Sempra

Phase A — Understand the business

Lens 1 · Company Overview

Sempra (NYSE: SRE) is a San Diego-headquartered energy-infrastructure holding company whose principal businesses are regulated electric and gas utilities in California and Texas, plus a (being-divested) infrastructure arm focused on North American LNG export and Mexican energy. It reports three segments:

• Sempra California — the wholly-owned regulated utilities SDG&E (electric to San Diego + southern Orange counties; gas to San Diego County) and SoCalGas (the largest US gas distribution utility, serving most of Southern California). Earnings here are a function of CPUC-authorized rate base × allowed ROE; commodity cost is a pass-through. FY2025 earnings $1,428M.
• Sempra Texas Utilities — an 80.25% economic interest in Oncor, Texas's largest regulated T&D wires utility, held as an equity-method investment (Oncor is ring-fenced and not consolidated; Sempra recognizes its share as "equity earnings"). FY2025 earnings $861M; carrying value of the Oncor Holdings investment $17,472M. This is the structural growth engine.
• Sempra Infrastructure (SI Partners) — LNG (Cameron LNG JV, ECA LNG, Port Arthur LNG), Mexican pipelines/Ecogas, renewables. As of Sept 2025 this is "held for sale": $31,024M assets / $11,704M liabilities reclassified. FY2025 booked a $(160)M loss — but that was a one-time tax artifact, not operations (see Lens 5).

Business model in plain terms: Sempra is overwhelmingly a rate-base compounder. It spends capital building wires and pipes; regulators (CPUC, PUCT/Texas, FERC) let it earn an authorized return (~10% ROE) on that invested capital and recover costs in customer rates. Growth = capex that grows rate base. Revenue ($13.7B FY2025) is a poor headline metric because the commodity portion is a pass-through with no markup — earnings, rate base and EPS are what matter, not the top line.

Contract structure / payment terms: Utility revenue is tariff-based and largely decoupled in California (SDG&E/SoCalGas recover authorized revenue regardless of volumes via balancing accounts), so weather/volume risk is muted but regulatory risk is concentrated. The SI Partners LNG contracts are long-term, US-dollar, mostly take-or-pay/tolling — but those are being sold down. Customers: ~25M+ consumers across the three utility footprints; Oncor serves Texas load including the explosive data-center demand. Suppliers: Bechtel (LNG EPC), gas/power wholesale markets (pass-through). Competitors: none head-to-head (regulated monopolies in their service territories); the competition is for capital allocation and regulatory ROE vs other utilities.

The strategic pivot (the whole story): In Sept 2025 Sempra agreed to sell 45% of SI Partners to KKR (with CPPIB co-investing) for ~$9.99B and in Dec 2025 to sell Ecogas for 9.0B MXN ($500M), both targeted to close Q2/Q3 2026. Post-close Sempra deconsolidates SI Partners and holds a 25% equity stake. Management's stated goal: become a "leading U.S. utility growth business" — i.e., shed the volatile, FX-exposed, capital-hungry Mexican/LNG businesses and redeploy into California + Texas regulated wires.

Lens 2 · Supply Chain

For a regulated wires/pipes utility the "supply chain" is really a capital-and-regulation chain plus a physical energy-delivery chain. Named stakeholders:

Capital → asset chain (the one that matters for earnings):

• Capital providers → debt investors (Sempra is a serial bond issuer: $11,282M debt issued FY2025 ); equity via a $3.0B ATM program; and now KKR + CPPIB as the SI Partners co-investors and contingently redeemable NCI providers ($5,294M CRNCI contribution in FY2025 financing — the KKR pre-funding).
• → Sempra (holdco) allocates to SDG&E, SoCalGas, and contributes equity to Oncor.
• → Regulators set the return: CPUC (California cost-of-capital + GRC), PUCT/ERCOT (Texas, via Oncor), FERC (interstate transmission). These are the true gatekeepers of Sempra's earnings — a hostile CPUC decision is the single largest swing factor (proven Feb 2025).
• → EPC/construction: Bechtel (fixed-price EPC on Port Arthur LNG Phase 1 + Phase 2), plus transmission contractors for Oncor's Permian Basin Reliability Plan.

