Duke Energy: $103bn grid build priced like a bond, but the data center load wave hasn't hit the income statement yet

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

4/5/2026

Summary

Duke Energy is the largest fully regulated U.S. utility by capital deployment, running a $103 billion infrastructure programme through 2030 across near-monopoly franchises in the Carolinas, Florida, and the Midwest. The structural insight is straightforward but underappreciated: Duke's rate base is compounding at 9.6% annually, yet the stock trades at a forward multiple that reflects bond-proxy sentiment rather than the accelerating earnings inflection visible in the 2026–2028 EPS trajectory. The market is pricing Duke as a slow-growth income stock at a moment when 4.5 GW of signed data center Energy Supply Agreements and a 9 GW pipeline are beginning to translate into 1.5–2% incremental retail load growth — load growth that, in a regulated utility framework, flows almost entirely into authorised returns on an expanding rate base. This is not a speculative growth story; it is a regulated earnings ramp with unusually high visibility. Recent financial performance has been distorted by reporting discontinuities. FY2023 actuals show $29.1 billion in revenue and $3.54 EPS, while the FY2024 figures ($1.7 billion revenue, $5.71 EPS) reflect segment reclassification and divestitures rather than an underlying business collapse — the regulated electric segment generated approximately $1.9 billion and regulated gas $1.7 billion in FY2024, with unregulated operations essentially wound down. The cleaner read is management's 2026 EPS guidance midpoint of $6.675, which is anchored to South Carolina Phase 2 rate implementation in Q1 2026, the North Carolina multiyear rate plan from January 2027, and continued grid rider growth in the Midwest and Florida. That guidance carries unusually high credibility given the cost-of-service regulatory framework and the mechanistic relationship between capital spend and rate base accretion. We apply a 20x forward P/E multiple to derive price targets, appropriate for a regulated utility with a 5–7% EPS CAGR, Sunbelt franchise exposure, and a visible data center demand catalyst that peers without comparable ESA pipelines do not possess. At $132.22 current price against our FY2025 EPS estimate of $4.85, the stock trades at approximately 27x near-term earnings — optically expensive, but the 2025 figure understates run-rate earnings power because rate case timing means the full benefit of approved increases does not flow through until 2026. Applying 20x to our FY2026 EPS of $6.68 yields a price target of $133.60, rising to $143.20 on FY2027 EPS of $7.16 and $153.60 on FY2028 EPS of $7.68. The 2027–2028 targets, where data center load ramp begins to contribute materially, represent 8–16% upside from current levels plus a dividend yield of approximately 3.5–4%, for a total return profile that is compelling relative to the utility sector and attractive in absolute terms for yield-sensitive institutional holders rotating from fixed income.

