Pinnacle West Capital reported third-quarter 2026 results on Tuesday, November 3, and the headline is a strong one: net income attributable to common shareholders of $434.2 million, or $3.49 per diluted share, against $413.2 million, or $3.39, a year earlier. The company raised full-year 2026 earnings guidance to $5.05 to $5.25 per diluted share, up from $4.55 to $4.75 — a 50-cent lift at both ends of the range, on the strength of customer and sales growth and favorable weather through the first three quarters. For a utility whose investment thesis is Arizona load growth, the quarter is the thesis working. The catch is that the thesis is now working faster than the regulatory machinery built to price it.
Pinnacle West is new to this publication’s earnings coverage — the first Arizona utility in the rotation, after Sempra, Entergy, Dominion, Eversource, and NorthWestern — and the quarter deserves the same two-number reading this publication’s earnings checklist prescribes. Start with the $21 million increase in net income, up 5.1%, against a 2.9% increase in earnings per share. The gap is the share count: new equity issued to fund the capital plan is already diluting the per-share result. Then read the composition. The release attributes the gain to higher customer sales and usage, favorable weather versus a year earlier, and customer growth — partly offset by higher depreciation and amortization and higher interest charges. The per-share bridge in the company’s earnings presentation is more precise: sales and usage added $0.35 a share and weather added $0.09, together $0.44 against a $0.10 rise in earnings per share, with a larger share count taking away $0.07. Weather is not load growth, and $0.09 of the quarter’s per-share gain is weather. Strip it out and the underlying improvement is thinner than the headline suggests.
The demand behind the quarter
The load numbers are the strongest in the release, and they are the reason the guidance raise happened. Weather-normalized retail sales grew 9.5% in the quarter, against 5.4% in the third quarter of 2025. Commercial and industrial sales rose 14.5% while residential sales grew 4.5%, and customer growth was 1.8%, including 1.9% residential. Arizona Public Service set an all-time peak demand record of 9,166 megawatts on August 2, topping last year’s peak of 8,648 MW; demand exceeded 9,000 MW on three occasions during the quarter, and APS passed the old peak a dozen times. It was the fourth consecutive year of record-setting summer demand. The quarter was also substantially warmer than normal — residential cooling degree days were up 3% from the same period in 2025 — so part of the sales gain is weather that will not repeat. But the weather-normalized figure is the one that matters, and 9.5% weather-normalized growth is not a weather story.
The forward view is more striking than the quarter. The earnings presentation puts committed load at 5.8 gigawatts, with more than 84% of it not yet served, and guides 2026 weather-normalized sales growth of 8% to 10% — including 5% to 7% from new large manufacturing facilities and several large data centers — with long-term sales growth of 8% to 10% a year through 2030. The release looks further out still: customers’ annual electricity needs are projected to more than double by 2041, while peak demand is expected to rise more than 5,000 MW above today’s peak. APS recently filed its integrated resource plan with the Arizona Corporation Commission, outlining the resources and transmission needed over the next 15 years. The load forecast is no longer a judgment about whether growth arrives; it is a schedule of named customers whose facilities are already under construction.
The TSMC contract
Three days after the earnings release, on November 6, APS filed with the Arizona Corporation Commission a proposed 25-year Direct Pricing Contract with TSMC Arizona — the agreement that makes the load-growth story concrete and, at the same time, puts it on trial. TSMC’s north Phoenix semiconductor fabrication campus would draw about 1.9 gigawatts at full buildout under the filing’s rounded figure — 1.88 GW in the company’s own earnings presentation — the electricity use of roughly 300,000 homes, and the largest single customer load APS has ever contracted. TSMC is the first customer on APS’s subscription model for extra-large energy users, a structure in which the customer commits to pay for the new generation and service infrastructure its load requires, rather than having those costs folded into the general rate base.
The terms, as APS describes them, are the heart of the “growth pays for itself” argument. Under the contract, TSMC Arizona assumes full long-term financial responsibility for 100% of the new power plants, transmission lines, equipment, and shared system usage required to serve its facilities, and APS says existing residential and business customers will pay nothing for TSMC’s energy growth. The commitments run whether or not the fabs reach full buildout. The new generation tied to the deal includes the proposed Desert Sun natural gas plant west of Gila Bend, which is proposed, not approved. The public description of the deal carries no rate level and no plant cost, and none is stated here. The contract was signed in October and filed for commission approval on November 6; it is subject to that approval, and the commission has not yet ruled.
The filing is the first live test of the Pledge for Responsible Energy Growth that Chairman, President and CEO Ted Geisler reiterated on the earnings call. The pledge has three principles — safeguarding reliability, protecting customer rates, and preserving capacity for Arizona’s homes, businesses, and communities — and its operative promise is blunt: “Growth should pay for growth.” APS says it will not connect new extra-large customers faster than the grid can reliably support, will require those customers to fund the costs of serving their growth, and will preserve sufficient capacity for existing and future customers. The TSMC contract is the pledge reduced to a single document, and the commission’s ruling on it will set the template for every data center and manufacturer that follows.