Physical energy chain:

• Upstream gas supply (wholesale markets, pass-through) → SoCalGas/SDG&E/Ecogas distribution → ~25M end customers.
• Oncor: ERCOT-dispatched power → Oncor's transmission + distribution wires → Texas load, increasingly hyperscaler data centers (the demand chokepoint flipped from supply-constrained to interconnection-constrained).
• LNG (being sold): Tangguh PSC / US gas feedstock → Cameron / ECA / Port Arthur liquefaction → global LNG offtakers.

Chokepoints / single-source dependencies:

1. Regulatory approval is the binding constraint — not steel or gas. Rate-base growth only converts to earnings if the CPUC/PUCT authorize it at an adequate ROE.
2. Interconnection capacity at Oncor is now the bottleneck for the data-center boom — Oncor's queue (200+ GW of requests; 127 GW substantiated per Q1-26) far exceeds buildable capacity.
3. The KKR close is itself a chokepoint: ~$9.99B of expected proceeds underpins the funding of the $65B plan without excessive equity dilution. Slippage in ECA LNG Phase 1 first production (targeted June 2026) could affect SI Partners' deal economics.

Lens 3 · Competitive Advantages (moats)

The moat is the regulated monopoly + the franchise location. Specifically:

1. Regulated-monopoly economics. SDG&E, SoCalGas and Oncor are franchised monopolies with no direct competitor in their territories. Switching cost for a customer is infinite (you cannot choose another wires provider). This is the classic utility moat — durable, but capped by the regulator's allowed return.
2. Irreplaceable franchise footprint — and this is the differentiator vs. peers. Oncor sits on top of the single most attractive load-growth geography in the United States: Texas/ERCOT, the Permian Basin, and the Dallas data-center corridor. Management frames transmission as "the best low-risk, high-growth way to play the AI-energy boom". ~200+ GW of interconnection requests (255 GW data-center-attributable at Q4-25; 127 GW substantiated at Q1-26) backed by $3.5B of customer collateral deposits — that collateral is a real signal the demand is funded, not vaporware.
3. Scale + cost of capital. A $110.9B asset base and investment-grade ratings give Sempra reasonable access to debt to fund a $65B plan. But note this is a contested advantage right now (negative outlook — Lens 9/10).
4. Bargaining power: weak over regulators, improving over customers. Sempra has almost no pricing power over its true counterparty (the regulator sets ROE). Over data-center customers, Oncor's leverage is rising — hyperscalers need the interconnection and are posting collateral. This is a genuine, improving moat at the Texas franchise.

Net: The moat is real but ordinary at the California utilities (and was just shown to be regulator-capped). The exceptional asset is Oncor's location-moat over AI load growth. The strategic pivot is essentially "sell the commodity-cyclical businesses and concentrate the moat where it's best — Texas wires."

Lens 4 · Segments

Earnings by segment, $M:

Revenue by market, FY2025: Gas $7,965M, Electric $4,453M (contracts-with-customers basis); total revenues $13,702M (2024: $13,185M; 2023: $16,720M — the 2023→24 drop is lower pass-through gas cost, not a demand decline).

Reading the trend:

• California decelerated and is the problem child. FY2025 earnings fell $418M (−23%), driven by a $432M regulatory disallowance (2024 GRC Track 2 final decision, covering 2019–2024) plus lower tax benefits and higher interest. Rate base is still growing, but the segment just demonstrated negative regulatory convexity.
• Texas/Oncor is the accelerator. +$80M (+10%) FY2025 and +$25M (+17%) Q1-26, driven by the new Utility Transmission Mechanism (UTM, effective June 2025), the System Resiliency Plan, rate updates on invested capital, and customer growth. Oncor's estimated regulatory rate base reached ~$31.5B at YE2025 (up from $26.6B YE2024). This is where the next decade of EPS lives.
• Infrastructure's FY2025 "(160)M loss" is an accounting illusion. It was caused by a $703M income-tax charge on classifying SI Partners/Ecogas as held-for-sale (deferred-tax on outside basis differences) plus $445M of adverse FX/inflation on Mexican monetary positions. Operationally the segment was profitable — Q1-26 earnings of $262M (vs $146M) prove it. After the sale, this volatility leaves the P&L (a feature of the pivot).

