Sempra (NYSE: SRE) — A Premier Texas Wires Asset Wearing a California-Wildfire-and-LNG Conglomerate Discount

An independent fundamental research note · Report date: 2026-06-21 · Coverage: Initiation

⚡ Claude’s Take

This block is the author’s own subjective opinion and general information only — not investment advice. The analysis that follows takes no position and carries no price target; the only view expressed anywhere is inside this fenced block.

Verdict: HOLD / accumulate-on-weakness. A genuine quality regulated-growth utility — arguably owner of the single best T&D growth asset in the United States (Oncor) — but one that has already recovered ~33% off its February-2025 lows and trades at a fair-to-full ~18x forward earnings, in the 89th percentile of its own valuation history. This is not a bargain; it is a good business at a reasonable-not-cheap price with real embedded optionality. Directional fair-value zone ~$95–115 (≈18–21x FY27E adjusted EPS of ~$5.40, crediting a partial pure-play re-rate); I would get more interested accumulating in the low-to-mid $80s (~16–17x), where the California wildfire tail and SI-close execution risk are better compensated. Not-a-short.

The market is pricing Sempra as a complicated, California-encumbered, serially-diluting conglomerate — a discount that was earned in 2025 (a botched guidance reset, a –19% single-day crash, an over-levered balance sheet, and a low-return LNG build). But the company is actively dismantling the reasons for that discount: it is selling 45% of Sempra Infrastructure to KKR/CPP for ~$10B (a deconsolidation that, by management’s own words, eliminates the need for new equity issuance across the 2025–2029 plan and removes the dilution overhang), pivoting to a ~95%-regulated pure-play, and tilting ~60% of its rate base toward Texas — where Oncor’s rate base is compounding ~18–21% a year into the ERCOT data-center supercycle, with a freshly-improved 9.75% ROE and lag-killing trackers (UTM). The variant-perception kicker is structural: Voss Capital is publicly agitating for a tax-free Oncor spin, and management’s own simplification is walking toward the same sum-of-the-parts logic. The framing is a self-help de-risking / SOTP re-rate story — not momentum, not a falling knife (beta 0.52, low-vol, sector-dominated, +26% over the last year). What keeps me at HOLD rather than BUY: the re-rate has largely happened, returns are structurally sub-WACC (the whole cohort’s curse), the California inverse-condemnation wildfire tail is real and SB254 Phase 2 is an unresolved binary, the comp plan has no return-on-capital governor, and management only sells its own stock. Conviction: medium. Flips bullish on: SI close + a rating-agency outlook upgrade + a firmed-up Oncor incremental-capex number (or a credible Oncor spin). Flips bearish on: a material California wildfire ignition or an adverse SB254 Phase 2 liability cap, or SI-close slippage that forces equity back into the plan.

Tag: “The wires are in Texas; the discount is in California.”

📈 Stock Price Action — Five-Year Event Map

Sempra’s five-year chart is a round-trip through a self-inflicted air-pocket. On a split-adjusted basis (2-for-1 split effective 2023-08-22) the stock climbed from ~$52 (early 2021) to the high-$80s by late 2024, cratered –19% in a single session on 2025-02-25 on a guidance reset, bottomed near $68 that spring, and has since clawed back to $90.69 (2026-06-18) — roughly 9% below its 52-week high of $99.75 (rs_peak –9.08%) and near the upper half of its 52-week range ($74.06–$99.75). Today the stock sits above its pre-crash level, having fully recovered the air-pocket and then some.

Cycle narrative. (1–2) Sempra spent 2021–2023 as a premium-multiple utility carrying an LNG growth option, then got compressed with the whole group as rates rose. (3) Through 2024 the Texas/Oncor data-center story re-inflated the multiple. (4) The defining event is 2025-02-25: alongside the Q4’24 print, management reset the forward plan — bigger capex, a larger embedded equity need, and worries about California regulatory lag/cost-of-capital — and the stock lost nearly a fifth of its value in a session, the worst single-day move in its modern history outside the 2020 COVID crash. (5) It based through spring 2025 as independent directors stepped in to buy near $70. (6) The September-2025 announcement that KKR/CPP would buy 45% of Sempra Infrastructure for ~$10B was the inflection — it solved the equity-dilution problem the market most feared. (7) From there a chain of constructive catalysts — Oncor’s April-2026 rate-case win, the 127 GW large-load RTP filing, visible progress toward closing SI and Ecogas, and Voss Capital’s public Oncor-spin campaign — carried the stock back to the low-$90s. The price is a fact; the attributed drivers are interpretation cross-referenced to the earnings prints, the SI 8-K, and the news feed.

1. Executive Summary

Sempra is a San Diego–based energy holding company in the middle of a deliberate identity change: from a three-legged conglomerate (California regulated utilities + ~80% of Texas’s Oncor + a global LNG/Mexican infrastructure arm) into a ~95%-regulated, Texas-tilted, pure-play utility. The transformation is being financed by selling down the non-regulated piece — 45% of Sempra Infrastructure to a KKR/CPP consortium for ~$10B — and recycling the proceeds into rate base, primarily at Oncor.

The investment tension is a classification problem. On consolidated GAAP, Sempra looks mediocre: FY2025 GAAP EPS of $2.75 (a ~60% effective tax rate and a $464M California rate-case disallowance crushed it), GAAP ROE of 5.8%, ROIC structurally below cost of capital (~5%), and chronic negative free cash flow funded by debt. But GAAP badly misrepresents the year. On the company’s own reconciliation, FY2025 adjusted EPS was $4.69 (versus $4.65 in 2024), the dividend rose for the 16th consecutive year to a $2.58 annualized rate, and — most importantly — the engine inside the conglomerate is excellent. Oncor’s rate base reached $31.5B at year-end 2025, up from $26.6B and $23.1B in the prior two years (~18–21% annual growth), its earnings are compounding ~30% a year through the 2027 guidance midpoint, and it just secured a 9.75% authorized ROE with a 43.5% equity layer and lag-reducing trackers. Texas — “ground zero” for AI/data-center power demand, with Oncor reporting a 289 GW interconnection queue (271 GW data-center) and 127 GW of “substantiated” large load — is where ~59% of the $65B five-year capital plan is directed.

