Centerpoint Energy Inc CNP

Thesis

CenterPoint Energy (CNP) is a pure-play regulated utility holding company whose two main legs are (1) Houston Electric, a transmission-and-distribution utility (TDU) serving roughly 2.9 million metered customers in fast-growing greater Houston, and (2) CERC, a multi-state natural-gas LDC franchise across Texas, Indiana, Minnesota and Ohio (Ohio LDC is being divested to NFGC). After exiting its Louisiana/Mississippi gas LDCs (LAMS) and announcing the Ohio sale, CNP has sharpened itself into a Texas-Indiana wires-and-pipes story with a sizable 10-year capital plan funded by rate-base growth, securitization bonds (Restoration Bond Companies II/III) and equity issuance.

The compounding case is straightforward: rate base grows, allowed return on rate base flows through to EPS and dividends, customer growth in Houston (data centers, refining/exports, advanced manufacturing) plus storm-hardening capex authorized post-Beryl creates a long runway. Buffett's MidAmerican franchise [2][5] is the template — 'massive investments in both transportation and energy,' regulators that 'allow us an appropriate return on the huge amounts of capital we must deploy.' That model can compound book value at allowed-ROE for decades.

The price disagrees. The deterministic scorecard places CNP at composite 53/100 with a base-case IV of $25.92 and a high-case IV of $38.78 versus the $43.35 quote — a price/IV of 1.67x. To justify today's price, a reverse DCF demands ~12.4% growth in owner earnings; that is roughly double a normal allowed-ROE utility's earnings power and three to four times typical rate-base CAGR. Net debt/EBITDA at 6.31x and 5-year FCF conversion at -78% (capex outrunning operating cash flow) confirm this is a capital-consuming compounder, not a self-funding one. With ROIC 4.4% versus a likely 5-6% cost of capital, every incremental dollar deployed earns at most a thin spread over WACC. The work is to wait. Owning CNP makes sense in the high-$20s where IV-base meets price; today's quote leaves no margin of safety, only optimism.

Moat

CenterPoint's moat is the regulatory franchise — the legal monopoly to deliver electricity and gas inside its certificated service territory. We test all five moat types.

Pricing power. CNP cannot raise prices like a consumer-brand business; it earns an allowed return on rate base set by the PUCT (Texas), IURC (Indiana), MPUC (Minnesota), PUCO (Ohio, exiting). What CNP DOES have is the right to file for cost recovery — base rate cases plus rider mechanisms (TCOS, TCRF, GRIP, TDSIC, DCRF) that translate capex into rates with relatively short regulatory lag. Beryl restoration costs (~$1.1B) are being securitized via Restoration Bond Company III rather than absorbed. That is real pricing power, but it is conditional on regulators continuing to honor the 'symbiotic' bargain Buffett describes [4][5] — capital invested earns a fair return because society needs the wires. Verdict: present, but not durable in the way a Coke is durable.

Switching costs. Effectively infinite for the wires customer; an end-user cannot choose a different TDU in Houston. Note however that in deregulated Texas the energy commodity is sold by Retail Electric Providers (REPs); CNP is paid by REPs for delivery. So CNP bears REP credit risk but enjoys 100% delivery monopoly. Indiana operations are vertically integrated and similarly captive on the gas/electric distribution layer.

Network effects. None in the social-network sense, but real economies of density: one set of poles, wires, transformers and substations serves all customers in the territory. Adding a data center load on an already-engineered grid is high-incremental-margin if regulators allow recovery of incremental investment. Buffett's MidAmerican commentary — 'each business has earning power that even under terrible conditions amply covers its interest requirements' [3] — applies in spirit to a well-run TDU but is NOT yet validated at CNP, where interest-coverage is unreported in the scorecard and net debt/EBITDA is 6.31x (vs MidAmerican-style utilities typically 4-5x).

