A record €9.9 billion backlog, 94% ROCE and a 17.5% Energy margin — plus the odd strategic problem of having nothing left to sell until 2030.
WRT1V.HE · Earnings Call · 2026-10-27
The best quarter, and the strangest answer
Wärtsilä's Q3 was close to flawless on the numbers. Order intake rose 33% to €2.4 billion — 41% organically — with energy orders up 121% and equipment up 53%. Net sales grew 7% to €1.5 billion, organic growth 16%. The comparable operating result climbed 19% to €212 million, a 14.2% margin, and cash flow from operations of €583 million was, per the CFO, an all-time high for a third quarter. Return on capital employed hit 94.1% and solvency improved to 39.4%. Energy's comparable margin stepped from 15.9% to 17.5%, helped by what the CEO calls solid project execution — delivering plants below budgeted cost. The order book reached a record €9.9 billion, book-to-bill 1.59.
And yet the most interesting moment was an admission of constraint. Asked whether the €4.8 billion of trailing-twelve-month Energy orders could repeat, Håkan Agnevall did not hedge:
If you are a new customer that comes to Wärtsilä today and asking for when we can deliver, we are into the 2030s. There is clearly a balance of how far we can stretch out the time and also the capacity that we have available.
Vlad Sergievskiy of Barclays pushed: so you may not book another 5 GW in the next twelve months? "Yeah, I think that is clearly so," came the reply. This is the rare industrial call where demand is not the question — fulfilment is. It also explains why a company with 121% Energy order growth guided the next twelve months' demand environment as merely "similar." That guidance isn't caution about the market. It's a ceiling built out of engines.
The lead-time paradox
Lead time is now the company's number-six keyword, and it has inverted meaning. For two years it was the reason customers came: as Agnevall said in July, "if I take the U.S., it's so focused on lead time. I mean you can basically sell any technology if you have a short lead time." “I mean you can basically sell any technology if you have a short lead time.” — Håkan Agnevall, Chief Executive Officer (CEO) · 2026-07-21 Now Wärtsilä is the one with the long lead time, and analysts are asking whether that brands it uncompetitive. Agnevall's rebuttal is a moat argument — fuel efficiency versus high-speed engines, industrial gas turbines and aeroderivatives, no water consumption, no derating at heat or altitude, plus the strongest mid-speed service network. “The current start is a proof point that we are very competitive because people are willing to wait.” — Håkan Agnevall, Chief Executive Officer (CEO) · 2026-10-27 Is it a moat, or is it a queue that competitors will walk around when capacity arrives? Agnevall himself flagged HiMSEN's expansion and conceded "there is more capacity coming in, clearly." That is the single biggest risk embedded in a €9.9 billion backlog.
The capacity answer is stepwise and slow: first step early 2028, second early 2029, at Vaasa, with the supply chain as the binding constraint — “Critical is the supply chain here.” — Arjen Berends, Chief Financial Officer (CFO) · 2026-10-27 In July management framed this as 75% utilisation of 2025 technical capacity rising to 135% of that base by early 2028. Nine months later they are already into the 2030s.
Meanwhile the analyst obsession that will not die — euro per kilowatt — resurfaced again. Management has spent quarters trying to bury it; in July, when disclosing a 500-basis-point uplift in the Energy equipment backlog margin, Agnevall was almost testy: "I know there has been a lot of focus on euro per kilowatt, but we have also been very clear that, that is not the right KPI to try to assess the margin content of the backlog." “We have also been very clear that, that is not the right KPI to try to assess the margin content of the backlog.” — Håkan Agnevall, Chief Executive Officer (CEO) · 2026-07-21 This quarter the disclosure simply stopped: asked how Q3 added to the 500 bps, he demurred. Scope, EPC percentage-of-completion and delivery timing make the KPI meaningless — but the refusal to re-anchor it leaves the market's only crude pricing lens dark.
What quietly left the stage
The Q&A was more revealing for its absences. Energy Storage — the business that dominated 2024 and 2025 calls, with its U.S. standstill, 175% battery tariffs and Chinese cell sourcing — has been folded into the Valo joint venture and booked as discontinued operations, a €0.07 hit to EPS, with a €10–20 million negative running through share of associates from Q4. A keyword that once ranked fourth for this company, Energy Storage, is now a footnote. So is portfolio business, ranked first as recently as the June quarter: eleven business units, divested, done, one sentence.
And the guidance framework itself is being retired. "We are changing our guidance principle as we go into next year," the CEO said, confirming the demand-environment outlook — a fixture of every call in the archive — will be discontinued in favour of revenue guidance. New new financial targets were published the same morning, with a midterm window of three to five years and an 18% Energy profitability ambition, all to be detailed at a Capital Markets Day next week. Also parked there: the buyback question. Max Yates of Morgan Stanley asked directly whether the absence of a repurchase meant none was coming; the CFO's answer was "I would park that question to next week." With a balance sheet this strong and net working capital still deeply negative, that non-answer is itself a signal.
The tape disagrees — for now
Here is the tension worth holding. Wärtsilä's order book is at a record precisely because of data center demand — Agnevall says "as far as I know, everything is going to hyperscalers" — and electricity demand from electrification of industry, transport and cooling that he insists is structural: "It's not only data centers that are driving this growth." The push into Engine Power Plants as grid-balancing assets, plus balancing power, is the commercial core of the next five years.
Yet the listed proxy for that theme has de-rated. The global AI data center keyword basket is up roughly 54% over 360 days but down about 5.6% over both the last 90 and 30 days, with names like IREN, HUT, CORZ and MARA among the decliners. The public market has started to charge AI-power stories for financing and execution risk even as Wärtsilä books record orders from the same customers. Wärtsilä's own equity tape was unavailable in this dataset, so the divergence is unverified on the company side — but the setup is asymmetric: if the AI-power complex is right, the backlog is fine and the margin is the question; if the equity market is early, the 2030s order book is a liability dressed as an asset.
The margin quality question is live. Max Yates asked whether 17.5% in Energy was a jumping-off point; Agnevall offered only “We are on a journey of continuously improving profitability.” — Håkan Agnevall, Chief Executive Officer (CEO) · 2026-10-27 Arjen Berends then volunteered the caveat: “A result in a quarter can vary a lot, and it depends on many different things... the project margins are always unique.” — Arjen Berends, Chief Financial Officer (CFO) · 2026-10-27 The €10–20 million Valo drag and a 0.6–0.7 percentage-point IFRS 18 restatement headwind on EBIT both land in 2027.
So: the strongest operating quarter in the company's history, a management team that has stopped talking about storage and portfolio exits because there is nothing left to say, and a market that must now decide whether a backlog sold into the 2030s is proof of franchise or the beginning of share loss. Wärtsilä is no longer a demand story. It's a capacity, pricing and credibility story — and next week's Capital Markets Day will be the first time they've had to tell it that way.