Exelon is now a pure-play regulated T&D utility earning a regulated return on a rate base growing ~7-8% annually, but at $46.50 the market is paying 1.69x our base intrinsic value and pricing in 9.4% perpetual owner-earnings growth that the rate-base math does not support.

Plain English

Exelon owns the electric wires that connect houses and businesses to the grid in six places: Chicago, Philadelphia, Baltimore, Washington DC, Delaware, and New Jersey. State officials let them earn a fixed percentage profit on the equipment they install. As long as they keep adding wires, transformers, and substations, profits grow slowly and steadily — like a savings bond. The business is safe and dull. The problem today is that the stock price already assumes a lot of growth, so even though the business is fine, owning it at this price is not a bargain.

Thesis

Exelon (EXC) is the largest pure-play regulated transmission and distribution utility in the U.S., running six state-regulated wires-only utilities (ComEd in Illinois, PECO in Pennsylvania, BGE in Maryland, Pepco in DC, DPL in Delaware, ACE in New Jersey) after spinning off the merchant/nuclear generation business as Constellation Energy in February 2022. The business is exactly what Buffett describes for BNSF and MidAmerican [1]: very long-lived, regulated assets funded by large amounts of long-term debt, where 'earning power...amply covers their interest requirements' and the fundamental compact is that the utility delivers safe, reliable service and capex, and regulators in turn allow a fair return on the ever-increasing capital invested [5].

What could compound: a roughly $46B rate base growing 7-8% per year through ~2027 (electrification, reliability, grid hardening, data-center load in PJM), translating to ~5-7% EPS and DPS growth at a regulated ROE of ~9.5-10%. That is a defensible bond-like compounding stream — but a bond-like one.

The problem is price. The scorecard's deterministic IV range is $18.94 (low) / $27.51 (base) / $40.85 (high) against a current price of $46.50, giving a price/IV ratio of 1.69x — i.e., the stock trades above even the optimistic case. The reverse DCF implies the market is pricing 9.4% perpetual owner-earnings growth, roughly double what a regulated wires utility can structurally deliver without persistent rate cases at favorable ROEs. ROIC of 3.85% over ten years is exactly the 'low return on capital' Charlie warned about [5], and net debt / EBITDA of 6.98x is at the upper end of utility comfort.

Verdict: this is a fine business, badly priced. Owning makes sense in the low-$30s; today is a pass.

Moat

Exelon's moat must be evaluated as a regulated wires utility, not a competitive enterprise. The five moat lenses:

1. Pricing power (regulatory franchise). EXC's six operating utilities (ComEd, PECO, BGE, Pepco, DPL, ACE) each hold an exclusive franchise to distribute electricity (and in some cases gas) across a defined service territory. That franchise is the moat. There is no competing distribution wire in ComEd's Chicago territory; building one is economically and politically inconceivable. Pricing is set by state public utility commissions (Illinois ICC, Pennsylvania PUC, Maryland PSC, etc.) and FERC for transmission, on a cost-of-service basis: rate base x allowed ROE + recovery of operating costs. This is structurally the same pricing mechanism Buffett describes for BHE [4][5]: 'we look to our utilities' regulators (acting on behalf of our customers) to allow us an appropriate return on the huge amounts of capital we must deploy.' The ceiling is the allowed ROE (currently ~9.5-10% across EXC's jurisdictions); the floor in a fair commission is roughly the cost of capital. Pricing power is genuine but capped — it is the right to earn a regulated rate, not the right to set price.

2. Switching costs. Captive customers cannot switch — EXC owns the wires. Even where retail electricity supply is deregulated (Illinois, Pennsylvania, Maryland), the distribution charge belongs to the wires owner. Customer concentration is essentially zero (~10.5M customer accounts).

3. Network effects. Modest. The grid itself is a network, but the utility does not capture additional value as more nodes connect — it captures the regulated return on the incremental rate base. Data-center load growth in PJM is a tailwind for rate-base growth, not a true network-effect flywheel.

4. Intangibles. The most important intangible is regulatory goodwill. Buffett puts this exactly: 'we hope to buy more regulated utilities in the future – and we know that our business behavior in jurisdictions where we are operating today will determine how we are welcomed by new jurisdictions tomorrow' [3]. Exelon's record here is mixed-to-poor. The 2020 ComEd bribery deferred-prosecution agreement (Madigan-related) cost $200M in penalties, badly damaged ComEd's standing in Springfield, and contributed to multi-cycle constructive-but-not-generous rate outcomes in Illinois. Compare to BHE's record [4]: ranked top-decile in safety, no rate increase in Iowa for 16 years, welcomed by every regulator. EXC is not in that category.