Management guides FY2026 adjusted EPS to $4.80–$5.30, FY2027 to $5.10–$5.70, and a long-term growth rate of 7–9% — among the highest in the regulated-utility sector — underpinned by ~11% rate-base CAGR. At $90.69 the stock trades at ~18x the FY2026 guidance midpoint and ~17x FY2027, a fair-to-full multiple that is roughly in line with the premium-growth utility cohort (XEL, AEP) and well above its own depressed 2023–24 trough multiple — the AZI own-history composite sits in the 89th percentile.

The bull case is a re-rate: as the LNG complexity and equity-dilution overhang fall away and Texas dominates the mix, Sempra should trade like the pure-play premium-growth utility it is becoming — and the sum-of-the-parts (Voss’s tax-free Oncor spin) could surface more. The bear case is that the California inverse-condemnation wildfire tail is unquantifiable, SB254 Phase 2 is an open binary, returns remain sub-WACC, the data-center queue is heavily speculative, and the easy recovery money has already been made. This note takes no position and sets no price target; it frames the embedded expectations and the falsification tests below.

2. Business Overview

Sempra (CIK 0001032208; incorporated 1996; renamed from “Sempra Energy” to “Sempra” in May 2023) is a holding company that owns and operates regulated energy-delivery utilities and, for now, a minority-and-shrinking stake in energy infrastructure. It reports in three segments.

(1) Sempra California. Two regulated franchises serving Southern California:

• San Diego Gas & Electric (SDG&E) — a combined electric-and-gas utility serving ~3.7M people (~1.5M electric meters) across San Diego County and southern Orange County. Regulated by the California Public Utilities Commission (CPUC) for distribution and by FERC for high-voltage transmission.
• Southern California Gas (SoCalGas) — the largest natural-gas distribution utility in the United States, serving ~21M people across a 24,000-square-mile service territory in central and Southern California. CPUC-regulated; a pure gas-distribution/storage/transmission business.

California is a decoupled, forward-test-year regulatory regime (revenue is set by a multi-year General Rate Case with annual “attrition” escalators of ~4–5%, and recovery is decoupled from volumes), which is constructive on cash-flow stability but carries the defining California risk: inverse condemnation — strict liability for wildfire damage caused by utility equipment regardless of negligence.

(2) Sempra Texas Utilities. Sempra owns ~80.25% of Oncor Electric Delivery, the largest transmission-and-distribution (“wires”) utility in Texas, serving ~13M+ Texans across 120+ counties with ~140,000 miles of lines and ~4.0M points of delivery. Oncor is a pure wires monopoly: it does not own generation and bears no commodity-price risk (Texas is a deregulated-generation/retail market). It is regulated by the Public Utility Commission of Texas (PUCT) and ERCOT. The remaining ~19.75% of Oncor is held by minority owners (Texas Transmission Investment / an OMERS-led consortium), which is why Sempra carries a large non-controlling interest (NCI) on its balance sheet. Oncor is the growth engine and the crown jewel.

(3) Sempra Infrastructure (SI). Develops, builds, and operates LNG export terminals (Cameron LNG in Louisiana; Energía Costa Azul (ECA) Phase 1 on Mexico’s Pacific coast — first LNG expected June 2026; Port Arthur LNG Phases 1 and 2 in Texas) plus Mexican energy infrastructure (pipelines, the Ecogas distribution business, renewables). SI is being deconsolidated: KKR (with CPP Investments) is buying 45%, taking the consortium to 65% (ADIA holds 10%, Sempra retains ~25%). This is mostly contracted/take-or-pay infrastructure, but its returns and risk profile differ from regulated rate base, and Sempra has decided to monetize it.

Revenue & earnings mix. Consolidated FY2025 revenue was $13.7B, but revenue is a poor lens for a utility (it includes pass-through fuel/commodity costs). On adjusted earnings, the FY2025 economic split (GAAP earnings-to-common by segment, the closest disclosed proxy) was roughly: Sempra California $1,428M, Sempra Texas/Oncor (80.25%) $861M (+10% YoY), Sempra Infrastructure –$160M (a non-cash tax/FX loss, not an operating loss), and Parent –$333M. Post-transformation, Sempra’s earnings will be ~95% regulated, with Texas the largest single contributor by the end of the decade.

Recurring vs. non-recurring. Essentially all of Sempra’s go-forward earnings are recurring regulated-utility earnings (rate-base × allowed return, plus volumetric/decoupled recovery). The LNG piece has long-term contracted offtake but is being reduced to a ~25% equity-method stake. This is among the most recurring, lowest-cyclicality revenue bases in the entire equity market — the trade-off is that the return is legally capped.

3. Industry Dynamics

Structure. Regulated electric and gas delivery is the textbook regulated natural monopoly: a single set of wires/pipes per territory, prices set by a regulator to allow recovery of prudent costs plus an authorized return on invested capital (“rate base”). Barriers to entry are absolute (no one builds a duplicate grid), demand is inelastic and non-cyclical, and the regulatory compact exchanges monopoly protection for a capped return. The structural attractiveness is stability, not upside: through-the-cycle ROIC sits below the cost of capital for nearly the entire group (Sempra’s ROIC ~1.6–5%), because cost-of-service regulation is designed to prevent monopoly rents from accruing to shareholders.

The supercycle. What makes this moment unusual for utilities is the AI/data-center load boom, and Texas (ERCOT) is its epicenter. Oncor’s own filings quantify the scale: a 289 GW large-load interconnection queue (271 GW data-center-related), of which 127 GW is “substantiated” under ERCOT’s compliance criteria (up from ~38 GW a year earlier) and 102 GW was submitted in the 2026 Regional Transmission Plan (RTP). For context, ERCOT’s all-time peak demand is ~85.5 GW. Even if only a fraction of the queue materializes, it implies an enormous multi-decade transmission build — high-voltage 765-kV “superhighway” lines that Oncor is uniquely positioned to construct. Management frames this as the “incremental to the incremental”: a base $47.5B Oncor plan, plus ~$10B of identified incremental capex, plus a further leg from the 127 GW load that extends “well into the middle part of the next decade.”

