Carolinas Settled at 9.8%, a Clean 10% Beat, and a Data-Center Pipeline That Added 200 Megawatts

Results vs. Consensus

Q2 2026 Scorecard

Consensus providers disagreed by a cent on earnings and by $61 million on revenue. The most widely carried earnings figure was $1.30, with the range running $1.29 to $1.32; on the low end the beat is 10.9%. On revenue the miss is 0.9% against the $7,659 million figure and 1.7% against the wider $7.72 billion estimate. A utility revenue line is a weak signal in either direction, because roughly a quarter of it is commodity cost passed straight through to customers at no margin, and this quarter is the clearest illustration of that in a year.

Year-over-Year Comparison

Sequential Comparison

The sequential decline is the shape of the business, not a signal. The first quarter carries the heating season and Piedmont's peak gas throughput, which is why regulated natural gas revenue falls by more than two thirds between the two quarters. The third quarter is the largest of the four, and the CEO framed it that way: "With our largest quarter still ahead of us, we remain firmly on track to achieve our 2026 guidance range of $6.55 to $6.80." Depreciation is the line to watch across the sequence, because it barely moves with volume and rose again to $1,700 million.

The Earnings Bridge

Two features of this bridge separate it from the first quarter. There is no operations-and-maintenance line at all, meaning that expense net of recoverable costs was neutral to the quarter after taking $0.09 out of the first. And volume contributed $0.08, more than the $0.03 it contributed in the first quarter, while weather turned from a $0.04 help into a $0.02 headwind. Rate cases and riders still carry the largest share at $0.18 combined, which is what a regulated utility is supposed to look like, but this is the first quarter under our coverage where the load itself did visible work.

Segment Performance

Electric Utilities and Infrastructure

The segment is 94.5% of consolidated revenue and produced $1,310 million of adjusted income on $7,175 million of revenue. Reported segment income of $1,271 million is after a $49 million impairment charge tied to the North Carolina settlement, which management excluded as a special item. Operating income rose $210 million, and the disclosed composition is worth reading closely, because two of the largest line movements are the same transaction seen from opposite sides.

On the revenue side the segment reports a $278 million decrease in storm recovery revenues at Duke Energy Florida. On the expense side it reports a $255 million decrease in operation, maintenance and other "primarily due to lower storm amortization at Duke Energy Florida". Those two are the roll-off of the 2024 hurricane cost recovery mechanism, and the net effect on operating income is a $23 million drag. Everything else is the underlying business: $145 million of higher fuel revenue that is offset in expense, $117 million of rate-case pricing, $84 million of weather-normal retail volume, $74 million of rider revenue and $52 million of higher wholesale revenue net of fuel.

Duke Energy Ohio net income covers the combined electric and gas business, so it is not directly comparable to the electric-only revenue column beside it. The two negatives in the right-hand column deserve more attention than they received. Duke Energy Carolinas, the largest utility in the fleet and the one that just settled a rate case, earned $36 million less in the first half than a year earlier on $427 million more revenue, as fuel rose $189 million, operations and maintenance $94 million, depreciation $143 million and interest $49 million. Duke Energy Indiana earned $13 million less on $148 million more revenue, with total sales down 5.5% and wholesale power sales down 31.0%.

Assessment: The segment is doing exactly what the regulated model promises, converting capital into earnings through rate cases and riders with a short lag, and the second quarter's mix was healthier than the first. The concern is beneath the aggregate. Two of the five utilities went backwards on earnings in the first half, and one of them is the franchise whose allowed return was just reduced. That is not visible in the segment line and it was not mentioned on the call.

Gas Utilities and Infrastructure

Revenue fell 8.9% to $449 million and segment income was $10 million against $6 million. The release describes this as flat, which is fair at this scale but understates what changed: this was the first full quarter without Piedmont's Tennessee business, sold to Spire on March 31 for approximately $2.5 billion. Management named the effect directly, saying results were driven by infrastructure recovery, "offset by lower earnings from the sale of Piedmont's Tennessee business." The segment's first-half reported income of $542 million is dominated by the $368 million pre-tax gain on that sale, which is excluded from adjusted earnings; on an adjusted basis first-half segment income was $371 million against $355 million a year earlier, an increase of 4.5%.

Assessment: Gas is now a rounding error in a summer quarter and a meaningful contributor in a winter one. The strategic question is not the earnings line but whether more of it gets sold. The Tennessee disposal fetched a price the parent could not have raised by issuing equity, and Duke Energy Ohio's gas reporting unit remains the only one in the group whose fair value did not materially exceed carrying value at the last annual goodwill test.

Other

The holding-company segment lost $204 million against $228 million, a $24 million improvement worth $0.03 per share. The CFO attributed it to "the expected benefit of lower interest expense resulting from the Tennessee and Florida transaction proceeds, which have reduced holding company financing needs as well as higher market returns." Interest expense at the parent fell to $313 million from $318 million while consolidated interest expense rose $60 million, which is the transactions doing precisely what they were sold as doing: moving financing need from the parent to the operating companies where it earns a return.

Assessment: This is the cleanest evidence so far that the asset-sale funding strategy works mechanically. It is also the smallest of the three effects, and it does not survive the arithmetic of the whole: consolidated interest expense is still rising 6.7% year over year on a debt balance that grew $993 million in the quarter.

Load and Operating KPIs

The weather-normal retail number is the one that matters, and at +1.3% it is the best print of the year against +0.4% for the first half. Residential grew 2.1% weather-normal on 1.6% more customers, which means usage per residential customer rose. Industrial fell 2.7%, or 0.5% weather-normal, on 2.1% fewer industrial customers. That combination, more households consuming more each and fewer factories, is the shape of the Southeast's economy and it is not the data-center story; contracted large-load volume has not started arriving.