Phase B — Measure performance

Lens 5 · Earnings Result

Most recent print — Q1 2026 (filed 2026-05-07):

• Diluted EPS $1.58 (Q1-25: $1.39) — +13.7% YoY; basic $1.59.
• Earnings to common $1,037M (Q1-25: $906M), +14.5%.
• Total revenues $3,655M (down from $3,802M — but this is mostly lower pass-through gas cost; ignore the top line).
• Income before tax & equity earnings $848M (vs $651M); equity earnings $367M (Oncor, vs $325M); net income $1,150M.
• The market reaction was negative despite the beat — the stock dipped on the Q1 print. The tell: the beat was partly held-for-sale accounting tailwinds (lower D&A, deferred-tax benefits) and commodity-derivative mark-to-market gains at Infrastructure, not clean operating upside. The market is (rightly) discounting non-recurring quality.
• NCI jumped to $107M (from $2M) — this is the KKR contingently-redeemable NCI now sharing in Infrastructure earnings; a structural change to how much of consolidated profit accrues to Sempra common.

FY2025 (filed 2026-02-26):

• GAAP diluted EPS $2.75 (2024: $4.42; 2023: $4.79) — a large GAAP decline, but driven by the one-offs above (held-for-sale tax, $651M total regulatory disallowances, FX). Adjusted/operating EPS is far higher — management guides FY2026 adjusted EPS of $4.80–$5.30.
• Earnings to common $1,796M; net income $2,072M.
• Equity earnings $1,604M — larger than pretax income before equity earnings ($1,169M). This is critical: a huge share of Sempra's profit comes from the unconsolidated Oncor stake, which means a big chunk of "earnings" is non-cash to the parent until Oncor distributes (FY2025 distributions from investments $1,120M ).

Guidance / outlook: Affirmed FY2026 adj EPS $4.80–$5.30, FY2027 $5.10–$5.70, and a 7–9% long-term EPS growth rate on a record $65B capital plan. Capex for PP&E + investments 2026–2030 expected to total $64.9B; Oncor's standalone 5-yr base plan is $47.5B (+32% vs prior), with up to $10B incremental upside.

Balance-sheet flags:

• Cash collapsed to $29M (from $1,565M) — optically alarming, but $112M sits in the held-for-sale disposal group and the holdco runs on commercial paper / revolvers, so this is normal for a utility mid-build. Short-term debt rose to $4,166M.
• Operating cash flow $4,565M vs capex $12.6B → FCF deeply negative (~−$6.0B pre-investments, ~−$8.1B all-in). This is the defining feature: Sempra does not self-fund its growth — it relies on debt + asset recycling (the $9.99B KKR proceeds) + the ATM. The dividend ($1,603M) is paid out of financing, not FCF.
• Regulatory balancing accounts a $(829)M cash drag (timing of cost recovery).

Unusual vs. its own history: The held-for-sale reclassification, the negative-segment-earnings optics at Infrastructure, the $651M of disallowances, and the cash draw-down are all one-time/transition items. The clean read is: underlying utility earnings grew (Texas strongly, California pressured), masked by a noisy GAAP year.

Lens 6 · Earnings Calls (sentiment trend)

No transcripts on the research-layer shelf (transcripts=0); sentiment reconstructed from web coverage of the last ~3 calls.