Capital cycle (Marathon lens). Normally, capital flooding into an industry erodes returns. Regulation inverts this for utilities: more capital deployed = more rate base = more earnings at a guaranteed return, so the asset-growth anomaly runs in reverse — growth is value-additive so long as the regulator keeps allowing the return and the customer can afford the bill. The genuine risk the capital cycle surfaces here is affordability and political backlash: a recent headline (“AI power surge sparks political revolt against utility profits”) captures the live debate over whether data centers (not ratepayers) should bear the cost of the build. Texas has begun to address this with cost-allocation legislation (SB6) assigning large-load interconnection costs to data centers; California’s affordability politics are more fraught and feed directly into the wildfire debate.

Regulatory construct comparison. The two Sempra jurisdictions differ materially:

• Texas/PUCT (Oncor): Among the most constructive large-state regimes for a wires-only utility. The April-2026 rate case set a 9.75% ROE, 43.5% equity layer, 4.94% cost of debt, and — critically — Texas allows interim cost-recovery trackers (TCOS for transmission, DCRF for distribution, and the new UTM mechanism filable every 365 days) that collapse regulatory lag. Oncor’s earned ROE was below 8% eighteen months ago and is moving toward the authorized 9.75%. No commodity risk, no generation, no fuel.
• California/CPUC (SDG&E, SoCalGas): Decoupled and forward-looking (constructive on cash flow), but the December-2025 cost-of-capital decision cut authorized ROEs ~30bps (to SDG&E 9.93% / SoCalGas 9.78%, 52% equity), and the regime carries the wildfire-liability overhang. California is a higher-political-risk, lower-incremental-growth jurisdiction.

Verdict: structurally good (stable, monopoly, inelastic) but return-capped industry, currently enjoying an unusually favorable growth tailwind in Texas. Sempra is the purest large-cap pure-wires play on the ERCOT data-center build, which is the most attractive growth pocket in the sector — but the same growth that excites also strains balance sheets and stokes affordability politics.

4. Competitive Position

Naming the moat (Greenwald taxonomy). Sempra’s utilities possess the strongest form of competitive advantage — a legal and geographic monopoly combining supply-side scale, customer captivity (zero switching: you cannot choose another set of wires), and regulatory barriers. Market share is ~100% and perfectly stable. By Greenwald’s tests, this is as durable a franchise as exists. The crucial caveat — and it applies to the entire regulated cohort — is that the moat protects cash flow, it does not generate excess returns: cost-of-service regulation caps the allowed return near the cost of capital, so realized ROIC is structurally sub-WACC (~5% for Sempra). This is a protective moat, not a compounding one. A moat that would not let returns deteriorate if removed is a moat; here, removing the monopoly would destroy the business, so it qualifies — but the financial signature is low-and-stable, not high.

Where Sempra has a genuine within-cohort edge: Oncor’s scale and supply-chain head start. In the ERCOT build-out, the binding constraints are not capital (regulators will fund it) but physical: 765-kV transformers, transmission equipment with multi-year lead times, and skilled labor. Oncor has been pre-positioning for ~5 years — securing equipment slots, ordering 765-kV gear ahead of need, and tripling its contract-labor base — and its board granted authority years ago to contract beyond the then-current plan. Management explicitly frames this as “a competitive advantage in the industry.” A large, contiguous, multi-year backlog is also a labor magnet (workers prefer years of stable local work to region-hopping). This is a real, if modest, execution moat that lets Oncor convert the load boom into rate base faster and with less risk than a sub-scale peer. It is the most defensible part of the bull case.

LNG is not a moat — and management agrees. Sempra Infrastructure’s LNG terminals are contracted merchant infrastructure competing in a global commodity-logistics market (against Cheniere, Venture Global, Qatar, and others). It has scale and a useful dual-coast (Pacific + Atlantic) footprint, but it earns project returns, not monopoly rents, and it consumed enormous capital at returns the market did not reward inside the conglomerate. That Sempra is selling it down to recycle proceeds into higher-multiple regulated rate base is a rational multiple-arbitrage and a tacit admission that LNG was not a competitive-advantage business for Sempra’s shareholders.

Direct comparison. Versus the regulated cohort, Sempra screens as top-of-group on growth (7–9% EPS / ~11% rate-base CAGR vs. ~5–7% typical) and mid-of-group on authorized returns (9.75% Oncor / ~9.8–9.9% California vs. a 9.5–10.5% range). Its differentiation is the Texas concentration (purest ERCOT-wires exposure of any large-cap; ETR is Gulf-South integrated, AEP multi-state/multi-grid, EXC PJM with a hostile Illinois regulator, XEL upper-Midwest). Its disadvantage is the California asterisk — a wildfire-liability regime no Texas-only or Midwest peer carries.

Verdict: a durable, top-tier regulated franchise with a genuine (if narrow) execution edge at Oncor, offset by a California liability profile peers lack and by the cohort-wide reality that the moat caps rather than compounds returns.

5. Growth History and Forward Opportunities

History. Sempra’s reported revenue is noisy (commodity pass-throughs swing it: $12.9B → $14.4B → $16.7B → $13.2B → $13.7B across 2021–2025), so the right lens is rate base and adjusted EPS. Adjusted EPS has been a steady-but-unspectacular grower historically, and notably flat in 2024→2025 ($4.65 → $4.69, +0.9%) — the California Track-2 disallowance and the costs of “loading the balance sheet” for the Texas build masked the underlying Oncor growth. This flat trailing print is a fair bear flag: the 7–9% story is forward-loaded, not yet visible in the trailing numbers.

The engine: Oncor rate base. This is where the growth lives. Oncor’s rate base compounded from $23.1B (2023) → $26.6B (2024) → $31.5B (2025), roughly 18–21% annually, and its earnings are guided to grow ~30% per year through the 2027 midpoint as the new rate case, the UTM lag-reduction, and the higher equity layer lift the earned ROE from sub-8% toward 9.75%. Oncor placed $4.4B of T&D assets into service since January 2025 (the inaugural UTM filing), and ERCOT has already awarded ~$2.9B of South Dallas transmission projects largely to Oncor.

