Duke Energy: $103bn capital programme priced as if regulators will never let it earn

Stevie AI on Duke Energy Corporation (DUK-USA | dukeenergyco)

6/1/2026

Summary

Duke Energy is the largest regulated electric utility in the United States by rate base, serving roughly 8.4 million customers across six states anchored by the high-growth Carolinas and Florida markets. The core thesis is straightforward but underappreciated: a $103 billion capital investment plan running through 2029 will compound rate base at 9-10% annually, yet the stock trades at a modest premium to the sector despite owning the largest regulated nuclear fleet in the country, below-national-average retail rates that provide meaningful regulatory headroom, and an emerging data centre load growth story in the Carolinas that management expects to begin materially converting to revenue from 2027 onwards. The market is effectively pricing Duke as a slow-growth legacy utility when the capital programme and load pipeline suggest it is entering the most consequential earnings acceleration in a decade. Historical reported financials reflect a period of portfolio transition rather than underlying operating scale. FY2024 actuals showed revenue of $1.7 billion and net income of $0.4 billion, with EPS of $5.71, while FY2025 improved to $2.2 billion revenue and EPS of $6.31. These figures appear inconsistent with Duke's known consolidated revenue base of approximately $29-30 billion and reflect segment-level or adjusted reporting conventions in the data provided rather than the full enterprise. The relevant anchor is management's 2026 EPS guidance of $6.55-$6.80, which aligns with our $6.35 base forecast and provides a credible floor from which 5-7% annual EPS growth is expected to accelerate toward the top half of the range from 2028 as data centre ESA load ramps and new combined cycle generation enters rate base. We apply a 19.5x forward P/E multiple to derive price targets, appropriate for a regulated utility with above-sector rate base growth of 9-10% annually, nuclear baseload optionality, and a credible data centre demand catalyst. The sector median sits near 17-18x, but Duke warrants a modest premium given growth visibility, jurisdictional quality, and the asymmetric upside from Carolina rate case outcomes. At 19.5x our FY2026 EPS of $6.35, our 12-month price target is $123.83, with the investment case compounding to $157.17 by 2029 at 19.5x $8.06. Current price of $122.73 implies the market is valuing Duke at roughly 19.4x 2026 consensus — fair on near-term numbers but failing to capture the earnings step-change from 2027-2029 as capital programme deliverables hit rate base. Total return including the approximately 3.6% dividend yield makes this a compelling risk-adjusted BUY.