• Q4 2024 call (Feb 25, 2025) — the rupture. Management cut 2025 guidance ~11% (to $4.30–$4.70 from ~$5.15 consensus) and the stock fell >18% — its worst day ever. Tone: defensive, surprised by the CPUC; Goldman and UBS downgraded (UBS to Neutral, PT to $78). The "things they stopped saying": the prior confident California-ROE narrative.
• Q4 2025 call (Feb 26, 2026) — the reset-and-redirect. Pivot messaging takes over: announce the $65B plan, the KKR/Ecogas sales, and reframe the story around Texas/Oncor data-center load (273 GW queue, $3.5B collateral). Tone: rebuilding credibility by leaning into the cleanest growth vector and de-risking the balance sheet via asset sales.
• Q1 2026 call (May 7, 2026) — execution proof points. Affirmed guidance and 7–9% growth; updated Oncor queue (127 GW substantiated; $2.9B South Dallas projects released; $10B incremental capex bucket); confirmed FERC/HSR/Mexico/Korea approvals for the KKR deal. Tone: confident on Texas, methodical on the close. Stock still dipped — the market wants the close done and California de-risked before re-rating.

Recurring phrases / focus: "leading U.S. utility growth business," "best low-risk high-growth way to play AI," "transmission," "Permian Basin Reliability Plan," "FFO-to-debt cushion," "simplification." The arc is a credibility-repair story: management is trading a complicated, FX-exposed conglomerate for a simpler, faster-growing, US-regulated compounder — and asking the market to trust the execution.

Lens 7 · Comps

Peer set: large US regulated utilities with data-center/transmission growth exposure. Multiples are `` with date, or n/a. Provenance-critical: I did not fabricate any figure.

EV/EBITDA conflict flagged: GuruFocus reports SRE EV/EBITDA 20.19x as of Apr 2026 ("30% above 10-yr median of 15.52"), while a Q3-2025 datapoint showed 14.26x. The gap likely reflects the held-for-sale distortion to EBITDA and timing — treat SRE's EV/EBITDA as unreliable until post-close financials clean up; the forward P/E (~18.5x) is the more trustworthy gauge.

Read: On forward P/E, SRE at ~18.5x trades at a discount to NEE (~24x), SO (~20x), AEP (~19–20x) and roughly in line with/below DUK (~19x) — despite guiding 7–9% EPS growth (comparable to or above SO's ~7% and AEP). The discount is the Sempra penalty: the Feb-2025 California shock, the un-closed KKR deal, the negative credit outlook, and the equity-method complexity. The thesis is that this discount narrows toward AEP/SO as the conglomerate complexity is sold off and California stabilizes. AEP is the cleanest mirror — same data-center-transmission growth driver, similar EV/EBITDA (~14.3x), and a comparable mega-capex plan ($78B vs Sempra's $65B).

Lens 8 · Stock-Price Catalysts (>5% moves, ~5 years)

Pattern reconstructed from web:

1. Feb–Mar 2025: −18%+ in a single session (worst day ever) — Q4-24 miss + ~11% cut to 2025 guidance on California regulatory setbacks (SDG&E GRC, FERC TO5 adder rejection). Goldman + UBS downgrades. This is the dominant catalyst in the name's recent history.
2. Sept–Dec 2025: positive re-rating on the strategic-transaction announcements (KKR 45% SI Partners $9.99B; Ecogas; Blackstone 49.9% of Port Arthur LNG Phase 2; positive FID + Bechtel NTP on PA LNG Phase 2) — the market rewarded de-risking and simplification.
3. 2026 YTD: grind back to ~$94 (~+8% YTD per TIKR), tracking guidance affirmations and Oncor queue updates.
4. Recurring driver: CPUC/PUCT regulatory decisions — the single most reliable mover. California cost-of-capital and GRC outcomes are repeat catalysts (both up and down).
5. Texas data-center queue updates — increasingly a positive catalyst as the queue grows (273 GW → substantiated 127 GW).

What the tape reveals: The market reacts hardest to (a) California regulatory surprises (the asymmetric downside) and (b) guidance credibility — not to revenue or even quarterly EPS. The 2025 crash proved that the quality and predictability of the regulated earnings is what's priced; the recovery is the market slowly re-trusting management's reset plan. For SRE, regulatory de-risking is the re-rating catalyst, and the KKR close + a clean California cost-of-capital cycle are the specific unlocks.