The capital plan. Sempra’s record $65B five-year capital plan splits roughly Texas $38.2B (59%) / California $23.5B (36%) / Infrastructure $3.2B (5%), supporting ~11% consolidated rate-base CAGR and ~60% of rate base in Texas by 2030. On top of the base plan, management has identified ~$9–10B of incremental opportunity (almost entirely Texas; only ~$1B LNG), and flags the 127 GW substantiated load as a source of “record capital spending at Oncor well through the middle part of the next decade” beyond even that. Few utilities have this length of visible runway.

California growth. Lower but real — SDG&E and SoCalGas file their next GRC (test years ~2028+) in Q2-2026, focused on safety/reliability, grid modernization, and affordability. SDG&E’s FERC TO6 transmission settlement (10.28% ROE, 54% equity, retroactive to 6/1/2025) is a constructive incremental positive awaiting 2H-2026 FERC approval.

LNG (de-emphasized). ECA Phase 1 reaching first LNG (June 2026) and Port Arthur Phases 1–2 (on time/budget) will add contracted cash flows, but Sempra is reducing capital allocation here and will own only ~25% of SI. LNG is now an optionality/cash-recycling story, not a growth pillar.

Verdict: high-quality, long-duration growth — among the best in the sector — but the quality sits almost entirely in Texas/Oncor, the trailing print is flat, and delivery depends on executing a very large capex program at acceptable customer cost.

6. Financial Quality

Earnings quality — GAAP is not the right number for 2025. FY2025 GAAP EPS of $2.75 understates the business badly. The bridge from GAAP earnings-to-common of $1,796M to adjusted earnings of $3,066M ($4.69/sh) comprises ~$1,273M of legitimate add-backs: +$512M of net tax from classifying SI Partners and Ecogas as held-for-sale (a $693M deferred-tax charge on SI’s outside-basis difference, partly offset by a $191M valuation-allowance benefit — non-cash, one-time); +$457M for the SDG&E 2024-GRC Track-2 regulatory disallowance ($651M pretax/$464M after-tax); +$180M of Mexico FX/inflation; +$78M of foreign-tax-credit valuation allowance; and ~$43M of mark-to-market on derivatives. None of these are recurring operating items, and the held-for-sale tax charge in particular is an artifact of the transformation, not a deterioration. Use $4.69, not $2.75. (Symmetrically, FY2024 GAAP $4.42 ≈ adjusted $4.65 — there was no large one-time GAAP gain inflating 2024; 2025 is the distorted year.)

Cash flow. Operating cash flow was $4,565M in FY2025; OCF/NI ran ~2.2x (depreciation and deferred taxes are large for a capital-intensive regulated business), so cash conversion is healthy and earnings are not flattered by accruals. The catch is the denominator on the other side: capex was $12,627M ($10,612M property, plant & equipment + $2,015M equity contribution into Oncor), against which OCF leaves free cash flow of roughly –$8.1B. Like every growth utility (PPL, PCG, AEP), Sempra is structurally FCF-negative and funds the gap with external capital. This is normal and not a red flag per se — but it makes the cost and form of that external capital (debt vs. equity) the central financial-quality question.

Margins / returns. EBITDA margin ~41%, operating margin ~22% — typical regulated economics. ROIC ~1.6–5% and normalized GAAP ROE ~9–11% (the 5.8% 2025 print is the disallowance/tax artifact). These returns are sub-WACC, as for the cohort; the value is created by growing the capital base at the allowed return, not by earning excess returns on it.

Balance sheet & leverage. Total assets ~$111B; total debt ~$35B; net debt ~$35B. Consolidated net-debt/EBITDA screens at a heavy ~6.2x — but this materially overstates holdco leverage because it consolidates (a) ring-fenced Oncor debt (Oncor is structurally separated, with its own balance sheet and dividend restrictions) and (b) non-recourse SI project debt that disappears on deconsolidation. The negative outlook from the rating agencies (S&P/Moody’s) reflects the stretch from the Texas build; the cure is the SI deconsolidation plus parent-debt paydown, with threshold improvement expected ~six months after close (late-2026/early-2027). Sempra redeemed its $900M Series C preferred in October 2025, simplifying the capital stack. The held-for-sale reclassification (~$31B moved from non-current PP&E to current assets, with a corresponding NCI/CRNCI build) is an accounting consequence of the pending sales, not an impairment.

The non-controlling-interest complexity. A genuine analytical wrinkle: ~$235–238M of FY2025 earnings, and a large slice of the balance sheet, belong to minority owners (Oncor’s ~19.75% and SI’s partners). Consolidated rate base therefore overstates Sempra’s economic share by ~20% — Oncor’s $31.5B rate base is ~$25B on a Sempra-economic basis. Any sum-of-the-parts must work in economic, not consolidated, terms.

Verdict: economics improve with scale only in the regulated sense — more rate base at the allowed return. Earnings quality is sound once GAAP noise is removed; the balance sheet is stretched but on a credible, self-funded de-leveraging path. Adjusted EPS is the honest number, and it is barely growing on a trailing basis — the thesis is entirely about the forward rate-base trajectory.

7. Capital Allocation

The centerpiece: capital recycling via the SI sell-down. On 2025-09-22 Sempra agreed to sell 45% of Sempra Infrastructure to a KKR-led consortium (with CPP Investments) for ~$10B cash, implying SI equity value of ~$22.2B and enterprise value of ~$31.7B. Post-deal ownership is KKR-consortium ~65% / ADIA ~10% / Sempra ~25%, and SI deconsolidates. Cash is phased (~47% at close, ~41% by end-2027, the balance ~2033), with ~$337.5M of KKR fees and a ~$340M development credit netted. The strategic prize is what it removes: by management’s own statement it “eliminates the need for equity issuances in the 2025–2029 plan” — killing the dilution overhang that drove the February-2025 crash — while delivering parent-debt-paydown firepower. A parallel ~$500M Ecogas Mexico sale furthers the simplification. The recycling arc is clean rather than opportunistic: KKR first bought 20% of SI in 2021 at ~$25.2B EV and ADIA 10% at ~$26.5B EV, so the implied valuation has crept only ~20–25% in four years — a fair monetization that funds higher-return regulated rate base, not a windfall.