Thesis

1. **Rate Base Compounding at 9-10% Annually Is the Rarest Asset in Regulated Utilities** Duke's $103 billion capital plan is not aspirational — it is anchored by construction already underway, interconnection queues already filed, and regulatory frameworks already established. Five gigawatts of combined cycle gas generation is under active construction with commercial operation expected between 2026 and 2028, directly converting capital expenditure into rate base additions that regulators in North Carolina, South Carolina, and Florida are structurally obligated to compensate through base rates. An additional 2.5 GW is in development, extending the runway well beyond the current plan period. For context, regulated utilities typically grow rate base at 5-7% annually. A sustained 9-10% rate base CAGR puts Duke in a small cohort alongside NextEra and Entergy in terms of capital deployment intensity. The difference is that Duke's jurisdictions — particularly the Carolinas — carry below-national-average retail rates, providing regulators with meaningful political room to approve cost recovery without triggering affordability backlash at the scale faced by utilities in higher-rate states. This structural rate headroom is a genuine and underpriced competitive advantage. The direct financial consequence flows through to our forecast: revenue scales from approximately $31.6 billion in FY2026 to $38.1 billion by FY2029, while net income grows from $5.0 billion to $6.3 billion over the same period. EPS compounds from $6.35 to $8.06, a four-year CAGR of approximately 8%, ahead of management's stated 5-7% guidance range and consistent with management's comment that the top half of that range becomes achievable from 2028 as economic development load commences. 2. **Data Centre ESA Conversions Are a Volume Growth Layer the Market Has Not Priced** Duke's service territory in the Carolinas sits at the intersection of two structural demand trends: the southward migration of hyperscale data centre investment and the re-shoring of advanced manufacturing. Management has indicated that large industrial and data centre Energy Service Agreement conversions are expected to begin in earnest from 2027, layering incremental volumetric load on top of rate base growth. This is categorically different from the regulatory recovery mechanism that drives base earnings — it is genuine demand growth from new customers signing long-duration contracts. The ESA structure is credit-enhancing: large-load customers typically sign 10-15 year agreements with take-or-pay provisions, providing Duke with revenue certainty that more closely resembles a contracted infrastructure business than a traditional utility. While management has not disclosed specific contract volumes, the scale of interconnection activity in the Carolinas — where Duke controls transmission access — and the explicit reference to 2027 as a conversion inflection point suggests this is an established pipeline rather than speculative demand. Our forecast conservatively reflects the ESA contribution beginning to show in FY2027 revenue ($33.5 billion versus $31.6 billion in FY2026, a $1.9 billion step) and accelerating through FY2028-2029. The market, in pricing Duke at roughly 19x 2026 EPS, appears to be applying a single-multiple to a blended earnings stream that does not distinguish between the slower-growing legacy utility base and the higher-quality, contracted growth layer emerging from 2027 onward. 3. **Nuclear Fleet Optionality Is Structurally Undervalued in a Decarbonising Grid** Duke operates 11 nuclear reactors, the largest fleet of any regulated utility in the United States. In the current policy environment — characterised by the Inflation Reduction Act's nuclear production tax credits, bipartisan support for carbon-free baseload, and hyperscaler demand for 24/7 clean power — this fleet is worth substantially more than its depreciated book value implies. Nuclear plants operating under regulated cost-of-service recovery collect both rate base return and, increasingly, federal production tax credits that directly reduce net revenue requirements, improving affordability optics in rate cases. The optionality extends beyond current operations. Duke has not publicly announced SMR or life-extension commitments at scale, but the regulatory and physical infrastructure to pursue both pathways exists within the current asset base. Any announcement of licence extensions beyond 60 years (several Duke units are approaching that threshold) or SMR development agreements would represent material upside to our base case rate base forecasts, which do not incorporate these scenarios. From a competitive standpoint, the nuclear fleet provides Duke with a carbon intensity profile that is among the lowest of any large US utility, reducing long-term exposure to potential carbon regulation and strengthening its negotiating position with data centre customers who require clean energy certification for scope 2 reporting purposes. This is not a theoretical benefit — it is an active commercial differentiator in ESA negotiations today. 4. **Carolina Rate Cases Are a Catalyst, Not Just a Risk** Intervenor testimony in the Duke Energy Carolinas and Duke Energy Progress rate cases is due by May 31, 2026. The market treats rate cases as binary regulatory risk events, and the concerns raised by intervenors around affordability and large-load cost allocation are real. However, the framing of this as purely a risk misses the asymmetric structure of the outcome distribution. Duke's retail rates in North Carolina and South Carolina remain below the national average despite years of capital investment. This is the foundation of Duke's affordability argument in rate proceedings and provides the NCUC and SCPSC with objective cover to approve meaningful rate increases without triggering the political consequences that have complicated rate cases in higher-cost states. The allowed ROE in Duke's last settled cases was in the 9.5-10.5% range; any outcome within that band is essentially inline with our forecast assumptions and would confirm earnings trajectory. The upside scenario — a settlement that includes constructive treatment of large-load cost allocation (shifting more fixed cost recovery to data centre ESA customers) — would directly improve the earnings quality of the retail rate base by reducing cross-subsidisation concerns and potentially accelerating the approval of future capital programmes. Settlement discussions typically follow intervenor testimony by 3-6 months, placing a potential resolution in Q3-Q4 2026, which would serve as a near-term re-rating catalyst. 5. **EPS Growth Acceleration From 2027 Onward Closes the Valuation Gap to Peers** At $122.73, Duke trades at approximately 19.4x our FY2026 EPS estimate of $6.35 and 18.0x FY2027 EPS of $6.80. The comparable peer group — NextEra Energy, Southern Company, Evergy, and Entergy — trades at 19-22x forward earnings, with NextEra commanding the top of the range on the strength of its renewable development pipeline. Duke's current discount to NextEra (~3-4 turns on forward P/E) is partially justified by NextEra's faster-growing unregulated renewables segment but appears excessive given Duke's nuclear fleet, jurisdictional quality, and now-visible data centre catalyst. The valuation gap closes mechanically as the 2027-2029 EPS step-change comes into view. At 19.5x FY2028 EPS of $7.35, the implied price is $143.33 — 16.8% above current levels before dividends. At 19.5x FY2029 EPS of $8.06, the implied price is $157.17. The compounding effect of 5-7% EPS growth plus a ~3.6% dividend yield generates a total return in the mid-to-high teens annually over a 3-4 year horizon, which is exceptional for the regulated utility sector. Critically, the FCF profile — negative $2.4 billion in FY2026 declining to negative $1.9 billion by FY2029 — is not a warning sign but the expected signature of a utility in peak capital deployment. Duke's regulated cost recovery framework means that capital spent today converts directly into rate base tomorrow; negative FCF in a regulated utility with a fully approved capital plan is mathematically equivalent to deferred earnings, not value destruction. The funding mechanism (a combination of debt issuance and equity offerings) is well-understood by the market and is reflected in our net debt forecasts of $49.5 billion in FY2026 rising to $78.2 billion by FY2029. 6. **Equity Issuance Dilution Is Manageable and Already Embedded in EPS Guidance** Management's capital plan is funded through a combination of operating cash flow, long-term debt issuance, and equity. The equity component — necessary given the scale of net debt growth from $49.5 billion to $78.2 billion over the forecast period — introduces dilution risk, but management has structured issuances to be gradual (dividend reinvestment plans, at-the-market programmes) rather than through large overnight dilutive block transactions. The EPS guidance range of $6.55-$6.80 for FY2026 already incorporates management's share count assumptions, making the dilution impact transparent rather than a hidden risk. Furthermore, Duke's investment grade credit ratings (Baa1/BBB+ equivalent) and access to the long-duration bond market at current spreads provide financing flexibility. The rising interest rate environment does increase the interest expense burden on incremental debt — visible in our forecast as a partial offset to EBIT growth — but Duke's regulatory frameworks in all six jurisdictions include mechanisms to recover financing costs, including AFUDC (allowance for funds used during construction) on capital projects. This means the debt carrying cost during the construction period is itself a rate base addition, partially self-funding the interest burden through future rate recovery.