5. Cost advantages. Limited. The cost-of-service model essentially passes through costs; efficiency gains accrue first to ratepayers via the next rate case, then partially to shareholders via earned-vs-allowed ROE. EXC's earned ROE has historically run 50-150 bps below allowed across the portfolio — meaning some operational underperformance leaks. There is no scale-driven unit-cost advantage analogous to GEICO's [3] or BHE's Iowa efficiency [4].

Competitor stress test ($10B + 5 years): Irrelevant in the conventional sense — no entrant can build parallel distribution wires. The real 'competition' is: (a) rooftop solar + storage + microgrids reducing kWh throughput per customer, (b) state legislatures or commissions tightening allowed ROE or disallowing capex, (c) FERC reducing transmission incentive ROE adders. None of these kills the business; all of them compress it.

Erosion risk: The franchise is durable for decades. The economic value of the franchise erodes if (a) regulators authorize lower ROEs (mean-reversion from ~9.6% toward cost of capital, perhaps 8.0-8.5% in a structurally lower-rate world — though current rates argue the other way), (b) rate-case lag widens during inflationary capex cycles, (c) political backlash to data-center cost allocation pushes large-load customers off-system or onto special tariffs that reduce the residential rate-base contribution.

The ROIC of 3.85% [scorecard] is the giveaway: this is a business where 'incremental sums' earn 'decent returns' [5] but never the 'terrific economics' of 25%+ unleveraged returns [6]. That is the regulated bargain — accept low returns on huge capital in exchange for low business risk.

Moat verdict: NARROW.

Management

Exelon's capital allocation is largely predetermined by the regulated business model. The five capital choices in this case:

1. Reinvest in the business. This is ~95% of the capital story. Management has guided to roughly $38B of capex over 2024-2027, taking rate base from approximately $46B today to $65B+ by 2027 (~7.5% CAGR). The reinvestment is not optional — it is the mechanism by which the company earns. Every dollar of rate-base growth, multiplied by allowed ROE (~9.5-10%) net of taxes and equity issuance dilution, is the EPS engine. The ROIIC 5-year of 10.77% [scorecard] is consistent with this math: incremental capital earns roughly the regulated return on the equity portion plus debt leverage. The question is quality of reinvestment — is the capex defensive maintenance (good, accretive at allowed ROE) or growth/electrification capex with regulatory lag risk (still accretive, but on a delay)? The split is roughly 60/40 in favor of growth, which means meaningful working-capital and AFUDC drag — and that shows up in FCF conversion of -1.57x [scorecard].

2. Acquisitions. Effectively zero post-Constellation spin. The 2012 Constellation merger and 2016 Pepco Holdings acquisition are in the rear-view. Management has signaled organic-only growth, which is the right call given the company's regulatory rebuild required in Illinois.

3. Debt. Net debt / EBITDA of 6.98x [scorecard] is at the upper end of the regulated-utility comfort zone but not alarming for a wires-only business with ~80% rate-regulated cash flows. Coverage metrics imply BBB / Baa-area credit (which is where EXC and its operating subs sit). The capital structure is fine, but there is no margin to take it higher; the next ~$20B of capex will require sustained equity issuance — the share count is up only 0.88% over ten years [scorecard], but that masks the post-spin reset and the ~$700M-1B annual ATM equity program now in place.

4. Buybacks. None. This is correct. A regulated utility issuing equity to fund rate-base growth should not also be buying back stock; doing so would just be a tax-inefficient round trip. There is no question of 'average P/IV when buying' to evaluate, which is itself a signal of disciplined capital allocation.

5. Dividends. Post-spin payout was reset to roughly 60% of operating EPS, currently ~$1.60/sh annual, ~3.4% yield at $46.50. Management has guided to dividend growth in line with EPS growth (~5-7%). This is the right framework: dividends as a return-of-capital mechanism rather than a hostage to be defended at all costs.

Communication quality. EXC's investor materials are detailed, with rate-base build, capex, financing plans, and earned-ROE bridges presented cleanly. CEO Calvin Butler (since Jan 2023) is a known regulated-utility operator and has spoken candidly about the ComEd legacy. The company avoids non-GAAP gymnastics of the kind Buffett ridicules [Munger excerpt 1] — operating EPS adjustments are limited to mark-to-market and pension items, not stock-comp.

