Capital Power's Record Quarter Comes With a Footnote

Record EBITDA, falling AFFO, a raised capex bill and a CEO who says the company no longer needs M&A — the story is in what got de-emphasized

CPX.TO · Earnings Call · 2026-11-02

A record that isn't a cash record

Capital Power opened its third quarter with the word "record" — $553 million of adjusted EBITDA, up $76 million year over year, the best quarterly contribution in company history. Hummel Station and Rolling Hills carried it on stronger PJM power prices and summer demand, with Goreway and the Ontario batteries pitching in. Year to date, EBITDA is $1.308 billion, up $142 million. Then the footnote. AFFO fell $63 million to $306 million. Year to date it is down $34 million to $788 million despite the EBITDA gain — the clean-technology investment tax credits booked in Q2 ($174 million) largely offset by maintenance. Kevin MacIntosh was plainly aware of how that reads: “Despite higher adjusted EBITDA, AFFO reflects higher sustaining capital spending, mainly from a heavy maintenance schedule across our Alberta fleet, rotor replacements at Rolling Hills and Harquahala and planned maintenance at Hummel Station and Rolling Hills.” — Kevin MacIntosh, Chief Financial Officer · 2026-11-02 And the maintenance bill got bigger, not smaller. 2026 sustaining capital guidance moved to $350–370 million from $290–330 million, driven by unplanned work on the U.S. flexible fleet and a planned outage pulled forward from 2027 into Q4. Management frames the increase as "event-driven," with a "significant portion" tied to insurable events and OEM concessions expected to offset the cash impact. Believable — but it is now the third consecutive quarter in which the sustaining capital line moved the wrong way against a record EBITDA headline. The dividend told the same story in miniature: a 2% raise to $2.79 per share, explicitly at the low end of the 2%–4% through-2030 corridor, the 13th consecutive increase. Growth in the payout from $1.51 in 2016 is real — about 85% — but the pace has been deliberately throttled to fund something else.

The $1.25 billion that doesn't need to be bought

What that something else is: the embedded upside management now pegs at up to $1.25 billion of incremental annual adjusted EBITDA, up from the $1 billion framed at the 2024 Investor Day. Roughly $400–550 million of that is recontracting and upgrades across ~3 GW; another $375–700 million is merchant upside across ~5 GW tied to Alberta and PJM power prices. Approximately 3 GW of portfolio capacity rolls off starting 2029, and the company has already hedged over 40 TWh of PJM and Alberta power from 2027 forward — ~80% hedged over a three-to-four-year window. The crucial attribute is that it requires “no new shares, no new project debt and limited growth capital, which makes it highly accretive to AFFO per share.” — Avik Dey, Chief Executive Officer · 2026-11-02 That is a foundation of investment grade counterparties — 75% of EBITDA contracted, 80%–90% of that contracted base with investment-grade names, more than 70% with utilities or system operators, 9–11 year weighted average life.

I would not say a diminished focus on M&A. In fact, I think '27 is going to be more active than less, given that '26 was effectively a pause. But I would really call into focus that my comments were around the fact that we don't need it.
Avik Dey, Chief Executive Officer · 2026-11-02

That is the real shift this quarter, and it is a shift in framing rather than in numbers. Across the last four calls Avik Dey and Scott Manson described the M&A pipeline in escalating terms — in March, “it's a robust market for M&A, and there are a lot -- there's a number of opportunities in the marketplace for us” — Scott Manson, Executive (likely CFO or similar senior finance role) · 2026-03-04; in July, “The deal pipeline is bigger today than it was last year.” — Avik Dey, President and CEO · 2026-07-29 Now the language is regulatory moratoriums, a locked-up back half, a "pause," and a "more refined focus" on bilateral single-plant deals in PJM, with less ERCOT. The Apollo partnership is reaffirmed but still has not transacted. M&A has been demoted from growth engine to optionality — because the organic contracting opportunity in the existing fleet does the job.

Turbines, and a CEO refusing to call a trend

The freshest thing on the call was not a number. It was a nuance about equipment. Dey noted more credible listings of complete or unused turbine packages and the ability to swap delivery slots into 2028–2030, then spent two answers carefully fencing it in: “I wouldn't call it enough to pick the trend, which is what my comments were. But I think if we needed turbines on a specific project, we're seeing a better ability to capture them.” — Avik Dey, Chief Executive Officer · 2026-11-02 Turbine supply has been the binding constraint on every North American gas developer for two years; a CEO volunteering that it may be loosening — while insisting it is not a turn — is precisely the kind of pre-consensus datapoint that precedes new capacity decisions. The jurisdictional backdrop improved too. Alberta's October 19 referendum endorsed remaining in Canada — Dey called it "a great day for Capital Power as a strong and proud Alberta company" — and management expects AESO Phase 2 resolution "within weeks to low-digit months," with Genesee in active commercial discussions. Genesee remains the single biggest optionality in the story: 26,000 acres, water, fiber, high-voltage interconnection, and a Phase 2 head start relative to hundreds of U.S. moratoriums.

What the tape is saying back

Here is the tension. Capital Power's entire equity narrative rests on speed to power for hyperscalers and utilities. The market's power-for-AI complex is actively de-rating: global theme AI data center sits among the heaviest 90-day and 30-day decliners in the tape, with demand for power also negative across windows. Bitcoin miners, data-center landlords and semiconductor suppliers tied to that buildout are down sharply. Power producers are being asked to prove contracted offtake, not data-center adjacency — which is exactly why the ~75% contracted base and the $1.25 billion de-risked number matter more than the Genesee headline. And on the stock itself, Benjamin Pham asked the question management would rather not have fielded: with shares down appreciably quarter over quarter, why not buy back stock? “Obviously, it's an option, but it's not one that I think we're going to execute unless we see considerable further weakness. We're in an enviable situation in that we have very little in terms of major capital commitments as it relates to organic growth.” — Kevin MacIntosh, Chief Financial Officer · 2026-11-02

Together, that is a business with the stability of the utility and the growth of an IPP.
Avik Dey, Chief Executive Officer · 2026-11-02

That is the pitch, and it is a coherent one. But note the ordering of the quarter's facts: record EBITDA, lower AFFO, higher sustaining capital, a minimum-sized dividend raise, no buyback, no M&A, and a stronger-for-longer message built on cash flows that management says will mostly land before 2030 but will not be pinned to a quarter. The stability is real and the contracting machine is real. The question investors will keep asking is whether the growth half of that sentence shows up as AFFO per share — or just as adjusted EBITDA with a maintenance bill attached.