Assessment: The generation mix moved in the wrong direction for a company running on an affordability platform. Nuclear output fell 6.7% and the fleet capacity factor dropped six points to 93%, with the shortfall picked up by coal, which rose 17.8%. Duke Energy Progress and Duke Energy Carolinas both cite higher nuclear outage costs in their expense variances. The cost of that substitution flows to customers through fuel clauses rather than to shareholders, which is why it never reaches the earnings bridge, and it is also why it was not discussed.

Key Topics & Management Commentary

Overall Management Tone: Settled and unhurried, and noticeably less defensive than the first quarter, when the entire call read as pre-positioning for a rate case. With the largest of those cases resolved, management stopped arguing the affordability case and started describing execution. The one place the posture stayed guarded was the growth rate itself, where repeated invitations to raise the long-term algorithm were met with a procedural answer about the normal fourth-quarter update rather than an appetite to commit.

1. The Carolinas Settlement: 9.8% and Half the Ask

On July 17 Duke Energy Carolinas filed a comprehensive settlement with the North Carolina Public Staff and the other intervening parties resolving every remaining revenue-requirement issue in the case it filed in November 2025. The terms are a 9.8% allowed return on equity with a 53% equity ratio, and a net retail revenue increase of $286 million in year one and $210 million in year two, $496 million in total, or 7.4% cumulative. The original filing asked for $1,002 million and a 15.0% increase at a 10.95% return. In June the company had already cut its own ask to $622 million at 10.48% in rebuttal testimony. The settlement therefore took roughly half of the original request off the table, and the concession on the return alone accounts for $133 million of the reduction.

"Last month, we were pleased to reach a comprehensive settlement with North Carolina Public Staff and other interveners in our DEC rate case, building on our long track record of collaborating with stakeholders to achieve constructive regulatory outcomes. ... The settlement includes a 9.8% ROE, 53% equity capital structure, and the continuation of the multiyear rate plan framework." — Harry Sideris, President and CEO

The number to hold on to is that 9.8% is lower than the 10.1% Duke Energy Carolinas has been earning under the 2023 order, and lower than the 9.99% Duke Energy Carolinas settled for in South Carolina in December. The allowed return in the company's largest jurisdiction went down. In May we set a base case of a negotiated settlement at 10.2% to 10.5% with a phased increase. That was too optimistic by forty to seventy basis points, and the historic rate base agreed in the settlement, $25.7 billion, is $800 million below the $26.5 billion filed.

Assessment: Read narrowly this is a worse outcome than we underwrote. Read against the alternative it is the removal of an unhandicappable tail from two thirds of the electric franchise, negotiated rather than litigated, with the intervening parties committed to pursuing "a substantially similar settlement framework" at Duke Energy Progress. A regulated utility that concedes 115 basis points and keeps its capital programme intact has bought certainty at a price it can afford. Duke reaffirmed 5% to 7% growth on the same call.

2. What the Settlement Bought Besides Certainty

Four provisions in the stipulation matter more than the headline return, and none of them was quantified on the call. The earnings sharing mechanism survives, allowing the company to earn 50 basis points above the allowed return before sharing begins, which effectively sets the realistic band at 9.8% to 10.3% rather than a hard 9.8%. The multi-year rate plan continues, now covering approximately $3.8 billion of North Carolina retail capital over two years, down from the $4.4 billion filed, and it now carries an annual refund mechanism keyed to how many projects and how much capital actually go into service against what was approved. The parties agreed to support a separate large-load tariff proceeding, to be completed before the new rates take effect. And the company may defer its next base rate filing to no earlier than November 2028 if it is permitted to defer costs on the Person County combined-cycle unit and the Marshall combustion turbines at a full weighted average cost of capital from the moment each plant goes into service.

"The agreement also retains the earnings sharing mechanism that allows us to earn 50 basis points above the allowed ROE up to 10.3%. Finally, we agreed to pursue discussions with intervenors to reach a substantially similar settlement framework for the DEP rate case." — Harry Sideris, President and CEO

The deferral provision is the most valuable and the least discussed. Regulatory lag on new generation is the structural cost of a build cycle this size, and full-cost-of-capital deferral from the in-service date removes it for the two largest units in the near-term programme. The refund mechanism cuts the other way, converting the multi-year plan from a commitment customers fund regardless into something closer to a true-up.

Assessment: The settlement trades headline return for structural protection, which is the right trade for a company whose earnings growth depends on deploying more than $1 billion a month for the next five years. A 9.8% return earned promptly on a larger base beats a 10.5% return earned two years late on a smaller one.

3. The Charge Was Guided at $40 Million and Landed at $51 Million

The July 17 exhibit told investors the stipulations were "expected to result in one-time pre-tax accounting charges of approximately $40 million, to be recognized by DEC in 2026." The quarter recorded $51 million of pre-tax charges within impairment of assets and other charges, split $29 million at Duke Energy Carolinas and $22 million at Duke Energy Progress. Net of a $12 million tax benefit the after-tax effect is $39 million, or $0.05 per share, and it is the entire gap between the $1.43 adjusted and the $1.38 reported figure. The consolidated impairment line went to $49 million from $3 million.

Two details are worth noting. The overrun is 28% against a figure furnished eighteen days before the quarter closed. And the charge appeared at Duke Energy Progress, whose own case had not settled as of the call, on the strength of the Duke Energy Carolinas framework.