The ComEd episode is the asterisk. The 2020 deferred prosecution agreement with the U.S. Attorney for the Northern District of Illinois — admitting to a years-long bribery scheme involving Speaker Madigan — is precisely the kind of behavior Buffett warns destroys the regulator-utility compact [3]. The $200M fine is not the issue; the issue is the multi-decade trust deficit in Springfield. Current management was not the perpetrator and has invested in compliance, but the file is open and the cost shows up in less-generous Illinois rate outcomes than peers achieve in Pennsylvania and Maryland.

Capital allocator: B-.

B- because: the framework is correct (no buybacks, no acquisitions, dividend policy aligned with earnings power), the reinvestment opportunity is genuine, but (a) the historical ComEd governance failure is a permanent mark, (b) earned-vs-allowed ROE persistently runs below allowed, suggesting operational efficiency lags BHE-class peers [4], and (c) the FCF picture (heavily negative conversion [scorecard]) means the equity holder is underwriting growth via dilution.

Industry

Porter's Five Forces applied to U.S. regulated electric T&D:

1. Threat of new entrants — VERY LOW. Building parallel distribution wires in an existing service territory is economically irrational and politically impossible. Franchises are exclusive and granted by state legislatures and PUCs. The barrier is essentially absolute. This is the single most attractive feature of the industry.

2. Bargaining power of suppliers — LOW-MEDIUM. Key suppliers: transformers (currently in shortage, lead times 100+ weeks, prices up 60-80% over three years), utility-grade conductor and steel, EPC contractors for substation work, and capital itself (debt and equity markets). Supply-chain inflation is real but is largely passed through to ratepayers via fuel/cost trackers and rate cases — with regulatory lag of 12-24 months. The cost-of-capital squeeze from 2022-2024 (rates rose faster than allowed ROEs adjusted) is the more meaningful supplier-power story; that is now beginning to reverse.

3. Bargaining power of buyers — MEDIUM. Residential and small commercial customers have effectively zero bargaining power (captive). Large commercial and industrial customers have some indirect power via political pressure, special tariff negotiation, and the threat of behind-the-meter generation or relocation. The new force is hyperscale data centers in PJM (and specifically in Northern Virginia adjacent to EXC's Pepco/BGE territory, and AEP/Dominion territory): Amazon, Microsoft, Google and Meta are extracting bespoke contracts and pressuring commissions on cost allocation between data-center load and residential ratepayers. This is a two-edged sword for EXC: it drives load growth and rate-base growth, but it also creates political risk if residential rates spike to fund grid expansion 'for' data centers. Maryland and Virginia commissions are actively wrestling with this in 2025-2026.

4. Threat of substitutes — LOW-MEDIUM and rising. The substitute is self-supply: rooftop solar plus battery storage plus, eventually, microgrids. Net-metering rules vary by state but generally the wires utility loses kWh throughput while still bearing fixed grid costs. Rate-design responses (demand charges, fixed customer charges, time-of-use) lag the technology. For now this is a slow erosion (2-3% of residential load over a decade), not an existential threat. The longer-term substitute — full grid defection — remains uneconomic for >95% of customers and likely will for another decade.

5. Rivalry among existing competitors — LOW within service territory, MODERATE in capital markets. No price competition between utilities (each holds a monopoly). Competition is for capital allocation favor (relative valuation), for talent, and for regulatory benchmarking — commissions compare allowed ROEs and capital structures across jurisdictions, which puts mild downward pressure on outliers.

Value pool location and trajectory. The value pool sits squarely with the equity holders of well-managed utilities in constructive jurisdictions and with grid-equipment suppliers (Eaton, Hubbell, Quanta Services, transformer makers). EXC sits in the first bucket but with mixed-quality jurisdictions: Maryland and Pennsylvania are constructive, Illinois has been historically difficult and remains under shadow of the ComEd matter, New Jersey and DC are average. The value pool is growing in absolute dollars (electrification, data centers, reliability investment) but flat to slightly compressing in margin terms (allowed ROEs trended down 50 bps over the last decade and have only partially recovered with rates).

Industry Verdict: Good.

Good, not Excellent: the regulatory franchise moat is real and durable, the demand backdrop is the best in twenty years, and the business model is well-understood. Not Excellent because returns on capital are structurally capped (3.85% ten-year average ROIC for EXC [scorecard] is the giveaway), the regulatory compact is asymmetric (regulators can compress returns faster than the business can defend them), and the value of the franchise is highly dependent on local political dynamics that the operator only partially controls.