Intangibles. Two matter: the operating license itself (impossible to replicate) and the franchise rights granted by Texas/Indiana cities. There is also a softer intangible — regulatory goodwill — which Buffett emphasizes [2] ('It is in our self-interest to conduct our operations in a way that earns the approval of our regulators'). CNP's intangible has been DAMAGED by Beryl: the public response to the July 2024 outages was harsh, multiple class-action suits are pending [HurricaneBerylClassAction1/2/3 in the 10-Q], and Texas-legislature scrutiny has been intense (the TEEEF mobile-generation episode is being unwound at PUCT direction). Compare to Buffett's MidAmerican, which posted a 95.3% customer-satisfaction rating [2] and was 'glad to see' by regulators. CNP is, today, NOT in that position with the PUCT.

Cost advantages. Modest. Scale matters for procurement and back-office, but CNP is mid-cap relative to NEE, DUK, SO. The relevant cost edge is capital-cost — investment-grade ratings let it borrow cheaply, which keeps customer rates competitive. That is real but commoditized: every IG utility has it.

Competitor stress test. Could a competitor with $10B and 5 years take share? No — physically impossible to overbuild duplicate poles-and-wires in Houston, and FERC/PUCT would not certificate it. The franchise is competition-proof. The ACTUAL competition is regulatory and political: Texas legislators choosing to disallow Beryl recovery, force lower allowed ROEs, or tighten storm-hardening rules. That risk is real and live.

Erosion risk. (a) Rooftop solar / distributed generation eroding kWh sales (largely neutralized by decoupling and capacity-based rates). (b) Politicization of allowed ROE post-Beryl. (c) Cost-of-capital risk if rates stay elevated and rate base grows faster than allowed-equity-thickness. (d) Tail risk: a single catastrophic storm whose costs are NOT permitted recovery would impair equity directly.

Moat verdict: NARROW.

Management & Capital Allocation

CenterPoint's capital allocation is the standard regulated-utility playbook: reinvest the bulk of operating cash flow into rate base, fund the gap with debt and periodic equity, pay a steady dividend, do tuck-in M&A, divest non-core LDCs. We grade it on the five Buffett choices.

Reinvestment. CNP is mid-execution on a 10-year capital plan disclosed as a core forward-looking item in the 10-Q. That plan funds wires modernization, storm-hardening (Greater Houston Resiliency Initiative), substation buildouts to serve Houston-area data center / AI / refining / advanced-manufacturing demand, and Indiana generation transition (replacing retired coal with gas + renewables). Rate-base CAGR guidance has historically been roughly 9-10%. The economic question is the spread between earned ROIC and cost of capital. The scorecard shows a 10-year average ROIC of 4.4% and a 5-year ROIIC of 6.8%. Against a likely WACC in the 5.5-6.5% range, that is a thin-to-zero spread. This is reinvestment that grows the company but barely creates per-share value above cost of capital. Buffett's framing in 2009 [5] is precisely this: 'we are today quite willing to enter businesses that regularly require large capital expenditures. We expect only that these businesses have reasonable expectations of earning decent returns on the incremental sums they invest.' CNP's incremental returns are decent — not exceptional.

Acquisitions. The Vectren acquisition (closed Feb 2019) added Indiana / Ohio / Minnesota gas and Indiana electric. Integration appears complete. Strategic direction has since shifted toward simplification: LAMS (Louisiana/Mississippi) gas LDCs sold to Bernhard Capital affiliates in 2025; Ohio gas LDC sale to NFGC announced Oct 20, 2025 (Ohio Securities Purchase Agreement). These divestitures are consistent with focusing capital on highest-rate-base-growth jurisdictions (Texas, Indiana). Multiples appear reasonable; however the historical Vectren transaction was rich and the current Ohio sale price/EBITDA is the relevant benchmark.

Debt. Net debt/EBITDA of 6.31x is high vs the regulated-utility peer median (~5.0-5.5x), although consolidation includes securitization debt that is non-recourse and self-amortizing. Even adjusting, leverage is elevated and rating-agency-sensitive — Moody's, S&P and Fitch all monitor a roughly 13-15% FFO/Debt threshold. A downgrade would materially raise the cost of the very capital plan that is the entire bull case.