Assessment: Immaterial in dollars and mildly informative in what it signals. A charge booked at Duke Energy Progress before its settlement is filed tells you management considers the Progress outcome close enough to determined that it can be accrued. The hearing was scheduled for August 11.

4. The Operations and Maintenance Line Reversed, and It Was Florida Storm Amortization

This was the single most checkable commitment from the first quarter. The line ran $253 million above prior year in the first quarter on Winter Storm Fern response and a legal settlement, and the CFO called the impact largely timing while holding to flat operations and maintenance for the full year. Flat for the year required the remaining nine months to run roughly that much below prior year in aggregate, and no bridge was offered.

The second quarter delivered a $270 million decline, taking the first-half line to $3,137 million from $3,154 million, down 0.5%. Flat, on schedule, one quarter early. The mechanism is not cost discipline. The Electric Utilities segment discussion attributes the $255 million segment-level decrease to lower storm amortization at Duke Energy Florida, where the first-half operations and maintenance line fell $226 million for that reason alone, and the same segment reports a $278 million decrease in storm recovery revenues. Duke Energy Florida's first-half revenue fell $42 million as a direct result.

The bridge that reaches earnings tells the honest version. In the first quarter, operations and maintenance net of recoverable costs cost $0.09 per share. In the second quarter the line does not appear in the variance table at all, meaning it contributed nothing. The June year-to-date table still shows $(0.09). The first-quarter drag did not reverse. It stopped growing.

Assessment: Management's promise was kept on the letter and not on the substance. Investors reading only the income statement will conclude Duke took $270 million of cost out of the business in a quarter, and it did not; it stopped amortizing a hurricane it had already been reimbursed for. The good news is that the underlying line was contained rather than blown out, and the storm roll-off continues into the second half, which makes the full-year flat claim easy from here.

5. Weather Was a Headwind in the Bridge and a Tailwind in the Level

These two statements are both true and they get confused constantly. Against the prior-year quarter, weather cost $0.02 per share, because the second quarter of 2025 was itself hot; Carolinas cooling degree days ran 14.5% above normal this year against 18.5% above normal last year. Against normal, the first half was helpful in both directions, a cold January and a hot second quarter.

"Favorable weather has also contributed to our strong results through midyear with a colder-than-normal first quarter, then quickly shifting to a hot second quarter. ... As we look forward to the back half of the year, we may have the opportunity to reinvest some of the weather benefits back into our generating facilities to ensure these assets continue operating well for our customers. This would be consistent with our demonstrated ability to exercise O&M agility in both directions over the past several years." — Brian Savoy, EVP and CFO

That sentence is the guidance. It tells you the first-half outperformance is banked, that management does not intend to flow it to the bottom line, and that the second half carries a discretionary spending programme sized to whatever the weather gave them. It is also a reasonable thing for a regulated utility to do: spending the windfall on generating assets is what keeps the next outage from becoming a fuel-cost problem for customers.

Assessment: Do not model the first-half beat through to the full year. Management has told you plainly that it will not, and the guidance arithmetic in the next section confirms it.

6. Data-Center Agreements Added 200 Megawatts

Executed electric service agreements went to 7.8 gigawatts from the 7.6 gigawatts disclosed in the first quarter. That is roughly 200 megawatts of new contracted capacity in a quarter that followed one adding 2.7 gigawatts. The high-confidence late-stage pipeline stayed at 15.4 gigawatts, unchanged for a second quarter, and the conversion timeline was restated as the first half of 2027 rather than the "next twelve months" framing used in May.

"We feel very good about our 15 gigawatts pipeline. They're advancing, but these negotiations are taking a little longer at times because they're very complicated transactions. So we want to continue to work through that. We feel very confident we'll be able to land more of those. We're looking at landing all of that 15 gigawatts by the first half of next year, and we're on track to do that." — Harry Sideris, President and CEO

The CFO framed the same pipeline in terms of what has to happen before capital moves, saying the incremental spending "will be triggered when the ESAs are signed and the requisite generation and transmission is modeled for those contracts." Duke also reported $5 billion of economic development wins in the first half supporting more than 9,000 jobs, and named life sciences and advanced manufacturing as sources of demand outside data centres.

Assessment: One slow quarter is not a broken pipeline, and management's explanation, that these contracts are complicated and take time, is credible for arrangements carrying minimum take provisions, credit support and termination charges. But 200 megawatts against a 15.4 gigawatt target that must be signed within four quarters is a pace problem, and the entire incremental capital case sits behind those signatures. This is now the thing to grade every quarter.

7. The $5 Billion to $10 Billion That Is Not in the Plan Yet

The most commercially significant new disclosure of the quarter came in the deck and was clarified in the first two questions: $5 billion to $10 billion of capital above the $103 billion 2026 to 2030 plan, concentrated in Indiana and Florida, contingent on converting the late-stage pipeline.

"And when we bracketed the $5 billion to $10 billion, we contemplated this is within the current 5-year plan. So we're going to roll forward the plan in February, which obviously capital is accelerating as we are investing more into the late part of the decade than we are right now. But this is incremental to that. So think about this as the 4 years left in the 5-year plan that we're executing today as capital upsides." — Brian Savoy, EVP and CFO

That answer is more useful than the number. It places the upside inside the four remaining years of the current plan rather than in the 2031 to 2035 window that a February roll-forward would reach, which means it would carry earnings inside the 5% to 7% guidance period rather than after it. At the top end it is roughly a 10% increase to the capital programme, and on a regulated utility's arithmetic a 10% larger rate base compounds into the growth rate directly.