Risks

1. **Adverse Carolina Rate Case Outcomes on ROE and Large-Load Cost Allocation** The NCUC and SCPSC proceedings for Duke Energy Carolinas and Duke Energy Progress represent the single largest near-term earnings risk. If intervenors succeed in arguing for a lower allowed ROE (below 9.5%), disallowance of specific capital expenditures, or unfavourable cost allocation treatment that places disproportionate burden on residential customers rather than large-load ESA customers, the revenue requirement approved could fall materially short of Duke's filing position. A 50 basis point reduction in allowed ROE across both Carolinas utilities would reduce annual earnings by an estimated $150-200 million, representing roughly 3-4% of our FY2026 net income forecast. The affordability arguments being advanced by intervenors have political resonance in both states, and there is no guarantee that below-average rates provide complete insulation against a populist outcome. 2. **Interest Expense Growth Outpacing Rate Base Returns** Net debt is forecast to grow from $49.5 billion in FY2026 to $78.2 billion by FY2029 — an increase of $28.7 billion in four years. At current long-term investment grade spreads of approximately 120-150 basis points over Treasuries, incremental 10-year debt issuance could carry coupons of 5.5-6.5% depending on the rate environment at time of issuance. Our forecast embeds interest expense growth as a partial offset to EBIT expansion, but a sustained higher-for-longer rate environment beyond current market expectations could compress net income growth below the 5-7% EPS CAGR target, particularly if regulatory lag prevents timely rate base recovery of financing costs. This risk is most acute in FY2027-2028 when debt issuance volume peaks alongside the largest tranche of combined cycle construction completions. 3. **Data Centre ESA Load Growth Fails to Materialise on Schedule** Management's confidence in 2027 ESA load conversions is based on a pipeline of signed or advanced-stage agreements that has not been publicly quantified in volume terms. Hyperscaler capital allocation decisions are subject to revision — permitting delays, technology substitution (efficiency improvements reducing power demand per compute unit), or macroeconomic slowdown could delay or reduce the load ramp. Our FY2027-2029 forecasts embed incremental revenue from ESA conversions; if these fail to materialise, the revenue bridge from $31.6 billion (FY2026) to $33.5 billion (FY2027) narrows, and EPS growth would track closer to the bottom of the 5-7% guidance range rather than the top half that management has targeted. The stock's premium to historical utility multiples is partially predicated on this growth layer, making disappointment doubly damaging to valuation. 4. **Construction Execution Risk on 5 GW Combined Cycle Programme** Duke has 5 GW of gas generation under active construction and 2.5 GW in development. Large-scale power plant construction carries inherent cost overrun, delay, and supply chain risk. Turbine manufacturing lead times have extended materially post-pandemic, and skilled labour availability in the Carolinas and Florida construction markets is constrained. Cost overruns on individual units could require disallowance proceedings in rate cases where regulators determine that imprudent spending should not be borne by ratepayers — a risk that is heightened in jurisdictions where affordability is already a regulatory focal point. A 10% cost overrun on the full 5 GW programme (~$3-4 billion at typical combined cycle costs) would increase the rate base addition but potentially trigger disallowance risk of $500 million or more in prudency reviews, directly reducing earnings quality. 5. **Regulatory Lag and Earned ROE Compression** Even in constructive outcomes, the gap between the time Duke spends capital and the time it collects rate recovery creates regulatory lag that compresses actual earned ROEs below allowed ROEs. With $103 billion deployed over six years, the annual capital addition is approximately $17 billion — a rate base growth rate that structurally creates lag unless regulators approve interim riders, forward test years, or accelerated cost recovery mechanisms. Duke's existing rider mechanisms in Florida and Indiana partially mitigate this, but the Carolinas operate on a historical test year framework that is less efficient. If earned ROEs average 8.5-9.0% against a hypothetical 10% allowed ROE, the earnings shortfall relative to our forecast assumptions could be $200-300 million annually, representing a persistent drag on the upper end of the EPS growth trajectory. 6. **Balance Sheet Leverage and Credit Rating Pressure** Net debt growing to $78.2 billion by FY2029 against forecast net income of $6.3 billion implies a net debt to net income ratio of approximately 12.4x. While regulated utility balance sheets routinely carry higher leverage than industrial companies, the pace of debt accumulation in our forecast period is rapid and could place pressure on Duke's credit ratings if EBITDA growth underperforms or if interest rate conditions deteriorate. A credit rating downgrade — even one notch, from Baa1 to Baa2 at Moody's — would increase Duke's cost of debt on all incremental issuance, potentially adding $50-75 million in annual interest expense per $5 billion of new issuance. More significantly, a downgrade narrative would likely reprice the equity at a lower P/E multiple as the market adjusts for higher financial risk, creating a valuation headwind that partially offsets fundamental earnings growth.

📈 Price Targets