Buybacks. Effectively none — share count UP 4.5% over 10 years per scorecard. CNP is a net issuer, which is normal for a fast-growing rate-base utility but means there is no per-share leverage from repurchase. The Buffett P/IV question for buybacks is moot.

Dividends. CNP pays a steady, growing dividend (currently ~$0.86/yr at the time of writing). Payout is consistent with peer utilities. Reasonable, not exceptional.

Communication. 10-K/10-Q disclosures are standard, with extensive litigation disclosure around Beryl class actions and 2021 Winter Storm Uri MDL [BellwetherCasesMember, MultiDistrictLitigationMDLMember]. Forward guidance on the 10-year capital plan is provided. Beryl response — the politically embarrassing TEEEF mobile-generation contracts — is being publicly unwound (release of large units to San Antonio area, removal of medium units, rate reduction to reflect TEEEF removal). This is reasonable accountability, though the underlying procurement decision was a clear misstep.

Capital allocator: B-.

Industry Structure

Porter's Five Forces on a regulated electric/gas T&D utility:

Threat of new entrants — Very Low. Cannot certificate a duplicate utility; the franchise is legally exclusive. The Beryl episode shows that even when an incumbent stumbles, the response is regulatory tightening, not new entrants. Buffett [2][5]: regulators 'are glad to see us' once a utility builds a service reputation; conversely, they make life painful for those that don't.

Bargaining power of suppliers — Low to Moderate. CNP buys steel, transformers, copper, gas commodity, EPC services. Transformer lead times have stretched to 18-24 months industry-wide; that supply-chain tightness elevates project execution risk and cost. Capital-providers (debt and equity markets) have moderate power, especially if rates stay high — financing cost rises before regulators reset allowed ROE.

Bargaining power of buyers — Moderate, increasing. End-customers individually have zero power. But politically, customers act as a bloc through the legislature and PUCT. Affordability has become the single dominant regulatory theme. The 10-Q lists 'customer affordability' nine times in just the forward-looking statements section. Hyperscale data-center customers DO have power — they negotiate large-load tariffs and can locate the load elsewhere. CNP discloses growth from 'expansion of data centers (associated with...AI), energy refining and exports, advanced manufacturing.'

Threat of substitutes — Moderate, slowly rising. For electricity: rooftop solar + battery storage is a partial substitute for grid kWh but does not replace wires service for the average customer. For gas: heat-pump electrification is a real long-term substitute for gas LDC therms; Indiana and Minnesota are seeing modest electrification pressure, less so in Texas. The wires business is the more durable franchise; the gas LDC business is structurally challenged by 2050 climate policy.

Industry rivalry — Low (within franchise) / Moderate (for capital). No rivalry inside the territory. But CNP competes for capital against ~40 other US IOUs. NEE, AEP, SO, DUK, ED have stronger franchises and lower leverage. Capital that flows to higher-quality utilities does not flow to CNP.

Value pool location and trajectory. The economic value pool in T&D utilities is the spread between allowed ROE and after-tax cost of debt, multiplied by rate base. Three trajectories matter: (1) rate-base growth — strongly positive, especially in Houston where load growth is structurally high; (2) allowed ROE — flat to declining, regulators are sympathetic to affordability; (3) cost of debt — has stepped up materially since 2022 and is the most-watched line. Net effect: the absolute value pool is growing because rate base is growing, but per-dollar economics are softening.

Regulatory comparison: Texas (PUCT) is generally constructive but has been hostile post-Beryl on TEEEF and on storm-cost recovery scope. Indiana (IURC) is a top-quartile regulator; minimum-bill / TDSIC mechanisms are favorable. Minnesota is average-to-tough; Ohio is being exited.