M&A history (Marathon lens). The defining acquisition was 2018’s purchase of ~80% of Oncor for ~$9.45B out of the Energy Future Holdings bankruptcy — in hindsight the best capital-allocation decision in Sempra’s modern history, and now the very asset Voss wants spun. The 2021 IEnova (Infraestructura Energética Nova) take-private and the broader LNG/Mexican build were more questionable — capital deployed at returns the public market never credited, which is precisely the value now being unwound through the SI sale. Net: above-average on Oncor, mixed on infrastructure — and the current strategy is, encouragingly, a correction of the latter.

Dividend. The dividend rose to a $2.58 annualized rate in 2026 — the 16th consecutive annual increase (aristocrat-class durability), from ~$2.29 in 2021, a slow ~2.4% CAGR. Forward guidance is modest (~2–4%). Crucially, the February-2025 –19% crash did not threaten the dividend; payout on adjusted EPS is a comfortable ~55%. Sempra returned ~$10B to shareholders over 2020–25, overwhelmingly via dividends (FY2025 buybacks were a token $26M — buybacks are not part of the story).

Equity issuance / dilution. Share count rose modestly from ~626M (2021) to ~654M (2025). FY2025 net common equity issued was only +$32M (versus $1.2B in 2024); the current vehicle is a November-2024 ATM with forward sales (~$268M unsettled). The SI sale’s whole point is to make further equity issuance unnecessary through 2029 — if it closes as planned, the dilution risk that defined 2025 is resolved.

Executive compensation — a “governor” problem. Here the verdict turns critical. The annual bonus is ~80% driven by absolute “ABP Earnings,” and the LTIP is split one-third relative TSR / one-third adjusted-EPS CAGR / one-third options-and-RSUs. There is no ROE, ROIC, or any return-on-capital metric anywhere in the incentive design. As with PCG, PPL, and AEP, this rewards growing earnings and rate base without an explicit hurdle for the return on the capital deployed — a structural incentive to build, regardless of incremental returns, which is exactly the wrong governor for a sub-WACC business in a capex supercycle. CEO Jeffrey Martin is combined Chair-and-CEO (FY2025 realized comp ~$16.8M); say-on-pay has been routine.

Insider behavior. A 435-Form-4 sweep shows zero open-market purchases by any named executive officer (Martin, Sedgwick, Bird) in five years — management only sells shares or receives grants. The only open-market buys are by seven independent directors, ~$2.0M total, clustered in the two crash windows (March-2025 near $70 and March-2026 near $93; director Richard Mark most active). Net: a mildly positive board signal, neutralized by management’s uniform all-sell pattern.

Verdict: management is, on balance, allocating capital intelligently right now — the pivot to regulated, the SI monetization, and the dilution-overhang removal are genuinely value-additive and self-correcting prior over-build. But the incentive design lacks a return-on-capital governor, and management’s own buying behavior signals no personal conviction at these prices.

8. Changes and Headwinds — Last Two Years

Strategic transformation (positive). The single biggest change is the deliberate pivot from conglomerate to pure-play regulated utility: the SI 45%-to-KKR sale (Sept-2025), the Ecogas sale, the redemption of the Series C preferred, and the explicit ~60%-Texas-by-2030 rate-base tilt. This is the de-risking that the equity market is gradually re-rating.

The February-2025 guidance reset (the negative catalyst, now digested). The –19% crash on 2025-02-25 reflected a forward plan with heavier capex, a larger embedded equity need, and California cost-of-capital/regulatory-lag concerns. The subsequent year has been a methodical walk-back of those fears (SI sale removes equity need; Oncor rate case + UTM fix the lag).

Oncor rate case win (positive, April-2026). PUCT Docket 58306: 9.75% ROE, 43.5% equity, 4.94% cost of debt, ~+$560M/+8.7% revenue, plus a surcharge for Jan–Jun 2026 — materially improving Oncor’s earned-ROE trajectory and credit metrics during the build.

California cost-of-capital cut (mild negative, Dec-2025). Authorized ROEs trimmed ~30bps (SDG&E 9.93% / SoCalGas 9.78%) — a reminder that California is the lower-return, higher-friction jurisdiction.

The wildfire-liability overhang (the defining California risk). California’s inverse-condemnation doctrine imposes strict liability on utilities for wildfire damage from their equipment regardless of fault. AB1054 created a ~$21B state wildfire fund and caps a prudent utility’s liability at ~20% of T&D equity rate base (provided it holds a valid safety certification), but the fund can be exhausted and the cap then ceases. SB254 (2025) added ~$18B of incremental liquidity to the fund — but SB254 Phase 2, which would address a more durable/harder liability cap and recovery framework, is the unresolved binary: management expresses “reasonable confidence” of progress in the current legislative session, supported by an April-2026 California Earthquake Authority resiliency study that framed wildfire as a “whole-of-society problem.” Resolution would remove a major discount; an adverse outcome (or a major SDG&E ignition before resolution) would reprice the California stub sharply. SDG&E is widely regarded as the best wildfire-mitigated utility in the state (extensive hardening, the lowest recent ignition record), which materially lowers — but does not eliminate — the tail.

Activist pressure (Voss Capital). In May-2026 Voss Capital (~2M shares, ~0.3% — a voice-only stake) went public urging a tax-free spin of Oncor to cure the conglomerate discount and isolate California wildfire risk, claiming a standalone Oncor worth ~$78B by 2028 (aggressive). Sempra has not commented. Material because it legitimizes the SOTP thesis and rhymes with management’s own simplification — though an Oncor spin is structurally hard (the ring-fence, PUCT approval, minority owners, tax).

Verdict: the changes net positive — a credible, self-funded de-risking that is repairing the 2025 damage — with the California wildfire binary the one genuinely unquantifiable headwind, partly mitigated but not removed.

9. Risk Analysis (Risk Matrix)

The dominant, differentiated risk is California wildfire (low-probability, high-impact, unquantifiable) layered on top of SI-close execution (the linchpin of the whole de-risking narrative). Everything else is shared with the cohort or self-mitigating.

10. Valuation Discussion (Embedded Expectations)

This section sets no price target and makes no recommendation. It frames what the current ~$90.69 price embeds.

The multiple. On adjusted EPS, Sempra trades at:

• ~19.3x trailing FY2025 ($4.69),
• ~18.0x the FY2026 guidance midpoint (~$5.05), and
• ~16.8x the FY2027 midpoint (~$5.40).