Phase C — Judge people & books

Lens 9 · Management

CEO: Jeffrey W. Martin — Chairman & CEO of Sempra. Long Sempra tenure (CFO before CEO; CEO since 2018). Oversaw the formation of Sempra Infrastructure Partners (the KKR/ADIA/CPPIB-funded vehicle), the Oncor acquisition integration, and now the pivot back to a US-utility-pure model.

1. Track record: Built Sempra into a Texas + California rate-base compounder and a top-3 North American LNG developer. But the Feb-2025 guidance cut — an ~11% rebase that cost shareholders 18% in a day — is a genuine black mark on management's forecasting credibility and on its read of the California regulatory environment. The subsequent strategic-simplification plan is, in effect, management correcting its own over-complexity.
2. Tenure & skin in the game: Multi-year insider tenure; insider ownership is modest in absolute terms (typical for a mega-cap utility). insider-transactions.csv not on the shelf — n/a for precise holdings.
3. Capital-allocation history: Mixed-to-improving. Positives: the $9.99B KKR + $500M Ecogas + Blackstone PA LNG Phase 2 sales are disciplined "capital recycling" — selling lower-multiple, FX-exposed assets to fund higher-quality US regulated rate base, exactly the right direction. The decision to concentrate on Oncor/Texas wires is well-judged. Negatives: ROE has been mediocre — FY2025 ~5.7%, FY2024 ~9.4%, i.e., the company earns roughly its allowed return at best and below it in a bad year; and the dividend is funded by financing, not FCF, during the build.
4. Red flags: The held-for-sale $703M deferred-tax charge and the $651M of regulatory disallowances speak to (a) tax complexity from the Mexican/LNG structure and (b) a contentious California regulator. The CRNCI structure (KKR's contingently redeemable interest, $3,206M) adds complexity to "who owns the earnings." No related-party-deal or promotional-behavior flags surfaced.
5. Founder vs professional manager: Professional-manager archetype running a regulated monopoly — appropriate for the stage. The job here is capital allocation + regulatory relationship management, and management is mid-course-correcting on both.

Verdict on management: Competent operators who over-complicated the story, got punished by California in 2025, and are now executing a credible simplification. The bet on management is whether they can (a) close KKR on time and on price and (b) avoid another California surprise. They have earned skepticism, not blind trust.

Lens 10 · Forensic Red Flags

Acting as a forensic analyst. Items ranked by how much they should bother an owner:

1. Equity-method earnings dominate the P&L (highest-priority watch item). FY2025 equity earnings $1,604M > pretax income before equity earnings $1,169M. A majority of Sempra's reported profit comes from the unconsolidated Oncor (and historically Cameron) stake. This is legitimate (Oncor is ring-fenced by Texas regulators), but it means: (a) earnings quality depends on entities whose financials don't fully consolidate; (b) cash to the parent lags reported earnings (distributions $1,120M vs $1,604M equity earnings) — reported EPS overstates parent cash generation. Not fraud — but a quality flag every analyst must adjust for.
2. Cash flow diverges sharply from earnings — by design. OCF $4,565M vs net income $2,072M is fine, but FCF is ~−$6.0B and the dividend is paid from new debt/equity. Sustainable only while capital markets stay open and the KKR proceeds arrive. This is a financing-dependency risk, not an accounting trick — but it makes the balance sheet the thing to watch.
3. Held-for-sale accounting introduces non-recurring noise. Stopping D&A on SI Partners ($36–37M/qtr tailwind) and the deferred-tax swings flatter near-term segment optics. Q1-26's Infrastructure "beat" was partly this. Normalize it out.
4. Regulatory disallowances $651M in FY2025 — the income statement now carries a discrete "Regulatory disallowances" line for the first time in 3 years. This is the California regulator clawing back previously-recovered costs (2019–2024 GRC Track 2). It signals the recovery of capital is not guaranteed even after it's spent — a structural risk for a utility whose entire model is "spend, then recover."
5. Rising leverage on a negative-outlook rating. Short-term debt up to $4,166M; total parent-consolidated debt ~$35.0B plus $8.2B held-for-sale debt. SBC is small ($64M) and not a concern. Goodwill went to zero on the balance sheet (reclassified to held-for-sale, $1,602M) — clean.