Thesis

1. **Rate Base Compounding Is the Engine, Not a Hope** Duke's investment case begins and ends with rate base mathematics. A 9.6% annual rate base CAGR, sustained through 2030 via ~$13 billion of annual capital expenditure, mechanically produces earnings growth in a cost-of-service regulatory environment. Unlike earnings growth at technology or industrial companies, this growth does not depend on market share gains, product cycles, or competitive dynamics — it depends on regulatory approval of capital spend already in progress and on rate cases where Duke has a demonstrated track record of constructive outcomes. The $103 billion capital plan through 2030 is not aspirational; it is a pipeline of investments already approved or in advanced regulatory review, spanning grid hardening, grid modernisation, clean energy transition, and load interconnection infrastructure. The translation mechanism is direct: capital invested enters Construction Work in Progress (CWIP), which in many Duke jurisdictions earns a carrying return before entering rate base, then formally enters rate base upon project completion and begins generating authorised returns — typically in the 9.5–10.5% range — that flow to earnings through subsequent rate cases or formulaic grid riders. Depreciation and interest expense grow in tandem, but the regulated revenue increases are structured to cover these costs plus the authorised equity return. The result is a business where the income statement is essentially a lagged reflection of the balance sheet's rate base growth, with the lag determined by rate case timing rather than commercial uncertainty. FY2025 EPS of $4.85 rising to $6.68 in FY2026 and $7.16 in FY2027 is the direct output of this mechanism: the step-change between 2025 and 2026 reflects South Carolina Phase 2 implementation and the beginning of North Carolina multiyear rate plan recognition, not any change in business fundamentals. Investors who look at FY2025 earnings and apply a trailing multiple are systematically undervaluing the earnings power that is already locked in regulatory filings. 2. **Data Center Load Growth Is a Rate Base Accelerant, Not a Utility Cliché** Every major regulated utility now discusses data center opportunity in investor presentations, which has caused some analysts to discount Duke's pipeline as sector noise. This is a mistake. Duke's 4.5 GW of signed Energy Supply Agreements and 9 GW total pipeline is among the largest contracted data center load positions in U.S. regulated utilities, and it is concentrated in jurisdictions — the Carolinas and Indiana — where Duke already holds the monopoly franchise. This matters because new large load in a regulated utility territory does not simply add revenue; it adds revenue that requires capital infrastructure investment to serve, and that capital investment enters rate base, earns an authorised return, and expands future earnings capacity in a compounding loop. Management's characterisation of 1.5–2% retail sales growth beginning to ramp materially in 2027–2028 is conservative relative to the pipeline. The 9 GW total ESA pipeline, even at a 50% conversion rate, would represent load additions that are large relative to Duke's current total generation capacity of approximately 50 GW. Hyperscale data centre loads are also highly predictable — they run at high capacity factors, are creditworthy counterparties, and generate demand for grid services including reliability and backup capacity that require additional regulated infrastructure investment. The incremental load growth does not merely fill existing capacity; it creates the regulatory justification for accelerated capital recovery and new build approvals. The 2027–2028 EPS estimates of $7.16 and $7.68 embed only modest data center contribution relative to the pipeline potential. If conversion rates on the 9 GW pipeline exceed management's internal planning assumptions — plausible given AI infrastructure build acceleration — there is material upside to FY2028 and beyond that is not reflected in current consensus. 3. **Regulatory Jurisdiction Quality Is Underrated by the Market** Duke's three primary regulatory jurisdictions — North Carolina, South Carolina, and Florida — are materially above average in regulatory quality relative to U.S. utility peers. The Carolinas Public Utilities Commission and Florida Public Service Commission have both approved constructive multiyear rate plan frameworks, grid hardening cost recovery mechanisms, and formulaic riders that reduce the lag between capital deployment and earnings recognition. This regulatory construct is not universal; utilities in Ohio, Illinois, or California face adversarial commission dynamics that create earnings uncertainty. Duke's Sunbelt franchise operates in a political environment where economic growth and job creation from data centre investment create bipartisan support for infrastructure spending. The North Carolina multiyear rate plan filing — covering the combined Carolinas utility post-merger — is the single most important near-term regulatory event. A settlement outcome (vs. litigation) would accelerate the implementation timeline for rate relief beginning January 1, 2027, and remove a key source of analyst uncertainty around FY2027 EPS delivery. The merger itself, combining Duke Energy Carolinas and Duke Energy Progress, creates operational synergies estimated in excess of $500 million over 10 years, and the combined entity's larger rate base simplifies rate case administration. The regulatory calendar for 2026 — South Carolina Phase 2 in Q1, North Carolina decision by Q3 — provides a series of well-defined positive catalysts that the current stock price does not fully reflect. Florida is an additional underappreciated positive. Duke Energy Florida operates under a multiyear rate plan through 2026 that includes base rate increases, grid investment recovery, and storm hardening provisions — all without requiring annual rate case filings. Florida's population and commercial growth dynamics mean load growth there is secular and does not depend on data centre contracts specifically. 