(The GAAP P/E of ~28–33x and the AZI P/E percentile are distorted by the 2025 tax/disallowance noise — ignore them; read P/B 1.84x and the adjusted-EPS multiple instead.) EV/EBITDA is ~19x (TTM), and dividend yield is ~2.84%.

Own-history context. The AZI own-history composite valuation sits in the 89th percentile (P/B 88.8th, P/S 95th) — i.e., Sempra is expensive relative to its own recent history. But that history is dominated by the depressed 2023–2025 window; the ~18x forward multiple is not stretched in absolute terms for a 7–9% grower.

Cohort context. Premium-growth regulated utilities (XEL, AEP, WEC) trade ~17–20x forward; slower names (D, EXC, ETR before its recent re-rate) trade ~13–17x. At ~18x with the highest growth guide in the group (7–9%) and a top-tier rate-base CAGR (~11%), Sempra is priced roughly in line with the premium cohort — neither cheaply nor expensively — which is the crux: the market is no longer applying a conglomerate discount as deep as it did in early 2025, but it is not yet paying a clean pure-play premium either.

What the price embeds. At ~18x forward for a credibly-guided 7–9% grower with a ~2.8% yield, the market is underwriting:

1. Delivery of the 7–9% EPS growth (≈10–12% total return with the dividend) — i.e., that the $65B plan is executed at the allowed returns and rate base compounds ~11%.
2. SI closes and the balance sheet de-risks — no equity dilution through 2029, rating outlook stabilizes.
3. The California wildfire tail does not crystallize — no catastrophic ignition and a constructive (or at least non-destructive) SB254 Phase 2.

What it does not yet fully embed is a pure-play re-rate or an Oncor SOTP unlock — that is the upside option.

Scenario frame (illustrative, not targets):

• Bear (~$70–80): California wildfire scare or adverse SB254 Phase 2; SI close slips and equity returns to the plan; multiple compresses to ~14–15x on FY26 EPS as the conglomerate discount re-widens. This roughly revisits the 2025 trough.
• Base (~$90–105): 7–9% growth delivered, SI closes, rating stabilizes; multiple holds ~17–19x FY27; total return tracks earnings growth plus yield. The stock compounds with the rate base.
• Bull (~$110–125+): the market re-rates Sempra as a clean ~95%-regulated, Texas-tilted premium-growth utility (low-20s× FY27), and/or the Voss-style Oncor SOTP narrative gains traction. Voss’s ~$78B-Oncor-by-2028 math is aggressive, but even a partial pure-play premium gets the equity into the $110s.

Sum-of-the-parts (the variant lever). On an economic-interest basis, Oncor (~$25B Sempra-economic rate base, growing 18–21%, at a premium pure-wires multiple), plus the California utilities, plus a ~25% SI stake (marked at the KKR ~$22.2B equity valuation ≈ ~$5.5B to Sempra), less parent debt, plausibly exceeds the current ~$58–63B equity value — which is the entire activist thesis. The discount is real; the question is whether and how it closes (re-rate vs. structural spin), and a spin is hard.

Verdict: fairly-to-fully valued on near-term earnings, with the re-rate/SOTP optionality unpriced. The market has moved from punishing the conglomerate to fairly pricing the transition; the remaining upside is a quality re-rate, the remaining downside is the California binary.

11. Variant Perception

Consensus view. Sempra is a recovering, above-average-growth regulated utility de-risking its balance sheet via the KKR/SI sale, with a strong Texas/Oncor growth engine and a manageable California wildfire overhang; sell-side is broadly constructive on the 7–9% growth and the simplification, and the stock has re-rated accordingly back to the low-$90s.

The strongest bull case. This is the best growth asset in the regulated-utility universe (Oncor’s 18–21% rate-base CAGR into the ERCOT data-center supercycle) trapped inside a holding company that the market still partly values as a complicated, California-encumbered conglomerate. As management completes the pivot — SI deconsolidated, equity-dilution overhang gone, ~60% Texas by 2030, rating outlook repaired — the multiple should migrate toward the clean pure-play premium cohort, and the Voss-style sum-of-the-parts could surface a further leg. You are buying the highest-growth utility in the group at an in-line multiple, with a re-rate option attached.

The strongest bear case. The easy money is made — the stock has already recovered ~33% off the lows, sits at the 89th percentile of its own valuation, and trades in line with premium peers, so you are not getting the asset cheaply. Underneath, returns are structurally sub-WACC (~5% ROIC), trailing adjusted EPS was flat in 2025, the comp plan has no return governor (incentivizing build-for-build’s-sake in a capex supercycle), management never buys its own stock, the data-center queue is mostly speculative, and the one genuinely differentiated risk — California inverse-condemnation wildfire liability with an unresolved SB254 Phase 2 — is unquantifiable and could reprice the whole equity in a single ignition. If the pure-play re-rate doesn’t come, this is a ~7–9% total-return grind with a fat left tail.

The 3–5 assumptions that matter most:

1. SI closes on terms (Q2/Q3 2026) — the linchpin of the entire de-risking and rating narrative.
2. The California wildfire tail stays dormant and SB254 Phase 2 lands constructively.
3. Oncor executes the $65B build at the allowed returns without cost/labor/supply blowups, and the data-center load converts to rate base.
4. The market grants a pure-play re-rate (the difference between a base ~$90–105 and a bull ~$110–125+).
5. Rates/cost-of-capital stay benign enough that authorized ROEs hold and FCF-negative financing stays cheap.

Factor-positioning read (Public factor-model data). Beta 0.525 (low), Utilities-sector loading 0.98 (R² 0.58 — overwhelmingly a sector/rate vehicle, little idiosyncratic factor signature). Risk-adjusted track record has inflected up: 1-year return +25.6% (Sharpe 1.20), 6-month +10.7%, but the last 3 months –4.4% (a pause after the recovery). rs_peak –9.08% (near, not far from, its high); rs_12m +26.16%. This is a recovered, low-volatility, sector-driven quality utility — categorically NOT a falling knife and NOT a crowded momentum trade. The tape says the de-rating is over and the recovery is mature, which supports the “fairly-valued-with-optionality” framing in Claude’s Take rather than either a deep-value-knife or a chase-the-momentum read. Factor-similar names (ETR, D, WEC, AEE, NI) are the right cohort cross-check.