Regulatory findings (required sub-section):

• SEC Litigation Releases / AAERs: None. regulatory/regulatory-findings.md reports total_sec_findings: 0 — no LR or AAER naming Sempra in 2021-06-30 → 2026-06-30 via SEC EDGAR EFTS.
• Non-SEC enforcement (web search): No material federal enforcement action (FTC/DOJ/FDA/CFPB) surfaced against Sempra at the holdco level. The salient regulatory exposure is state-utility (CPUC disallowances; wildfire liability) and the long-tail SoCalGas Aliso Canyon gas-leak matter (largely settled in prior years) — covered under contingencies, not enforcement. No material non-SEC enforcement hit found.
• Item 3 (Legal Proceedings), most recent 10-K: Sempra's material litigation is dominated by California wildfire-related claims and routine regulatory proceedings. The FY2025 10-K's forward-looking risk factors lead with "California wildfires, including potential liability for damages regardless of fault and any inability to recover... from insurance, the Wildfire Fund and the Continuation Account". The 2025 Wildfire Legislation established an $18B Continuation Account to backstop the Wildfire Fund — a material mitigant, but the underlying inverse-condemnation exposure (strict liability regardless of fault) remains the largest tail risk for SDG&E.
• Conclusion: No accounting-fraud or securities-enforcement findings — verified via SEC EDGAR EFTS (LR, AAER), web search, and 10-K Item 3 as of 2026-06-30. The real "red flags" are earnings-quality (equity-method), financing-dependency (negative FCF), and regulatory-recovery risk (CA disallowances + wildfire) — all disclosed, none fraudulent.

Phase D — Project & stress-test

Lens 11 · Forward Projection

Building from the latest actuals + management guidance. Management affirms adj EPS $4.80–$5.30 (2026) and $5.10–$5.70 (2027) on a 7–9% long-term growth rate — I anchor on that and extend a year. All outputs ``; inputs labeled.

Base inputs:

• 2026 adj EPS midpoint $5.05.
• Long-term growth 7–9%; use ~8% base — driven by Oncor rate base ($31.5B→ growing on a $47.5B 5-yr plan), California rate base growth, partly offset by the ~25%-stake dilution of SI Partners earnings and higher interest.
• Share count ~654M, roughly flat (modest ATM dilution offset by buybacks of forward sales).

• Bull (~10–11%): KKR closes on time/price; Oncor incremental capex (the +$10B bucket + South Dallas) lands in rate base; California cost-of-capital cycle is benign; data-center load converts to authorized investment faster than planned. EPS compounds at the top of guidance.
• Base (~8%): Guidance delivered; Oncor strong, California stable-but-unspectacular, modest SI-stake dilution. $5.05 → $5.45 → ~$5.90.
• Bear (~5% or worse): A second California regulatory shock (adverse cost-of-capital or GRC), a delayed/repriced KKR close forcing equity issuance, or a credit downgrade raising financing cost. EPS growth stalls to ~5% and the multiple compresses — the Feb-2025 playbook repeats.

Forecast logging: Per --watchlist rules, the Brier forecast.ts create step is skipped in this unattended sweep. If promoted to a thesis, the natural tracked forecast would be: "SRE FY2026 adjusted EPS ≥ $5.05, p≈0.62, resolves 2026-12-31.".

Lens 12 · Bull vs Bear

Bull case. Sempra is a cheap, simplifying, super-regional regulated compounder levered to the best load-growth story in the country. The KKR/Ecogas sales convert a complicated, FX-exposed, capital-hungry LNG/Mexico conglomerate into a ~pure California + Texas wires utility — exactly what the market pays premium multiples for. The crown jewel is Oncor, sitting on a 200+ GW interconnection queue (127 GW substantiated, $3.5B customer collateral) in the Permian/Dallas data-center corridor, with a $47.5B 5-yr plan (+32%) and $10B+ of identified upside. Management affirms 7–9% EPS growth + a 15-year-streak dividend (~3% yield, ~52% payout on adjusted EPS) — a ~10–12% total-return algorithm. At ~18.5x forward, SRE trades at a discount to NEE (~24x), SO (~20x) and AEP (~19–20x) despite comparable-or-better growth. The re-rating thesis: as complexity is sold off and California stabilizes, the discount-to-AEP closes — that's ~10–20% multiple upside on top of EPS growth.