Assessment: This is the strongest new element in the investment case and it is entirely conditional. Every dollar of it requires an executed agreement and a modelled generation and transmission plan, and this quarter produced 200 megawatts of executed agreements. The option is valuable; it is not yet an asset.

8. Indiana Is the Next Affordability Problem

Three of the six analysts on the call raised Indiana, on the state's affordability rhetoric, on whether a separate generation company structure would help, and on the fate of the Cayuga coal plant. Management's answers were consistent and general.

"Affordability is top of mind. Our customers are struggling with gasoline prices, rent prices, health care prices. So we share the Commission as well as the Governor's focus on affordability and making sure that customers are protected from these large loads and that we're providing reliable service and low-cost service to our customers, and we'll continue to do that." — Harry Sideris, President and CEO

On structure, management said it had examined a generation company in the past and "didn't feel like we were needing that to accomplish what we're doing", and that as large loads are signed "that may be something that we're going to revisit in the future to be able to provide financing as well as another layer of protection for our customers." A multi-year rate plan filing is coming. Meanwhile the utility's own numbers are weak: Indiana gigawatt-hour sales fell 6.8% in the quarter and 5.5% in the first half, wholesale power sales fell 31.0%, and first-half net income declined $13 million on a $148 million revenue increase.

Assessment: The pattern that produced the Carolinas outcome, a large capital ask into a commission and a governor focused on bills, is now assembling in Indiana, and Indiana is also where a meaningful share of the $5 billion to $10 billion of upside is supposed to land. The Carolinas settlement gives Duke a template and a precedent, and the precedent is a 9.8% return.

9. Cash Generation Improved Sharply, and Deleveraging Stopped

First-quarter operating cash flow covered 37% of capital expenditure, the weakest reading of our coverage. The second quarter produced $2,760 million of operating cash against $4,152 million of capital expenditure, a coverage ratio of 66.5% against 87.3% in the same quarter last year. For the first half, $4,272 million of operating cash covered 51.8% of $8,240 million of capital expenditure, against 78.4% a year earlier. Capital expenditure rose 28.2% year over year in the first half; operating cash fell 15.2%.

The balance sheet moved the other way. Total debt of $91,238 million against $56,863 million of equity puts debt at 61.6% of total capitalisation, against 61.5% at March 31 and 62.9% at December 31. In other words the entire improvement happened in the first quarter, funded by $5.3 billion of asset proceeds, and the second quarter gave a basis point back as debt grew $993 million against $385 million of equity. Cash flow from financing was $2,424 million for the half against $1,245 million a year earlier, and accrued capital expenditure not yet paid rose to $2,574 million from $1,929 million.

Management restated the funds-from-operations-to-debt target without changing it, and extended the horizon.

"We are on track to achieve our FFO to debt target of 14.5% for the year. Longer term, we expect to reach 15% FFO to debt as additional proceeds from the DEF minority interest investment are received. This FFO to debt target has substantial cushion to our downgrade thresholds, provides financial flexibility and serves as a solid foundation as growth accelerates later in the 5-year plan." — Brian Savoy, EVP and CFO

Assessment: The trajectory is right and the level is still uncomfortable. Half-funding a capital programme from operations is normal for a utility in a build cycle and is exactly why the Brookfield tranches and the equity programme exist. But the first-quarter deleveraging was a one-time event dressed as a trend, and the second quarter shows what the underlying direction is once the asset proceeds stop arriving.

10. Equity Funding, the Convertible, and the Growing Minority Interest

Duke has priced $600 million under its at-the-market equity programme so far this year, settling at the end of 2027, and ruled out anything larger.

"Richard, we're being very opportunistic with the equity issuances. And like I mentioned in my remarks, locking in attractive pricing when the market is there for us. So you could see us continuing to leverage the ATM as we move through the plan, the DRIP program and being smart about equity issues over time, but no large block equity planned in our 5-year plan." — Brian Savoy, EVP and CFO

Two dilution channels are running that the share count does not capture. The convertible notes issued in March 2026 struck at a 22.5% premium to the March 9 closing price and are excluded from diluted earnings per share only because they are currently antidilutive; the conversion right becomes live for holders in any quarter where the stock trades at or above 130% of the conversion price. And noncontrolling interests took $53 million of consolidated income in the quarter against $23 million, and $80 million in the half against $48 million, as Brookfield's 9.19% of Florida Progress earns its share with further staged closings running through 2028. Noncontrolling interests on the balance sheet grew to $2,112 million from $1,177 million at year end.

Assessment: The funding plan remains genuinely less dilutive than issuing common at 18.6 times earnings, which is the whole point of it. But the minority-interest line is now large enough to matter to reported growth, it grows mechanically with each Brookfield tranche, and it appears nowhere in the earnings-per-share bridge management presents.

11. Generation Build, Turbines, and Nuclear Discipline

The build target moved up to 15 gigawatts of additions by 2031 from the 14 gigawatts carried at the first quarter, reflecting Duke Energy Florida's latest ten-year site plan. Turbines available under the framework agreement with GE Vernova rose to 26, the first was delivered to the Person County combined-cycle site in July, and roughly 5 gigawatts of gas capacity is under construction with a further 2.5 gigawatts in development. Management repeated that a single engineering and construction partner moves crews between Carolinas sites and that the sites are built to identical specifications.

On nuclear the posture was disciplined in a way that is easy to under-appreciate in a market that rewards announcements. Duke has subsequent licence renewals approved for two plants and will file for Brunswick by year end, and is completing roughly 300 megawatts of uprates on the existing fleet. On new build the answer was a refusal.