Inversion

I am now a short-seller. The pitch:

1. The single event that kills this. A combination event — and it is more probable than the bull case admits. Federal Reserve cuts pause at 4-4.5%, the long bond settles at 5%, and within 18-24 months the regulated utility comp set re-rates to a 14-15x P/E (against EXC's TTM 18.96x [scorecard] and 10-year average 18.37x [scorecard]) as the dividend-yield-vs-Treasury spread compresses. Simultaneously, FERC reduces the transmission incentive ROE adder by 50 bps following the latest D.C. Circuit guidance, and Illinois ICC denies a meaningful portion of ComEd's next multi-year grid plan filing. The result: EPS ~5% below current consensus, multiple compression of 20%, and a stock at $30-32 within 24 months. That is a 30-35% drawdown from $46.50 — and the dividend does not save you, because the dividend yield re-rates with the multiple.

2. Why the moat is narrower than bulls think. Bulls describe the regulatory franchise as a 'wide moat.' It is not. It is a narrow moat because the moat protects the franchise but not the return. The franchise is unassailable; the allowed ROE is reset every 3-5 years per jurisdiction and is highly mean-reverting toward the cost of capital. Look at the actual numbers: ROIC 10-year average 3.85% [scorecard]. That is the moat in action — for a decade. A wide moat produces 15%+ ROIC. EXC produces 3.85%. Bulls are confusing 'low business risk' with 'high economic moat.' Buffett distinguishes the two clearly [5]: capital-intensive regulated businesses are acceptable but explicitly not 'the best businesses by far for owners.'

3. Why management is worse than it appears. The headline is reasonable: clean post-spin focus, no buyback nonsense, dividend policy aligned with earnings. The substance has cracks. (a) The ComEd deferred prosecution agreement is not ancient history — it documents a culture, not an isolated act. The same governance regime that tolerated Madigan-era influence-peddling for years is the institutional muscle now negotiating Illinois rate cases. (b) Earned ROE has lagged allowed ROE by 50-150 bps across the portfolio for multiple years; that is operational underperformance, not 'regulatory lag' as the IR deck claims. (c) The negative FCF conversion of -1.57x [scorecard] is being narrated as 'growth investment' but in plain English the equity holder is funding part of the dividend by issuing new equity — a soft form of return-of-capital-as-return-on-capital that Buffett would call out [Munger excerpts on accounting]. (d) The 'data center demand' narrative is being aggressively monetized in investor communications as if EXC will capture outsized returns — but rate-base growth at allowed ROE delivers normal returns, not outsized ones.

4. What bulls are extrapolating that won't hold. Bulls are extrapolating: (a) 7.5% rate-base CAGR through 2030 — likely partially true but with downward revision risk if any major capex program is denied; (b) constructive regulatory outcomes everywhere — already not true in Illinois, with risk of spreading to NJ and MD as residential rates rise; (c) a 5-7% EPS growth annuity at a premium multiple — but the dividend-discount math at a 4% terminal growth rate, 9% discount rate, and current $1.60 dividend yields ~$32, not $46.50. The reverse-DCF implied growth of 9.42% [scorecard] is the smoking gun: the market is pricing the bullish extrapolation as if it were a base case.

5. Valuation trap (multiple compression / regime change). The setup has all the markings of a yield-stock trap. EXC trades at 18.96x TTM EPS [scorecard] — a multiple richer than its 10-year average of 18.37x [scorecard], at a moment when its long-bond competitor (the 30-year Treasury) yields more than its dividend (3.4% vs ~5%). The 'reach for yield' thesis is the bull case, and it works in a falling-rate world; it explosively reverses in a sticky-rate world. P/IV ratio of 1.69x [scorecard] against a deterministically-computed IV range — the price is above even the bull-case high IV of $40.85 [scorecard]. That is not a margin of safety. That is paying 14% above the optimistic case.

If I am right, the stock could be worth $28-32 within 18-24 months. That is roughly the IV base of $27.51 [scorecard] plus a small premium for the durable franchise. Downside-to-fair is roughly 32-40%. The bull case 'works' only via further multiple expansion or a sharp drop in long rates, neither of which I want to underwrite at this entry price.

Lollapalooza Bias Check

Biases active in me right now as the analyst:

Anchoring (strong). I am anchored on the deterministic IV range from the scorecard — $18.94 / $27.51 / $40.85 [scorecard] — and on the current price of $46.50 [scorecard]. That is by design (the brief instructs me to treat the scorecard as ground truth), but I should acknowledge that if maintenance capex were materially different from the 5-year average used in the calc, the IV could be 30-40% higher or lower. The scorer notes flag this explicitly: 'Maintenance capex uncertain (>50% spread); widen IV range' [scorecard notes]. I am treating $27.51 base as more precise than it deserves to be.