12. Fact vs. Interpretation Table

13. Open Questions

1. SI close mechanics & timing — will the Cameron-partner and Japanese-ECA consents land on schedule, and could terms be re-traded? The whole de-risking narrative hinges here.
2. SB254 Phase 2 — what does a durable California wildfire liability cap/recovery framework actually look like, and does it pass this session? Binary for the California stub.
3. Oncor incremental capex firm-up — how much of the ~$10B “I-squared” and the 127 GW load converts to a hard plan, and when (management flagged a Q2-2026 update)?
4. Does the pure-play re-rate actually arrive, or does Sempra stay range-bound as a “good utility at a fair price” — and would an Oncor spin (Voss) ever be structurally feasible given the ring-fence, minority owners, PUCT, and tax?
5. Earned-vs-authorized ROE convergence — does the UTM/tracker machinery actually lift Oncor’s earned ROE to ~9.75% on the timeline, or does the heavy build keep lag alive?
6. Rating-agency outcome — does the negative outlook resolve to stable/positive within close-plus-six-months, and what FFO/debt threshold do the agencies set post-deconsolidation?
7. What is the true Sempra-economic SOTP once NCI, parent debt, and the ~25% SI stake are netted — does it materially exceed the current equity value, and by how much?

14. What Must Be True

Bull case — what must be true:

• SI closes in 2026 on (roughly) announced terms; equity issuance stays off the table through 2029; the rating outlook moves to stable/positive.
• Oncor delivers ~18–21% rate-base growth and converts a meaningful chunk of the data-center queue, lifting earned ROE toward 9.75% and sustaining ~30%/yr Oncor earnings growth to 2027.
• The California wildfire tail stays dormant and SB254 Phase 2 is constructive.
• The market re-rates Sempra toward a clean pure-play premium-growth multiple (low-20s× FY27).
• Falsification test: if, by year-end 2026, SI has not closed or equity issuance reappears in the plan or the rating outlook is still negative, the de-risking thesis is broken and the conglomerate discount is deserved. A failure of adjusted EPS to grow toward the $4.80–5.30 (2026) / $5.10–5.70 (2027) guide would also break it.

Bear case — what must be true:

• The recovery is exhausted at ~18x and the pure-play re-rate never comes; total return reverts to the ~7–9% earnings grind.
• A California wildfire event or an adverse SB254 Phase 2 crystallizes liability and reprices the equity.
• The capex supercycle strains affordability/credit, authorized returns drift lower, and the no-return-governor comp plan drives value-neutral build.
• Falsification test: if SI closes cleanly and the rating outlook turns positive and Oncor’s incremental capex firms up materially (or a credible Oncor-spin path emerges) and SB254 Phase 2 lands constructively — all plausibly within 12 months — the bear’s “fairly-valued grind / fat-left-tail” framing is falsified and the stock should re-rate.

The two cases are unusually time-boxed: the SI close, the rating decision, the SB254 session, and the Oncor capex update all resolve within roughly the next 12 months, which makes the next four quarters genuinely thesis-determining.

15. Source Appendix

Primary sources: Sempra FY2021–FY2025 Forms 10-K (CIK 0001032208), Q1-2026 Form 10-Q, Q1-2026 earnings call transcript (2026-05-07), the SI/KKR transaction 8-K (Sept-2025), DEF 14A (executive compensation), the Form 4 corpus (insider transactions), PUCT Docket 58306 (Oncor rate case), CPUC cost-of-capital decision (Dec-2025), AB1054/SB254 (California wildfire), and quantitative feeds (ROIC.ai fundamentals/EV, AZI price & valuation-percentile data, Public factor-model data factor model). Peer cross-reads drew on public filings and disclosures for ETR, D, AEP, XEL, EXC, SO, DUK, PCG, and PPL.

This note is independent fundamental research and general information only — not investment advice. The body takes no investment position and contains no price target; the only position expressed anywhere is the clearly-fenced “Claude’s Take” block, the author’s own view.

Appendix A — Diligence Questionnaire — Sempra (NYSE: SRE)

Report date 2026-06-21. Labels: F = Fact, I = Interpretation, A = Assumption.

General

What thoughtful questions have other investors asked about this company?

• Will the SI/KKR sale actually close on terms, and does it really remove the equity-dilution need through 2029? (F: management asserts yes.)
• Is the Texas data-center growth real or a speculative queue? (F: 289 GW queue, 127 GW “substantiated”; I: base $47.5B Oncor plan is “largely indifferent” to the speculative tail.)
• What is California wildfire actually worth as a liability, and does SB254 Phase 2 cap it? (I: unquantifiable; binary on the legislative session.)
• Should Sempra spin Oncor to close the conglomerate discount? (F: Voss Capital is publicly pushing exactly this.)
• Is the dividend safe given negative FCF and a stretched balance sheet? (F: yes — 16th consecutive increase, ~55% payout on adjusted EPS, not threatened by the 2025 crash.)

Cyclicality & Earnings Nature

Are earnings at a cyclical high or low? I: Neither in the classic sense — regulated utility earnings are non-cyclical (rate base × allowed return). Trailing adjusted EPS is actually depressed/flat ($4.65→$4.69) by the California disallowance and balance-sheet build; the growth is forward-loaded. So earnings are at a transitional trough-to-inflection, not a cyclical peak.

Driven by external environment or internal actions? F/I: Internal — rate-base deployment, rate-case outcomes, and the strategic pivot. The external macro inputs (interest rates, AI-power demand) set the backdrop but the earnings are administratively determined.

How stable are revenues? F: Among the most stable in the market — decoupled California recovery + Texas wires (no commodity risk). Reported revenue swings on commodity pass-throughs but margin/earnings are stable.

Outlook for products/services / how big is the market? F: Electricity/gas delivery demand is growing (electrification + data centers); the Texas T&D build is a multi-decade, multi-$10B opportunity. Growing, domestic, with the LNG piece (international) being divested.

Business Quality & Competitive Moat

More or less competitive industry? I: Non-competitive by design (legal monopoly). The “competition” is regulatory and political (affordability, cost allocation), not market share.