"We want to continue to emphasize that additional financial protections are needed before we would propose a new nuclear project. Any structure to advance new nuclear must address first-of-a-kind and supply chain risks, provide financial risk protections for our customers and our investors and ensure a strong balance sheet during the construction cycle." — Harry Sideris, President and CEO

Assessment: The largest regulated nuclear operator in the country declining to announce a new nuclear project into the most receptive policy environment in forty years is a credibility asset, not a missed opportunity. Duke is buying gas turbines it can build on schedule and extending licences on reactors it already owns. The gap in the disclosure is operational rather than strategic: the existing fleet ran at a 93% capacity factor against 99%, and neither the release nor the call addressed why.

Guidance & Outlook

Implied second-half ramp. First-half adjusted EPS was $3.36 against $3.00 a year earlier, a 12.0% increase. That is 50.3% of the $6.675 guidance midpoint, against 47.5% of the $6.31 that Duke actually earned in 2025. Holding the midpoint therefore requires $3.315 in the second half against $3.31 delivered in the second half of 2025, which is growth of 0.2%. Reaching the top of the range requires $3.44, or 3.9% growth. The bottom of the range would allow the second half to decline 3.6%. A company that has just grown first-half earnings 12% is guiding to a second half that is flat at the midpoint, which is either substantial conservatism or a spending plan that has already been decided.

Street at. Consensus for 2026 sits at $6.72, four and a half cents above the midpoint and eight cents below the top of the range. On the Street's own number the second half needs $3.36, or 1.5% growth. The distribution of outcomes is narrow and the disagreement with management is small, which is normal for a regulated utility in August and tells you the debate has moved to the multiple and the 2028 inflection rather than to this year's number.

Guidance style. Duke has now reaffirmed rather than raised for two consecutive quarters while beating consensus in both, and it has beaten the published estimate in each of the last five quarters. The pattern is a company that banks outperformance rather than spending it on a guidance raise, and the CFO's remark about reinvesting weather benefits into generating facilities is the explicit statement of that policy. The fourth quarter is when the long-term growth rate normally gets revisited, and the CEO said as much when pressed.

What is not in the guide. Neither the $496 million Duke Energy Carolinas revenue requirement nor the pending Duke Energy Progress outcome was quantified against 2026 or 2027 earnings anywhere in the release or on the call. Year-one Carolinas rates are requested to take effect no later than January 1, 2027, and orders on both cases are expected by mid-November. Neither is the $5 billion to $10 billion of incremental capital in any published plan; the capital roll-forward comes in February.

Analyst Q&A Highlights

Whether the Long-Term Growth Rate Gets Raised

The opening question was the one the whole call turned on. With incremental capital identified, a large late-stage pipeline and management already claiming the top half of the range from 2028, the obvious follow-through is a higher long-term algorithm, and several peers have moved to a growth-plus framing. Management declined, offering a procedural answer about the normal fourth-quarter cadence rather than any appetite to commit, and redirected to execution on signing agreements.

Q: "So Harry, I mean, obviously, you guys are highlighting additional CapEx up to $10 billion. You've got 15 gigawatts in late stages. You're already sort of at the top end of the EPS CAGR. I guess how are you thinking about the 3Q update? Is there a point where we could see some upward pressure in the CAGR? And how -- I guess, how are you thinking about messaging around that, especially as many of your peers are now focusing on the plus part in their growth ranges."
— Shahriar Pourreza, Wells Fargo

A: "And we'll continue to evaluate where we need to be on our earnings per share growth rate, and we typically update that in the fourth quarter. But if anything changes materially like we did last year, we'll update you on that as we see fit. But our focus right now is to continue executing, getting those large loads signed to ESAs, making sure they're protecting our customers and paying their way as they go forward."
— Harry Sideris, President and CEO

Assessment: The refusal is informative in both directions. It confirms there is no near-term upgrade to the algorithm to trade on, and it also tells you management will not commit to more growth until the agreements that fund it are signed. Given that this quarter added 200 megawatts, that discipline is appropriate. The February capital roll-forward, not the third quarter, is the event.

How Much of the Incremental Capital Lands Inside the Current Plan

The single most useful answer of the call. The question was whether the newly disclosed capital upside would fall inside the five-year window that the 5% to 7% growth rate covers, or spill into the mid-2030s where it would be irrelevant to any investor's holding period. The answer placed it squarely inside, and attached a trigger condition.

Q: "So just on the potential $5 billion to $10 billion of additional capital for the large load in Florida and Indiana. I guess just as you guys -- or as we prepare for the roll forward in another 5 years, just how much of that do you think is eligible for like a 5-year plan versus being kind of well into the mid-2030s? How much of the $5 billion to $10 billion should we be thinking about can make its way into the roll forward?"
— Nicholas Campanella, Barclays

A: "And when we bracketed the $5 billion to $10 billion, we contemplated this is within the current 5-year plan. So we're going to roll forward the plan in February, which obviously capital is accelerating as we are investing more into the late part of the decade than we are right now. But this is incremental to that. So think about this as the 4 years left in the 5-year plan that we're executing today as capital upsides. And that will be triggered when the ESAs are signed and the requisite generation and transmission is modeled for those contracts."
— Brian Savoy, EVP and CFO

Assessment: This converts a headline number into a modellable one. Up to a tenth more rate base inside the guidance period is worth real basis points on the growth rate, and it is explicitly gated on executed agreements rather than on pipeline. The gate is the risk, and it is the same gate that produced 200 megawatts this quarter.