Bear case. Three risks that could permanently impair the equity story:

1. California regulatory repricing. Feb 2025 proved the CPUC can cut Sempra's earnings power ~11% overnight (cost-of-capital + GRC + FERC adder), and FY2025 booked $651M of disallowances — clawing back already-spent capital. If the regulator structurally lowers allowed ROE or disallows more WMP/wildfire capex, a third of the company's earnings base is impaired and the multiple stays penalized.
2. Wildfire tail risk (inverse condemnation). SDG&E carries strict liability for wildfire damage regardless of fault. The $18B Continuation Account helps, but a catastrophic San Diego-area fire could overwhelm the Wildfire Fund and force uninsured losses onto shareholders — the same risk that has periodically crushed PG&E and Edison.
3. Financing dependency on a negative-outlook balance sheet. FCF is ~−$6B/yr; the $65B plan + dividend is funded by debt, the ATM, and the KKR proceeds. With Baa3/BBB- ratings on negative outlook (Moody's + S&P), a downgrade or a delayed/repriced KKR close would raise the cost of capital and could force dilutive equity — directly impairing per-share growth.

Pre-mortem (18 months out, thesis broke): Most likely failure path — the KKR close slips or reprices (ECA LNG Phase 1 first production missed June-2026; SI Partners marked down), forcing Sempra to plug the funding gap with equity and a credit-rating downgrade, while the CPUC delivers another unfavorable cost-of-capital decision in the 2026–2028 cycle. EPS growth drops to ~5%, the multiple compresses back toward 15x, and the stock round-trips to the high-$70s — the Feb-2025 low. Second path: a major California wildfire event.

Are multiples too high? No — at ~18.5x forward for 7–9% growth + 3% yield, SRE is below its quality-utility peer group. The multiple already embeds the California penalty. The risk is to earnings, not to multiple.

Contrarian view (what the market refuses to see): The market is anchored on the Feb-2025 California trauma and is treating Sempra as a damaged conglomerate. What it's underpricing is that the KKR/Ecogas sales mechanically de-risk the story — post-close, ~85%+ of earnings come from US regulated wires (California + Oncor), the FX/LNG volatility leaves the P&L, and the comparable becomes AEP, not a Mexican-pipeline-plus-LNG hybrid. The re-rate is a re-classification event, not an operational one — and it's largely contracted to close within ~3 months.

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

Dismantling the bull case.