4. **The Capital Structure Is a Feature, Not a Bug, for Long-Term Holders** Duke's balance sheet is admittedly leveraged: net debt rises from $83.1 billion in FY2025 to $107.1 billion by FY2028, and free cash flow is persistently negative (ranging from -$4.5 billion in FY2025 to -$3.8 billion in FY2028) because the ~$13 billion annual capex programme cannot be funded from operating cash flow alone. This requires approximately $1.5–2 billion per year in equity issuance and $10–12 billion in long-term debt issuance annually. Equity issuance is dilutive, which is why EPS growth is lower than net income growth in some years — the denominator is expanding alongside the numerator. However, this structure is entirely standard for large regulated utilities in heavy capital deployment phases, and it is manageable precisely because the capital is earning a regulated return. The debt is investment-grade (Duke maintains Baa1/BBB+ ratings with stable outlooks), the regulated cash flows are highly predictable, and lenders and bond investors underwrite the capital plan with full visibility into the regulatory cost recovery framework. Rising interest expense is a real headwind — each 100 basis point increase in blended debt cost on a $100 billion net debt base is approximately $1 billion of incremental annual interest — but Duke's rate case filings include test-year interest expense, meaning higher financing costs are recoverable through the regulatory process rather than absorbed as a permanent margin drag. The negative FCF is also not a sign of business deterioration — it is the accounting consequence of investing more than current earnings, which is precisely what drives future earnings growth. The dividend (~$4.18 per share annual, ~3.2% yield at current price) is well covered by regulated earnings and is explicitly protected under management's capital allocation framework. 5. **Valuation Mispricing: The Market Is Applying a Rearward Multiple to a Forward Earnings Step** At $132.22, Duke trades at approximately 27x FY2025 EPS of $4.85 — which appears expensive relative to regulated utility peers trading at 17–22x forward earnings. But this comparison is structurally misleading. FY2025 EPS is depressed relative to run-rate earnings power because it precedes the South Carolina and North Carolina rate implementations that become effective in 2026 and 2027. A more representative valuation anchor is the FY2026 EPS of $6.68, which management has explicitly guided to ($6.55–$6.80 range), against which Duke trades at approximately 19.8x — at the midpoint of the regulated utility peer range and below peers with comparable rate base growth profiles. The correct comparison set for Duke is not generic regulated utilities but specifically large-cap regulated utilities with above-average rate base growth (9%+ CAGR), Sunbelt franchise exposure, and credible large load pipelines. That peer set — including NextEra Energy, Entergy, and Southern Company — trades at 20–24x forward earnings. Duke at 19.8x FY2026 consensus represents a modest discount to that peer group, which we view as unjustified given the data centre ESA pipeline, the clarity of the North Carolina regulatory calendar, and management's extension of 5–7% EPS CAGR guidance through 2030. The discount likely reflects residual investor uncertainty around the North Carolina multiyear rate plan outcome and the timing of data centre load contribution — both of which are Q1–Q3 2026 resolution events. Applying a 20x multiple to our FY2026 EPS of $6.68 yields $133.60, essentially at current price. The investment case is therefore not a near-term re-rating story but a 2027–2028 earnings delivery thesis: at 20x FY2027 EPS of $7.16 and FY2028 EPS of $7.68, price targets of $143.20 and $153.60 respectively represent 8–16% capital appreciation plus a 3.5–4% dividend yield. Total return of 12–20% over a 2–3 year horizon is a strong outcome for a regulated utility with this level of earnings visibility. 6. **Management Track Record and Capital Plan Credibility** Duke's management team under CEO Lynn Good has executed the divestiture of non-core commercial renewables and midstream assets, allowing full capital concentration on the regulated utility franchise. The 2026 EPS guidance midpoint of $6.675 is consistent with the bottom-up rate base math and has been explicitly reaffirmed at investor events through 2025. The extension of 5–7% EPS CAGR from 2029 to 2030, with management expressing confidence in 'top half' (6–7%) delivery from 2028 as data centre load ramps, reflects a disciplined approach to guidance that does not over-promise ahead of regulatory approval. The $103 billion capital plan has been reviewed and accepted by all three major rating agencies without negative rating action, which is a substantive validation of the plan's financial sustainability. Duke's ability to issue $10–12 billion annually in long-term debt at investment-grade spreads — even in the current rate environment — reflects bond market confidence in the regulated business model. For equity investors, this credit market validation provides a second opinion on the capital plan's risk profile that is often overlooked in purely equity-focused analysis.