Indiana Affordability and a Possible Generation Company Structure

A recurring line of questioning tested whether Indiana's affordability politics push Duke toward a separate generation entity for large-load supply, which would sidestep the certificate process and make the customer savings more visible. Management had studied the structure before and rejected it, and for the first time signalled it is back under consideration as contracts get signed.

Q: "And then just maybe sticking with Indiana, there's been obviously a lot of rhetoric in the state around affordability. You've seen what's happening with the commissions. So I guess with the potential opportunities that you guys have to serve that large load, would you guys consider a genco type structure just given the benefits around maybe bypassing the CPCN process and flowing the savings back to customers a lot more visibly."
— Shahriar Pourreza, Wells Fargo

A: "On the genco side, we are looking at that. We have looked at that in detail in the past and didn't feel like we were needing that to accomplish what we're doing. But as these large loads are signed, that may be something that we're going to revisit in the future to be able to provide financing as well as another layer of protection for our customers."
— Harry Sideris, President and CEO

Assessment: A generation company is a financing structure as much as a regulatory one, and management named financing first. That is worth watching, because it would put large-load generation outside the rate base that the 5% to 7% growth rate is built on. The answer was a soft yes to studying it, which is a change from the first quarter, when the topic did not come up at all.

Indiana Regulatory Expectations Ahead of the Hearings

A separate question asked management to set expectations for Indiana on the eve of state hearings on affordability, with Duke earlier in the queue of contemplated cases than its peers. The answer was a list of the company's operating credentials rather than an expectation, and it avoided the state's demand picture entirely.

Q: "Maybe just to kick off a little bit more on Indiana. Just set expectations, if you can. Obviously, we're having some hearings later this week on the backdrop of affordability and implications. I'd just love your open-ended comments on that front, if you can here just at the outset. I'd really love to hear a little bit more on that specifically, if you can, especially given you guys are earlier in the slate of contemplated cases."
— Julien Dumoulin-Smith, Jefferies

A: "We start in a strong position. We have great reliability. We have great storm response. We have low cost in Indiana. We're very active in economic development in the state and been successful in bringing jobs and other tax benefits to the communities that we serve, and we continue to do that."
— Harry Sideris, President and CEO

Assessment: An answer about reliability and jobs to a question about rate-case risk. The Carolinas case is the precedent for how this ends, and it ended with the company conceding 115 basis points of allowed return. Duke Energy Indiana's first-half earnings fell while its revenue rose, and neither fact reached the call.

New Nuclear, and What Would Have to Be True First

Two of the six questioners pressed on new nuclear, on technology choice, on federal involvement, and on when a commercial outcome might appear. Management gave the same answer three times, refusing to attach a timeline and repeating a precondition rather than a plan.

Q: "And the timeline on even seeing commercial outcomes there. I mean, again, there's all sorts of noise in the system around this, if you set any kind of expectation on this. And obviously, you're doing the updates, et cetera, but on the core new nuclear."
— Julien Dumoulin-Smith, Jefferies

A: "We're focused on going through the process and making sure that we can offset that risk. So no real timeline, and we're not putting ourselves under pressure of a timeline. We want to make sure that we offset the risk first and foremost."
— Harry Sideris, President and CEO

Assessment: The most disciplined answer on the call. Duke holds a construction and operating licence at Lee for two AP1000 units and an early site permit for small modular reactors at Belews Creek, and it is choosing not to use either. In a market that pays for nuclear headlines, refusing to generate one is a signal about how management thinks about construction risk, and it should be read alongside the equally disciplined answer on gas supply.

Supply Chain for the Next Phase of the Gas Build

With six more turbines secured in the quarter, the natural question is whether the fuel to run them is contracted, and whether gas supply rather than equipment becomes the binding constraint. Management gave the most specific commitment of the call.

Q: "And then just on the additional 6 gas turbines that you secured this quarter as you think about the next phase of resource needs. Are you also in progress on securing the gas supply for any incremental gas plants as part of that next phase? And is that something you could see as a potential constraint to the build-out?"
— Carly Davenport, Goldman Sachs

A: "So we have all the gas that we need through early 2030 secured, and we continue to work with our vendors on providing additional gas beyond that. And we feel confident that we'll be able to nail that down as those generation projects get further in their build."
— Harry Sideris, President and CEO

Assessment: Fuel secured through the early 2030s covers the entire five-year plan and the first tranche of contracted data-center load. That is a genuine de-risking of the build and it received almost no attention. The residual exposure sits beyond 2030, which is also where the load ramp is heaviest.

The Shape of Cash Flow as Tax Credits Roll Down

The most forward-looking exchange concerned what happens to cash generation once the accelerated flow-back of nuclear production tax credits reaches parity, given how much of the near-term funding case rests on monetising credits. The answer described a hand-off rather than a cliff, and put a date on the crossover.

Q: "And then a separate question, I guess, for Brian, on just thinking about the long-term cash flow of the company, it seems like you're capturing a lot of the tax credit cash flow from the nuclear and the batteries in the near term. What happens in later years? Does that roll down and then kind of the cash flow from recovery of all these new investments ramps up? And so does cash flow stay stable, rising? Just how should we think about the kind of the shape of cash flow?"
— Steven Fleishman, Wolfe Research

A: "And it's going to catch up with earning the tax credits kind of late in the '20s. So 2028, 2029. We about hit parity on the nuclear PTCs, which is a huge contributor to the tax credit profile of Duke, and it's going to save cost for customers. And as we get into the early 30s, that will turn. But like you said, the earnings power on the investments we're making will more than offset that. So the cash generation continues to grow and it's durable well into the late '30s."
— Brian Savoy, EVP and CFO

Assessment: Parity in 2028 and 2029 coincides exactly with the year management claims the growth rate moves to the top half of the range, and with the first full year of contracted load. The tax-credit tailwind and the load tailwind are not additive; the second replaces the first. That is a coherent story and it should be modelled as a substitution rather than an acceleration.