Customer-mix tailwinds: Houston load growth from petrochemicals expansion, LNG export buildout (Sabine Pass, Corpus Christi area), data-center investment, and population in-migration is structurally above national average. This is the single most attractive feature of CNP's portfolio.

Industry Verdict: Good. The franchise is excellent; the leverage and current regulatory friction pull it down to good rather than excellent.

Inversion (Bear Case)

I am now a short seller. I will write the strongest credible case against CNP at $43.35.

The single event that kills this. The PUCT issues an unfavorable ruling on Houston Electric's Beryl restoration cost recovery — disallowing a meaningful slice of the $1.1B+ securitization, OR setting a punitive prudency review precedent that lingers into the next Houston Electric base rate case (filed 2025-2026). Combined with a downgrade from one of the three rating agencies (Moody's, S&P, Fitch) for sustained FFO/Debt below 14%, CNP must issue equity into a falling stock and re-price its planned debt issuance. The capital plan is funded by a stable cost of capital; if that cost rises 75-100 bps and equity dilutes 5-7% to plug the gap, the per-share earnings power of the 10-year plan compresses by ~$2-3. Multiple compresses simultaneously. The stock re-rates from 29x to 17x on a now-lower EPS base.

Why the moat is narrower than bulls think. The bull narrative is 'regulated monopoly in growing Houston.' But the moat is conditional. Buffett's MidAmerican commentary [2] is explicit: regulators give a fair return because the operator earns approval. Houston Electric's 95th-percentile customer-satisfaction equivalent does not exist — Beryl's outage response was widely panned, and three separate class actions are now pending [10-Q HurricaneBerylClassAction1Member, Member2, Member3, plus HurricaneBerylPersonalInjuryMember]. The PUCT's TEEEF rule (effective January 2025) was a direct legislative slap. Bulls are pricing a MidAmerican-quality regulatory relationship; the actual relationship is closer to PG&E's pre-2018 — strained, politicized, with risks that have been understated for years.

Why management is worse than it appears. TEEEF was a failed procurement. Management committed to large mobile-generation leases that the legislature subsequently restricted, and is now publicly unwinding the contracts at unfavorable economics. This is not a one-time error — it is a pattern of optimistic capital allocation under regulatory stress. The 10-year capital plan is cited 8+ times in forward-looking statements; that is not vision, that is leverage. A management team that needs to deploy this much capital this fast is necessarily one that has lost flexibility to walk away. The capital allocator grade should be C, not B-.

What bulls are extrapolating that won't hold. Three things. (1) That Houston load growth is durably 2-3% organic when it is heavily reliant on a small number of large customer categories — petrochemical, LNG export, data center — each of which has its own boom-bust dynamic. A drop in LNG export economics (already happening as Brent / Henry Hub spread compresses) removes a leg. (2) That data center demand will pay for itself: hyperscalers negotiate hard, and behind-the-meter power options (Constellation–Microsoft Three Mile Island reactor restart, Talen–AWS Susquehanna deal) show that the largest customers are willing to bypass the grid entirely. (3) That allowed-ROE is sticky around 9.5-10%. In Texas in 2025, the political climate is moving toward affordability-first regulation; allowed-ROE compression to 9.0% is plausible.

Valuation trap (multiple compression / regime change). P/E of 29.3 vs 10-year average 20.6 — already 42% above its own historical multiple. P/IV of 1.67x. Implied DCF growth of 12.4% — utilities don't compound owner earnings at 12%. Suppose two things happen by 2027: rate-base growth slows to 7% (vs ~10% guided) and the multiple reverts to its 10-year average of 20x on a roughly flat earnings base. That puts CNP at $30 — about a 30% drawdown. If FFO/Debt slips and a downgrade ensues, the multiple could overshoot to 16x and the stock at $24. The asymmetric risk is to the downside: bull-case IV is $38.78 (lower than today's price); base-case IV is $25.92 (40% below today's price); low-case IV is $14.79.

If I am right, the stock could be worth $24-28 within 2-3 years.