Risks

1. **North Carolina Multiyear Rate Plan Litigation Risk** The North Carolina multiyear rate plan filing — covering the combined Carolinas utility post-merger — is the highest-stakes regulatory event in Duke's near-term calendar. If the North Carolina Utilities Commission rejects the merger value-sharing proposal, imposes conditions that delay implementation, or denies the multiyear plan structure in favour of traditional annual rate cases, the timeline for FY2027 rate relief could slip by 12–18 months. The impact would be material: FY2027 EPS of $7.16 assumes North Carolina multiyear plan rates effective January 1, 2027; a delay to mid-2027 or 2028 could reduce FY2027 EPS by $0.30–0.50 relative to our estimate, pushing the earnings inflection into 2028. Litigation risk is elevated because affordability has become a primary political concern in North Carolina, and intervenors representing low-income customer advocates are likely to contest rate increases aggressively. The settlement vs. litigation path is the most important binary event in Duke's 2026 regulatory calendar. 2. **Interest Rate and Refinancing Risk on $100 Billion+ Net Debt** Duke's capital structure requires $10–12 billion of annual long-term debt issuance to fund the capex programme. At a net debt level rising toward $107 billion by FY2028, a sustained increase in investment-grade utility borrowing spreads or benchmark rates directly increases interest expense before regulatory recovery is possible. While Duke's rate cases include test-year interest expense, the lag between cost incurrence and regulatory recovery can be 18–36 months, creating a period of earnings drag. A 50 basis point increase in blended new issuance cost on $11 billion of annual issuance represents approximately $55 million of incremental annual interest expense — manageable in isolation, but compounding over multiple years and multiple issuances could represent $200–300 million of cumulative unrecovered interest before the next rate case cycle. The FY2025–2027 period is particularly exposed because multiple rate cases are in process simultaneously, and a rising rate environment could cause regulators to scrutinise the reasonableness of embedded debt costs. 3. **Data Centre Load Ramp Timing and Conversion Risk** The 9 GW data centre pipeline represents total opportunity, not committed load. The 4.5 GW of signed ESAs is binding in commercial terms but depends on customer construction timelines, permitting, and interconnection queue progression that Duke does not fully control. If hyperscale customers delay facility construction due to AI infrastructure spending cycles, capital market conditions, or regulatory permitting delays, the incremental load growth expected in 2027–2028 may not materialise on schedule. Management's 1.5–2% retail sales growth assumption for the forecast period is predicated on conversion of a meaningful portion of the 9 GW pipeline. A scenario where only 2–3 GW converts by 2028 — plausible given hyperscale project delays observed industrywide — would reduce FY2027–2028 EPS by $0.15–0.25 relative to our estimates and delay the 'top half' CAGR delivery management has projected. 4. **Equity Dilution Suppressing Per-Share Value Creation** Duke issues approximately $1.5–2 billion of equity per year to partially fund the capex programme — equivalent to roughly 10–13 million new shares annually at current prices, or approximately 1.3–1.6% annual dilution on a share count of approximately 770–780 million. Over a four-year forecast horizon, this represents potential cumulative dilution of 5–6% before accounting for any scrip dividend reinvestment. While the dilution is offset by rate base growth that expands earnings in aggregate, the per-share compounding math is less attractive than a self-funding business would be. If capital markets access deteriorates — through credit spread widening, equity market volatility, or a decline in Duke's stock price — the equity issuance required to maintain the capital plan becomes more dilutive per dollar raised, creating a potential spiral where programme financing becomes more expensive precisely when it is most needed. This structural feature means Duke's EPS CAGR of 5–7% is a net-of-dilution number, and gross earnings growth is somewhat higher — but it also means shareholder value creation is more dependent on capital market conditions than peers with stronger FCF generation. 5. **Affordability Pressure and Political Intervention in Rate Setting** Across Duke's service territories, residential electricity rates have risen materially in the post-pandemic period alongside broader cost-of-living pressures. The political salience of utility bills — particularly for low-income customers — has increased regulatory scrutiny of rate increase requests in all three primary jurisdictions. North Carolina, South Carolina, and Florida have each seen legislative or gubernatorial commentary on utility affordability that could translate into commission pressure to limit approved rate increases or require customer protections that constrain cost recovery. If regulators approve rate increases below the levels embedded in test-year filings — by, for example, denying certain capital projects, lowering authorised ROE, or imposing bill credit requirements — Duke's authorised earnings would fall below the 5–7% CAGR trajectory. This risk is difficult to quantify but is real and increasing: Duke's average residential bill in North Carolina has risen approximately 20–25% over five years, and the North Carolina Attorney General's office has become a more active rate case intervenor. 6. **Hurricane and Extreme Weather Event Risk in Florida and the Carolinas** Duke Energy Florida and Duke Energy Carolinas both operate in regions with material hurricane and severe weather exposure. While Duke has storm hardening programmes and securitisation mechanisms that allow recovery of storm restoration costs through regulatory mechanisms, extreme weather events create short-term earnings volatility, operational disruption, and regulatory friction. A major hurricane affecting Duke's Florida service territory — approximately 1.9 million customers — could require $1–3 billion of restoration spending, funded initially from storm reserve funds and subsequently through securitisation. The process takes 12–24 months, during which Duke carries the cost on its balance sheet and faces potential credit rating pressure. More subtly, repeated storm events create political pressure on rate increases for grid hardening — the very capital investment Duke is relying on for rate base growth — as regulators face the optics of approving large rate increases following customer outage events. FY2024 results were partially affected by weather normalisation items, and the FY2025–2028 forecast assumes normal weather; a materially above-normal storm season in any forecast year would reduce EPS below our estimates.

📈 Price Targets