What They're NOT Saying

1. Duke Energy Carolinas earned less: the flagship utility's first-half net income fell to $902 million from $938 million, on $427 million more revenue. The registrant's own management discussion sets out the drivers. Neither the release narrative nor the call mentioned it, in a quarter whose headline event was that utility's rate case.
2. The storm accounting behind the operations and maintenance headline: the words storm amortization and storm recovery revenue do not appear anywhere in the call. Storm appears once, in the phrase "great storm response." The $255 million expense decline and the $278 million revenue decline that produced it are disclosed only in the segment discussion in the 10-Q.
3. Volumes, at all: the transcript contains no reference to gigawatt-hours, sales volumes, or wholesale sales in any form, by management or by any analyst. For a utility whose entire investment case rests on load growth, an earnings call with zero discussion of how much electricity it sold is a conspicuous omission. Indiana sales fell 6.8% in the quarter and wholesale power sales fell 31.0% in the first half.
4. Nuclear availability: fleet capacity factor fell to 93% from 99% and nuclear output fell 6.7%, with coal generation up 17.8% to fill the gap. Duke Energy Carolinas and Duke Energy Progress both cite higher nuclear outage costs in their filed expense variances. The call did not raise it and no analyst asked.
5. What the settlement is worth: the $496 million revenue requirement, the $286 million and $210 million year-one and year-two splits, and the reduction of the multi-year plan capital to $3.8 billion from $4.4 billion were all absent from the call. Management gave the return on equity and the equity ratio and stopped there. The revenue-requirement arithmetic reached investors only through the July 17 exhibit.
6. A $1,046 million cash-flow item: the other-assets line in the first-half cash flow statement consumed $1,046 million against $78 million a year earlier, a $968 million year-over-year swing and the single largest negative item in the reconciliation. Current regulatory assets rose to $2,645 million from $1,934 million at year end. Neither movement was explained.
7. Funds from operations to debt still appears in no filed document: the 14.5% target for 2026 and the 15% longer-term figure were stated on the call for the second consecutive quarter and appear in neither the earnings release nor the 10-Q. Both are non-GAAP metrics with no published reconciliation, and both are the primary evidence offered for balance-sheet health.
8. Duke Energy Ohio's gas goodwill: the 10-Q repeats the standing disclosure that this is the only reporting unit whose estimated fair value did not materially exceed its carrying value at the August 31, 2025 annual test. The next test falls on August 31, 2026, three weeks after this call, and it was not mentioned.

Market Reaction

• Pre-print setup: DUK closed at $124.28 on August 3, up 6.0% year to date against 11.0% for the S&P 500, up 0.1% over twelve months and down 4.1% over the trailing thirty days. The 52-week closing range entering the print was $114.00 to $133.46, placing the stock roughly in the middle of its own year.
• Reaction session: Duke reports before the open, so August 4 was the reaction day. The stock gapped down to $123.51, traded between $121.80 and $124.87, and closed at $124.27, one cent below the prior close. Volume was 5.7 million shares against a 3.9 million thirty-day average, 1.4 times normal.
• Relative move: the S&P 500 rose 1.8% on the same session. A 10% earnings beat therefore produced roughly 180 basis points of relative underperformance.
• The settlement was already in the price: the Duke Energy Carolinas comprehensive settlement was furnished after the close on July 17. The stock rose 0.7% to $125.85 on the following session against a 0.2% decline in the S&P 500, then gave back 1.2% between that close and the pre-print close.

The flat tape is the honest verdict on the quarter, and it is more informative than the beat. Every element that would ordinarily move a utility on results day had already been disclosed or was structurally unable to move the number. The settlement, the largest event of the quarter, was public eighteen days earlier and was received with a 0.7% rally at the time. The earnings beat came with no guidance raise, and the guidance arithmetic makes clear why. The one genuinely new disclosure, $5 billion to $10 billion of capital above the plan, is contingent on agreements that added 200 megawatts in the quarter.

What the session priced, in other words, is a company whose near-term earnings power is fully known and whose incremental value sits behind a signature. Elevated volume at an unchanged price is consistent with position rotation rather than repricing: holders who owned the stock for the rate-case resolution had their answer, and holders who want the data-center story were given a target date of the first half of 2027 rather than a contract. Underperforming a 1.8% market day on a 10% beat is a de-rating, and at 18.6 times the guidance midpoint against 19.1 times after the first-quarter print, the multiple has done exactly that across the two quarters.

Street Perspective

Debate: Does a 9.8% Return on Equity Break the Earnings Algorithm?

Bull view: the bull case on the Street is that the settlement is a non-event for earnings because what matters is the rate base and the speed of recovery, not the headline return. Duke reaffirmed 5% to 7% growth on the same call, kept the multi-year rate plan, kept an earnings sharing mechanism that allows 10.3%, and won the right to defer costs on two major plants at a full cost of capital until 2028. On this reading, Duke traded a number that appears in headlines for provisions that appear in cash flow.

Bear view: the bear camp contends that a company which halves its own revenue ask and accepts a lower allowed return than it already had has demonstrated its pricing power is capped by politics rather than by cost of capital. If North Carolina, historically among the most constructive commissions in the country, produces 9.8%, then Indiana and Ohio will not produce more, and the entire industry's capital-in-equals-earnings-out arithmetic is being repriced downward at exactly the moment capital plans are at their largest.