How profitable is the business (ROIC, ROE)? F: ROIC ~1.6–5% (sub-WACC); normalized GAAP ROE ~9–11% (2025’s 5.8% is a tax/disallowance artifact). The moat caps returns near cost of capital — value comes from growing the capital base.

How profitable is the industry / barriers to entry? F: Absolute barriers (no duplicate grid). Industry-wide returns are regulator-capped; profitability is stable-and-modest, not excess.

Can the business be easily understood? I: The utilities yes; the holding-company structure (large NCI from Oncor’s ~19.75% minority + SI’s KKR/ADIA partners, plus the in-flight deconsolidation) is genuinely complex and is part of the discount.

Undermined by foreign low-cost labor? F: No — physical local-monopoly infrastructure.

Do brands matter? F: No — captive customers, zero choice of provider.

Nature of competition / switching costs? F: No switching (monopoly). The relevant “competition” is for capital, labor, and equipment in the build (where Oncor’s scale/supply-chain head start is a genuine edge), and political competition over who pays.

Financial Condition & Balance Sheet

Assets not fully on the balance sheet? I: The economic value of Oncor’s growth franchise and the regulated rate base’s premium-multiple worth (the SOTP gap) is not visible in book value (P/B 1.84x).

Off-balance-sheet liabilities? F: The California inverse-condemnation wildfire exposure is a contingent liability not booked until an event; AB1054/SB254 fund participation. Large NCI obligations. Non-recourse SI project debt (leaves on deconsolidation).

How conservative is the accounting? I: Standard regulated-utility accounting (regulatory assets/liabilities, deferred taxes). The 2025 GAAP distortion (held-for-sale tax, disallowance) is disclosed and reconciled — not aggressive. Adjusted EPS is the honest run-rate.

How CapEx-hungry? F: Extremely — $65B five-year plan, capex ~$12.6B in FY2025 vs. ~$4.6B OCF → ~–$8.1B FCF. This is the defining financial characteristic; it requires continuous external capital.

Capital Allocation & Management

How much FCF, and how used? F: FCF is structurally negative (growth utility). “Discretionary” cash (dividend) is funded from the regulated cash flows; the growth is debt- and (historically) equity-funded. Philosophy: deploy maximum capital into allowed-return rate base, return a steady dividend, recycle non-core assets.

Significant acquisitions recently? F: The strategy is now divestiture — SI 45%-to-KKR (~$10B), Ecogas (~$500M). The defining past acquisition was ~80% of Oncor (2018, ~$9.45B) — the best deal; the LNG/IEnova build is the value now being unwound.

Buying back shares? F: No (token $26M FY2025). Not part of the story.

Issuing shares to insiders / dilution? F: Modest dilution (~626M→654M shares 2021–25); the SI sale is designed to halt further equity issuance through 2029.

Compensation policy? F/I: A governor problem — bonus on absolute “ABP Earnings” (80%), LTIP one-third relative TSR / one-third adjusted-EPS CAGR / one-third options-RSUs. No ROE/ROIC metric — rewards growth without a return hurdle (like PCG/PPL/AEP).

Motivations of management? I: Combined Chair/CEO Martin; comp incentivizes EPS/rate-base growth. Zero NEO open-market buys in 5 years (management only sells) signals no personal conviction at current prices; seven directors bought ~$2.0M in the two crash windows (mildly positive board signal).

Valuation & Market Data

ADR, MLP, or K-1 issuer? F: No — a standard US C-corp common stock (NYSE: SRE), 1099 not K-1.

Dividend policy? F: $2.58 annualized (2026), 16th consecutive annual increase, ~2.4% CAGR, ~2.84% yield, ~55% payout on adjusted EPS. ~2–4% forward growth guidance.

How profitable / net income vs. cash from operations? F: OCF/NI ~2.2x (cash flow well exceeds GAAP net income due to D&A and deferred taxes) — earnings are not flattered by accruals. The divergence runs the healthy direction.

Risks & Downside

What would cause the stock to decline? I: A California wildfire ignition or adverse SB254 Phase 2; SI-close failure/re-trade forcing equity back into the plan; a rating downgrade; a broad rate-driven utility de-rate; failure to deliver the 7–9% growth.

Risk of catastrophic loss? I: Low-probability but real — a major SDG&E-caused wildfire under inverse condemnation, if it exhausted the AB1054 cap/fund, is the genuine tail. SDG&E is the best-mitigated California utility, which lowers (not removes) it.

Chance of total loss? A: Very low — a diversified, ~95%-regulated, investment-grade utility with a fortress monopoly franchise and a 16-year dividend record. Total loss would require an unprecedented uninsured catastrophe plus capital-market closure.

Recent News & Events

Has the business environment changed recently? F: Materially and favorably on de-risking — SI 45% sale (Sept-2025), Oncor rate-case win (Apr-2026, 9.75% ROE), 127 GW load filing, ECA first-LNG (June-2026), Voss activist campaign (May-2026). The February-2025 guidance reset (–19% day) was the negative catalyst, now largely digested.

Significant acquisitions / accounting changes? F: No acquisitions; the in-flight divestitures drove the held-for-sale reclassification (an accounting consequence, not a policy change). Series C preferred redeemed Oct-2025.

Recent changes — new markets, facilities, management? F: Strategic pivot to pure-play utility / ~60% Texas rate base by 2030; ECA LNG Phase 1 commissioning; CEO Martin and CFO Sedgwick in place; Oncor’s Allen Nye and SI’s Justin Bird leading the segments.

Appendix B — Source Appendix — Sempra (NYSE: SRE)

Report date 2026-06-21. Primary sources first. All quantitative figures reconciled to filings where material.

Primary — SEC filings (CIK 0001032208)

Primary — earnings call transcript (ROIC.ai)

Primary — regulatory

Quantitative feeds (third-party; reconciled to filings)

Secondary / news

Peer cross-reads

Public filings and disclosures for ETR (Entergy), D (Dominion), AEP, XEL, EXC (Exelon), SO (Southern), DUK (Duke), PCG (PG&E), and PPL — used for cohort growth/ROE/valuation framing, regulatory-construct comparison, and data-center-supercycle context.