

Lens 12 · Bull vs Bear

Bull case. Dominion is the most direct large-cap utility play on the AI/data-center electricity supercycle, sitting under the densest data-center load on earth with a government-granted monopoly. The ~70 GW interconnection queue (3x current peak) and PJM's 5.4%/yr DOM-Zone load growth give it a rate-base growth runway almost no other US utility can match — and the GS-5 rate class + 14-year contract rule de-risk that demand into contracted, recoverable capital, while Stonepeak's 50% absorbs CVOW overrun risk. The balance sheet has been repaired ($21B de-levered), the dividend is covered (~74% payout), and management is signaling acceleration to the upper half of 5–7% in 2028–2030. If CVOW finishes clean and the dividend resumes growth, a beaten-down ~17.9x / 3.9%-yield utility re-rates toward the NextEra/Duke premium — that's the bull's multiple-expansion + EPS-growth double.

Bear case (permanent-impairment risks). (1) CVOW cost blow-through past $11.3B and especially past $13.7B — the band where Dominion eats overruns with no customer recovery — turns the flagship project into a value sink (cost already moved $9.8B→$10.7B→~$11.4B). (2) Balance-sheet fragility: BBB+ with Moody's on Negative outlook, $48.9B debt, structurally negative FCF, and a ~$65B build that must be funded through volatile capital markets; a rate-shock or downgrade raises the cost of the entire growth plan. (3) Virginia political/regulatory backlash: bills are rising, data-center cost allocation is contentious, and the SCC could disallow costs or compress allowed ROE — the moat is regulatory, so the regulator is also the kill-switch.

Pre-mortem (it's late 2027, the thesis broke — what happened?): CVOW's final turbine installation slipped into a second weather season and costs pierced $13.7B; Moody's downgraded to BBB; the dividend stayed frozen a sixth year; and a populist Virginia rate backlash forced cost disallowances. The "data-center supercycle utility" re-rated down to a distressed-balance-sheet multiple.

Are multiples too high? No — ~17.9x forward is below peers; the stock is not expensive, it's de-rated. The risk isn't multiple compression from a high base, it's that the de-rating is deserved and persists.

Contrarian view (what the market refuses to see): The consensus still files Dominion under "value trap / dividend-cutter." But the 2020–2024 repositioning quietly turned it into a contracted, de-risked, data-center-levered rate-base compounder — the market is anchored on the old story and is mispricing the new one. The single fact that breaks the value-trap frame is the GS-5 / 14-year-contract structure: it makes the data-center boom bankable rather than speculative. The re-rate is gated on one clean CVOW completion.

Lens 13 · Devil's Advocate (short-seller)

Dismantling the bull.

• What structurally breaks the model: disallowed cost recovery. The entire bull case assumes the SCC lets Dominion earn on $65B of capex. Virginia just made headlines raising residential bills ~$16/mo; the political economy of "utility builds gold-plated grid for Amazon and Microsoft while grandma's bill rises" is combustible. One adverse biennial review and the ROE/rate base assumptions collapse.
• Revenue concentration / what if it shifts: 28% of Virginia Power sales are data centers — a single end-market tied to the AI-capex cycle. If hyperscaler capex normalizes or a chunk of the 70 GW queue is speculative and cancels (the 10-K itself notes it requires deposits/firm-queue rules precisely because much of it is not firm), the demand-growth story that justifies the build softens — yet the debt taken on to build it stays.
• Why the moat is weaker than bulls think: it's a capped moat. A regulated monopoly can't earn above its allowed ROE, and the allowed ROE is set by a politicized body. The "moat" is really a permission slip the regulator renews — and renews on the regulator's terms.
• Most dangerous competitor bulls underestimate: not a rival utility — it's behind-the-meter / co-located generation. If hyperscalers build their own on-site gas/nuclear/SMR power (as several have begun signaling industry-wide) to skip the interconnection queue, Dominion's marginal load-growth thesis is partially disintermediated. PJM queue delays push customers toward self-supply.
• Worst capital allocation / incentives: SCANA/V.C. Summer (inherited a nuclear fiasco), the cancelled Atlantic Coast Pipeline, a dividend cut, a combined Chair/CEO with rising pay and lagging TSR. The incentive structure (89% performance-based, 93% tied to relative TSR and operating EPS) can be hit even while absolute shareholders are crushed — "relative TSR" against a weak utility peer set is a low bar.
• What must hold for today's price: CVOW lands ≤$11.3B and on time; the SCC keeps blessing rate base + recovery; credit stays investment-grade through the build; the dividend eventually grows.
• If growth disappoints 20–30%: if the 5–7% becomes 2–3% (overruns + dilution + rate pushback), the stock is a sub-$3.60-EPS, frozen-dividend, BBB-or-lower utility — fair value likely below the current ~$68, toward the high-$50s.
• Single scenario that permanently impairs: CVOW total cost pierces $13.7B (no recovery) and a Moody's downgrade and a hostile rate order — a simultaneous project + balance-sheet + regulatory failure. Plausibility: low-to-moderate individually, low jointly — but the asymmetry is what the ~3.9% yield is paying you for.

Lens 14 · Management Questions (ordered by information value)

1. CVOW: what is your current point estimate for total project cost, and what is the probability you pierce the $13.7B no-recovery threshold? What's left un-fixed in the remaining ~10% of costs?
2. Of the ~70 GW data-center queue, how much is contractually firm (deposits posted, 14-year GS-5 terms signed) vs speculative — and what's your real expected connected MW by 2031?
3. What is your credit-metric path (FFO/debt) through the peak-capex years, and what specifically gets you off Moody's Negative outlook vs a downgrade?
4. When does the dividend grow again, and what payout ratio / coverage trigger gates that decision?
5. How much equity issuance (and dilution) does the $65B plan require 2026–2030, and at what point does dilution offset the rate-base EPS growth?
6. What is your assumed allowed ROE in the next Virginia biennial review, and how do you defend it against the rising-residential-bill political backlash?
7. What is your exposure if hyperscalers shift to behind-the-meter / co-located generation (on-site gas, SMRs) to bypass the interconnection queue?
8. The FY2025 deferred-tax revision triggered a clawback analysis — what control failure caused it, and what's been remediated?
9. How do the 14-year large-load contracts allocate stranded-asset risk if a data-center customer defaults or cancels mid-term?
10. What is Stonepeak's economic position if CVOW costs run between $11.3B and $13.7B — how much incremental capital do you carry (the 67–83% band)?
11. What is the earnings/AFUDC drag mechanics once CVOW flips from construction (AFUDC) to in-service — net EPS accretion in the first full year?
12. Tariff exposure on imported equipment (~$0.6B flagged) — how much is recoverable vs absorbed?
13. Capital-allocation philosophy now that the portfolio is "pure-play regulated": any remaining asset sales or further partnership monetizations on the table?
14. Governance: why retain the combined Chair/CEO structure given the shareholder proposal and the 5-year TSR record?
15. Beyond CVOW and data centers, what is the next leg of rate-base growth (transmission, nuclear life-extension, SMRs) and its expected timing?

---