Our take: the bulls have the better of it for this cycle and the bears have the better of it for the next one. The 2023 case delivered $768 million over three years, about $256 million a year; this one delivers $496 million over two, about $248 million a year, at a lower return but on a larger base and with the November 2028 deferral option attached. The algorithm survives. What has genuinely changed is the terminal assumption: nobody should now model a Duke jurisdiction at above 10% allowed return, and the growth rate has to come from rate base rather than from rate of return.

Debate: Is the Data-Center Pipeline Slipping, or Just Slow?

Bull view: some desks argue that 200 megawatts in a quarter is noise in a business where individual contracts run to gigawatts and take a year to negotiate. The pipeline is unchanged at 15.4 gigawatts, the customers are in vertical construction, management reiterated conversion by the first half of 2027, and the contracts carry minimum take provisions, credit support and termination charges that justify the negotiation timeline. Lumpy is not the same as stalling.

Bear view: the skeptics note that the pipeline has now been 15.4 gigawatts for two consecutive quarters, that the conversion language moved from twelve months to the first half of 2027, and that the CEO volunteered that "these negotiations are taking a little longer at times." They also point out that the entire $5 billion to $10 billion of incremental capital, and the claim of top-half growth from 2028, both depend on that conversion happening on schedule.

Our take: one quarter does not make a trend, and the honest answer is that this quarter produced no evidence either way beyond a slow number and a slightly softer commitment. The right response is not to change the thesis but to change what gets graded: contracted gigawatts is now the single metric that determines whether the multiple is justified, and four quarters of 200-megawatt additions would leave the 2028 inflection unfunded. We would rather see the pipeline convert at 2 gigawatts a quarter than see it grow.

Debate: Is 18.6 Times Right for a Utility Growing 5% to 7%?

Bull view: a growing consensus view holds that Duke deserves a premium to the regulated group because its growth is contracted rather than forecast, its jurisdictions are demographically the best in the country, and it has funded a quarter of the plan without issuing a share. Add a 3.5% yield to a 6% growth algorithm and the base return is close to 10% before any re-rating, with a February capital roll-forward as the catalyst.

Bear view: the bear camp points out that 18.6 times for 5% to 7% growth is not obviously cheap against a group trading in the high teens, that the load which justifies the premium does not arrive until 2028, and that the funding gap between now and then runs through a capital programme half-funded from operations. Buying growth that starts after the guidance period at a multiple that already embeds it is the definition of paying up.

Our take: at 19.1 times in May we agreed with the bears, and we said so. At 18.6 times, with the rate-case tail removed, the growth rate reaffirmed and the yield at 3.5%, the arithmetic has moved. A constant multiple delivers roughly a market return; the case for outperformance rests on a modest re-rating, and the catalysts for one are dated: North Carolina orders by mid-November, the Duke Energy Progress settlement, and the February capital roll-forward carrying the incremental $5 billion to $10 billion. That is three identifiable events inside twelve months against a stock that has gone nowhere in a year.

Model Framework & Valuation

The table revises the framework we established at initiation. Where an assumption is unchanged, the reason is stated rather than repeated.

Valuation. At the August 4 close of $124.27 the stock trades at 18.6 times the $6.675 guidance midpoint and 18.5 times the Street's $6.72, with a 3.5% dividend yield on the newly raised $4.34 annualised rate. Carrying the middle of our 5.5% to 6.5% growth range off the 2026 midpoint gives approximately $7.08 for 2027. Holding the multiple constant at 18.6 times that figure produces $131.70, a 6.0% price return, which with the yield is roughly a market return and is precisely the arithmetic that made this a Hold in May.

The upgrade rests on the multiple rather than the estimate. We think a resolved regulatory position, a reaffirmed growth rate and three dated catalysts inside twelve months support 19.5 times, which on the same $7.08 gives $138. From $124.27 that is an 11.0% price return, and with the 3.5% yield a total return in the mid-teens. The 52-week closing high is $133.46, so this is explicitly a call for the stock to exceed its own twelve-month range, and it sits in the middle of the $135 to $145 band where sell-side targets have clustered since the print.

What would take us back to Hold. Two more quarters of sub-gigawatt additions to executed service agreements, which would push the load ramp past the guidance period and strand the incremental capital case. A Duke Energy Progress order materially worse than the Duke Energy Carolinas framework, which would mean the settlement template did not hold. Or a move back above 19.5 times without a corresponding upgrade to the growth algorithm.

Thesis Scorecard Post-Earnings

The scorecard grades the standing thesis established at initiation in May, pillar for pillar. Status tags in the right-hand column are the ones carried in the thesis of record.

Overall: strengthened. Of the five standing bear points, two improved, one deteriorated and two were unchanged, and the largest of them resolved without costing the growth algorithm. A sixth has been added for the below-the-line leakage that management does not bridge. The offset is that Bull 2, the contracted-demand pillar that justifies the premium multiple, moved to AT RISK on a single slow quarter of contracting. The thesis is now less exposed to regulatory tail risk and more exposed to execution on one measurable variable.

Action: buy. We are upgrading to Outperform with conviction at 7 out of 10 and a twelve-month framework of $138 on 19.5 times a 2027 estimate of $7.08, against $124.27. The rating is a judgment that a resolved regulatory position, a 3.5% yield and three dated catalysts inside a year are worth more than the market is paying at 18.6 times, and it is not a judgment that the data-center story has been proven. If executed service agreements do not resume converting at scale by the fourth quarter, the premium argument goes away and so does the rating.