• Structural break in how it makes money: The model is "spend capital → regulator grants a return." But FY2025 showed the regulator can disallow already-spent capital ($651M) and cut allowed ROE. The "moat" is rented from the CPUC, and the landlord just raised the rent against shareholders. If California's political environment turns more hostile to utility returns (affordability backlash, post-wildfire populism), the entire California third of the company de-rates structurally.
• Revenue/earnings concentration: Earnings are concentrated in (a) California regulatory outcomes and (b) a single equity-method stake (Oncor) that produces more than 100% of pretax-pre-equity income. If Oncor's Texas rate case (2025 comprehensive base rate review, order expected H1-2026) disappoints, or ERCOT/PUCT throttles the data-center interconnection economics, the growth engine stalls — and Sempra doesn't even control Oncor (80.25% economic, ring-fenced governance).
• The moat is weaker than bulls think: Data-center load is a queue, not revenue. 200 GW of interconnection requests ≠ 200 GW of rate-based investment earning a return — much of the queue is speculative, duplicative, or will never connect. The $3.5B collateral covers a fraction. Bulls extrapolate a queue into earnings; the conversion rate is unproven.
• Most dangerous competitor bulls underestimate: Not a rival utility — it's the regulator and the affordability lobby. And on the data-center thesis, AEP, NEE and others are chasing the same hyperscaler load with bigger balance sheets and cleaner stories ($78B AEP plan, 63 GW contracted) — Sempra isn't uniquely positioned, it's one of several.
• Worst capital-allocation/incentive flags: The CRNCI / KKR structure is complex and dilutive to common at the Infrastructure level (NCI jumped to $107M in Q1-26). The dividend is funded by financing while FCF is −$6B — a classic "borrowing to pay the dividend" pattern that works only until it doesn't.
• Assumptions that must hold for today's price: (1) KKR closes ~$9.99B on time and on price; (2) no second California shock; (3) ratings hold at investment grade; (4) the data-center queue converts to rate base at the assumed pace; (5) capital markets stay open for a serial issuer.
• If growth disappoints 20–30%: EPS growth of ~5–6% (vs ~8% base) re-rates SRE toward a 15x slow-utility multiple → mid-$70s, a ~20% drawdown from ~$94. The Feb-2025 low is the template.
• Single scenario that permanently impairs: A catastrophic SDG&E wildfire that exhausts the Wildfire Fund + Continuation Account and forces large uninsured, non-recoverable losses onto shareholders — inverse-condemnation strict liability is the PG&E-style equity-wipeout risk. Plausibility: low in any given year, but non-trivial cumulatively and structurally unhedgeable.

Lens 14 · Management Questions (ordered by information value)

1. The KKR/SI Partners close: confirm the expected net proceeds (~$9.99B base), the gating conditions remaining after FERC/HSR/Mexico/Korea approvals, and — critically — is the purchase price subject to downward adjustment if ECA LNG Phase 1 first production slips past June 2026? (Highest-value: it determines the funding of the entire $65B plan and the equity-dilution risk.)
2. California cost-of-capital 2026–2028: what allowed ROE and equity ratio did the final decision grant SDG&E/SoCalGas, and what is the earnings sensitivity to a 50bp ROE change? (The single biggest swing factor proven in Feb 2025.)
3. Of Oncor's 200+ GW interconnection queue (127 GW substantiated), what GW do you underwrite as rate-based investment in the $47.5B plan, and what is the historical request-to-energized conversion rate? (Tests whether the data-center story is earnings or hype.)
4. FFO-to-debt: where do you sit today vs the downgrade thresholds at Moody's/S&P, and what specifically removes the negative outlook? (Directly tests the financing-dependency bear case.)
5. Post-close earnings mix: pro forma, what % of EPS comes from US regulated wires, and what is the new run-rate FX/commodity earnings volatility? (Quantifies the re-classification re-rating thesis.)
6. Wildfire: what is your maximum probable uninsured loss from a catastrophic SDG&E event after the Wildfire Fund + $18B Continuation Account, and how much equity is at risk in a tail scenario?
7. The $651M FY2025 regulatory disallowances: what changed in the CPUC's posture on cost recovery, and what is your current at-risk balance of spent-but-not-yet-recovered capital?
8. Equity-method gap: equity earnings ($1.6B) exceed parent distributions ($1.1B) — what is the multi-year plan to convert Oncor earnings into parent cash to fund the dividend without external financing?
9. Texas Oncor base rate case (order expected H1-2026): what ROE/capital-structure outcome are you assuming, and what is the downside case?
10. Dividend policy: with ~52% payout on adjusted EPS but FCF at −$6B, at what point does dividend growth become self-funded rather than financed?
11. The +$10B incremental Oncor capex bucket (South Dallas $2.9B released): what's the probability-weighted timeline into the formal plan, and the associated financing?
12. After SI Partners is sold to 25%, what is your long-term intent for the residual LNG stake — hold for optionality, sell down further, or re-grow?
13. Capital allocation: with a ~9–10% achievable ROE, why is organic rate-base growth the best use of the KKR proceeds versus buybacks at a sub-peer multiple?
14. Affordability: California customer bills are politically charged — what is your assumption for rate-increase tolerance, and the risk of legislative/CPUC action capping returns?
15. What is the single risk on the 3–5 year horizon you think the market is most mispricing — in either direction?