Authority bias (medium). I am leaning heavily on Buffett's framing of regulated utilities [1][3][4][5] because it cleanly fits this case. But Buffett owns unregulated-tax-advantaged-permanent-capital utilities (BHE), not public-equity utilities. His framework somewhat overstates the equity-holder's case for a public utility because it ignores the dilution and dividend-policy frictions that public utilities face. I should discount his framework slightly when applying it to EXC.

Recency bias (medium). Data-center demand growth in PJM is the loud story of 2024-2025. I am partially extrapolating that this is a 5-10 year tailwind. The honest answer is that it is a 2-4 year tailwind with high uncertainty about regulatory cost allocation that could partially neutralize the rate-base benefit.

Confirmation bias (medium). I started this analysis suspecting EXC would be a 'fine business at the wrong price' — that is a comfortable, contrarian-flavored conclusion for a value analyst — and the data is conveniently confirming it. I should be honest that the bull case (lower rates + sustained data-center capex + constructive regulation) could plausibly support today's price; I am not giving that case its strongest form.

Deprival super-reaction (mild). EXC's 3.4% yield is attractive in absolute terms, and there is a mild loss-aversion pull toward owning income-producing assets in an inflationary era. I should note this and discount it.

Inactive biases: social proof (utility space is unfashionable, no FOMO), commitment (no prior position to defend), incentive (compensation not tied to outcome here).

Net effect. The active biases pull in opposing directions: anchoring + confirmation push me toward 'pass,' while authority + recency + deprival push me toward 'hold or buy.' That partially cancels, but I lean toward acknowledging that the deterministic IV may be conservative and that my recommendation should be 'Hold/Avoid' rather than 'Sell' to honor the uncertainty around the IV range.

10-Year Outlook

Same fundamental business model in 10 years? Yes, with very high confidence. EXC will still be operating the same six wires-only utilities under the same cost-of-service regulatory framework. The shape of the business — rate base x allowed ROE — is essentially fixed in U.S. regulated utility law and changes only at glacial speed. This passes Munger's '10 years forward' test cleanly.

Larger customer base? Modestly yes, in customer count (1-2% growth via population growth in BGE/PECO/ComEd territories) but more meaningfully in load (5-15% over a decade depending on data-center realization, EV adoption, building electrification). Rate base growth of 6-8% CAGR translates to roughly $80-90B by 2034 versus ~$46B today.

Higher profit per customer? Yes in nominal terms (rates rise with capex and inflation), roughly flat in real terms. Allowed ROE will likely be in a band of 9.0-10.0% throughout the period. Earned ROE may improve modestly if operational efficiency catches peers.

Wider moat? No. The franchise stays equally durable; the allowed-return ceiling does not widen. If anything, distributed energy resources and behind-the-meter substitution narrow the long-tail moat slightly.

Single biggest threat? Regulatory backlash to data-center cost allocation in Maryland, Virginia (adjacent to Pepco), and DC. If commissions decide that hyperscaler load growth must be paid for entirely by hyperscalers and that residential ratepayers must be insulated, then EXC's rate-base growth assumption shrinks meaningfully. Secondary threat: a sustained 5%+ long-bond environment that re-rates utility equities back to 14-15x.

Confidence in 10-year compounding ability: I can predict rate-base growth, allowed ROE, and the dividend with reasonable accuracy. I cannot predict the multiple at which the market will price this stream — and the path-dependent return to a public-equity holder over 10 years depends as much on multiple as on EPS. Underlying economic compounding is medium-confidence; equity total return is lower-confidence.

CONFIDENCE: medium

Position Guidance

• Recommendation: Hold (lean Avoid at current price)
• Conviction: Medium
• Target buy price: $32 (roughly IV base of $27.51 plus a small premium for franchise durability; gives ~25-30% margin of safety to base IV)
• Target trim price: $42 (above bull-case IV of $40.85; further upside requires regime change, not business performance)
• Position sizing if buying at target: 2-4% of portfolio. This is a bond-substitute, not a compounder. Do not concentrate.
• Action today at $46.50: Pass. Revisit if the stock drops 25%+ on rate fears, an unfavorable Illinois rate case, or general utility-sector de-rating. The dividend yield (~3.4%) is not high enough to compensate for the ~50% downside-to-base-IV risk.