The rate case: the other half of the story
The TSMC filing arrives alongside a rate case that is, in its own way, the more consequential proceeding. APS filed its current rate application on June 13, 2025, seeking a net revenue increase of $579.52 million, or 13.99%, on a test year ending December 31, 2024, and a return on equity of 10.70%. Commission staff recommended a return in the range of 9.55% to 9.80% — a range that starts at the 9.55% return now in rates, set in the commission’s 2024 decision. The gap between 10.70% and 9.55% is the whole ballgame for a utility whose rate base is growing as fast as APS’s: every tenth of a point of allowed return, applied to a $21.6 billion proposed fair value rate base, is real money, and the spread between the ask and the staff recommendation is worth roughly a point of return on equity.
APS is also asking the commission to adopt a formula rate plan, under which rates would adjust annually between general rate cases. On the earnings call, management indicated that earning within 50 basis points of the authorized return will take time, with the first formula rate filing not scheduled until May 2027 — the regulatory lag problem stated plainly. The rate case decision is expected in December 2026, though the presentation notes the dates are tentative and subject to change. The decision is expected, not certain, and the timing matters because the TSMC contract and the rate case will be decided by the same commission in the same season.
A commission that just changed
That commission is not the one that set APS’s current rates. On November 3 — the same day Pinnacle West reported earnings — Arizona voters elected Democrats Clara Pratte and Jonathon Hill to the two open seats on the five-member Arizona Corporation Commission, ending its all-Republican makeup. Republicans retain a 3–2 majority, and the results remain unofficial until canvassing and state certification, but the composition has changed in the direction that matters for APS: both winning candidates campaigned on the argument that the commission has been too deferential to the state’s utilities, and Pratte in particular campaigned against passing large-customer costs to residential ratepayers.
The timing is the story. The rate case decision is expected in December, weeks after the election, and the TSMC contract will be reviewed by a commission whose newest members have already stated their skepticism in public. That does not mean the commission will reject either filing — the TSMC contract is structured precisely to answer the cost-shifting objection, and the rate case record was built under the old commission — but it does mean the “growth pays for itself” model will be tested by the people most inclined to test it, at the moment its first and largest application is before them.
The capital plan and the funding behind it
The growth the quarter documents has to be built, and the build is large. APS’s capital spending totals $10.35 billion for 2025 through 2028, before a Cholla gas conversion expected to cost up to about $440 million. The presentation also lists an investment of about $200 million on a buyout option for nearly 100 MW of Palo Verde nuclear capacity previously under sale-leaseback, plus about $500 million of added Palo Verde capital spending over the next 10 years. The buyout shows up in the quarter’s per-share bridge as a $0.03 contribution from the purchase and end of two of the three Palo Verde lease agreements.
The financing plan is where the growth story meets the balance sheet. Pinnacle West’s plan includes a new equity need of $1.0 billion to $1.2 billion, of which about $920 million is priced under equity forwards through October 2026, and the company has started a new at-the-market equity program with about $235 million already priced. Guidance for 2026 interest expense, net of allowance for funds used during construction, is $410 million to $430 million. The equity issuance is the mechanism behind the $0.07 per-share drag in the quarter’s bridge, and it will continue: a capital plan of this size, funded at roughly half equity, means dilution is a standing feature of the growth story, not a one-quarter event.
The guidance math
Do the arithmetic the checklist prescribes. Nine-month net income attributable to common shareholders stands at $645.7 million, against $601.1 million a year earlier. The raised full-year guidance of $5.05 to $5.25 per share implies a fourth quarter that must carry the year’s momentum without the summer’s weather help — the third quarter is seasonally the strongest for an Arizona utility, and the fourth is not. The guidance raise is a 50-cent lift at both ends, but it is also a statement about the second half of the year: management is telling the market that the sales growth seen through September, not just the weather, will persist into the winter.
The durable-versus-one-off ledger is the right way to close the quarter. Durable: weather-normalized sales growth of 9.5%, commercial and industrial sales up 14.5%, customer growth of 1.8%, a 5.8 GW committed load queue, and a capital plan of $10.35 billion through 2028. One-off or timing: $0.09 a share of weather in the per-share bridge, the $0.03 Palo Verde lease benefit, and a guidance raise that leans on nine months of favorable weather as much as on load growth.
What to watch
Five markers, in order of consequence. First, the December rate case decision: the allowed return on equity, and whether the commission adopts the formula rate plan that would start closing the earned-versus-authorized gap in May 2027. Second, the TSMC contract proceeding: whether the commission approves the 25-year Direct Pricing Contract on the terms filed, and what conditions, if any, it attaches. Third, the new commissioners’ first votes: whether Pratte and Hill’s campaign skepticism shows up in the rate case decision or the TSMC review. Fourth, the fourth-quarter print: whether the $5.05 to $5.25 guidance range holds without summer weather. Fifth, the load queue converting to billed sales: whether the 5.8 GW of committed load materializes on the schedule the guidance assumes.
Pinnacle West’s quarter is a load-growth quarter: record demand, a guidance raise, and a capital plan sized to a doubling of the state’s electricity needs by 2041. The TSMC filing is the quarter’s real news — a 1.9 GW customer willing to pay for its own generation, filed with a commission that has just changed its makeup, in the same season as a rate case that will set the return on everything APS builds.
None of this is investment advice, and nothing here recommends buying, selling, or holding any security. It is a reading of one quarter’s evidence: a $3.49 print against $3.39 a year earlier, a guidance range raised to $5.05–$5.25, weather-normalized sales up 9.5% on a record 9,166 MW peak, a 25-year contract for roughly 1.9 GW of semiconductor load, and a rate case decision expected in December from a commission whose newest members campaigned on testing exactly these deals.