Lollapalooza Bias Check

Active biases in me, the analyst, right now.

Anchoring. The scorecard hands me an IV-base of $25.92 and an IV-high of $38.78. I am anchored to those numbers. They are model outputs based on neutral 12/17/22 P/FCF multiples (the scorer notes flag this) and on a -78% FCF conversion that almost certainly understates normalized owner earnings for a utility in a heavy capex cycle. If maintenance capex is closer to half of total capex (vs the conservative scorecard assumption), normalized owner earnings could be 50-100% higher and IV-base shifts up. I should not mechanically treat $25.92 as truth.

Recency. Hurricane Beryl is the loudest signal in my information set. I am over-weighting Beryl's regulatory implications because it is recent and emotionally vivid. Most utilities recover most storm costs most of the time. The base case is recovery, not disallowance. I am letting Beryl color my view of the entire moat.

Confirmation. Once I noticed P/IV of 1.67x, every fact I encountered fit the 'overpriced' narrative — leverage, ROIC, share count up, FCF conversion negative. I did not work as hard to find disconfirming facts: Houston's structural load growth, the simplification post-LAMS/Ohio, the post-Beryl resiliency capex that itself becomes rate base.

Authority. Buffett's MidAmerican commentary [1][2][5] is in the canon. I am using it as both template and benchmark, which biases me toward judging CNP harshly against an unusually high-quality comparable. Most utilities are not MidAmerican.

Deprival super-reaction (avoidance). I am steering toward 'Trim/Hold' because I do not want to be the analyst who said Buy at $43 and watched it go to $30. The asymmetry of being wrong on the upside (looking foolish) vs wrong on the downside (looking prudent but bored) is making me more conservative than my own model justifies.

Incentive. I have a self-image as a Buffett-Munger value analyst. P/IV 1.67x triggers my identity to say 'no.' That identity-protection is itself a bias.

Net effect: my recommendation is probably one notch more cautious than a fully calibrated answer. I will offset this slightly by setting conviction at medium rather than high.

10-Year Outlook

Same fundamental business model in 2035? Yes. CenterPoint will still be a regulated electric and gas T&D utility serving Houston and Indiana. The wires will still be wires; the gas mains will still be gas mains (though the gas LDC business faces 2040+ decarbonization pressure that becomes visible by 2035).

Customer base larger? Yes, very probably. Greater Houston is one of the few US metros structurally adding population, industrial load (LNG export, petrochemicals expansion, advanced manufacturing per CHIPS Act build-out), and data center load (AI training and inference). The 10-Q explicitly identifies these as growth drivers. Indiana is flatter but stable.

Profit per customer higher? Probably yes in nominal dollars, modestly so in real dollars. Allowed ROE may compress 25-50 bps over the decade as affordability pressures intensify. Rate base per customer will grow due to storm-hardening, decarbonization capex, and grid modernization. The combination is mid-single-digit nominal EPS growth — not explosive.

Moat wider? No. The moat is what it is — a legal monopoly. Wider moat would require deeper regulatory trust, which Beryl set back. Recovery is plausible by 2030 but not assured.

Single biggest threat? A combination scenario: (a) sustained higher cost of debt + (b) tightening allowed ROE + (c) one more catastrophic storm that triggers political (not just regulatory) action restricting cost recovery. Each is plausible alone; together they would impair equity value materially.

Other threats: gas LDC stranding risk in the 2030s; behind-the-meter data center deals bypassing the grid; cyber/physical attacks on critical infrastructure; tax policy changes (the 10-Q flags OBBBA, IRA changes, CAMT).

Positive case: Houston growth surprises to the upside, a constructive PUCT chairman emerges, allowed ROE holds at 9.5%, capex earns spread, FFO/Debt rebuilds, multiple re-rates.

Net: I can confidently say the business will exist and earn an allowed return. I cannot confidently underwrite the rate of compounding. The outcome distribution is wider than for a Mid-American or NextEra.

CONFIDENCE: medium