Open in interactive viewer → charts, metric popovers & call review

FLSmidth's Elephants Have Landed — Just Not in This Decade's P&L

Products orders tripled, cash flow swung positive, and a three-quarter sell-side argument about backlog conversion finally got an answer. The catch is buried in the lead times.
FLS.CO · Earnings Call · 2026-11-11

The Quarter the Math Finally Worked

For three straight calls, FLSmidth's analysts have been doing arithmetic at management and not liking the answer. In August, Lars Topholm of DNB laid out the backlog math line by line and concluded: “I simply don't understand what I'm missing in this picture. I can't make the numbers stack up.” — Lars Topholm, Analyst · 2026-08-19 Roland Andersen's reply that day was a confession dressed as a compliment: “we've been caught surprised a little bit by the high backlog conversion, you can say, which is basically positive but it means that we tie up work in progress for a while longer.” — Roland Andersen, CFO · 2026-08-19 Q3 was the quarter that argument got settled — in management's favour. Revenue of DKK 3.8bn rose 10% reported and 6% organically, with gross margin at 35% and adjusted EBITA at 16.3%, up a full point from 15.3%. Operationally the swing is unambiguous: cash flow from operations came in at DKK 358m positive after two slow quarters, free cash flow DKK 249m, leverage a sleepy 0.7x, and net profit DKK 388m. The revenue conversion that management had promised — and that the sell side had stopped believing — showed up. The mix tells you why. Over 80% of the quarter's revenue came from Service and Pumps, Cyclones & Valves, the two business lines that actually convert. PC&V grew 17% with both equipment and aftermarket rising, and management nudged its full-year organic guidance up to 9%-11% from 5%-8%. Group organic growth guidance was narrowed to 1%-3% (from 0%-4%) and the EBITA margin guided to 16.5%, the top of the old range.

Elephants, Not a New Normal

The number that stopped people mid-sentence was orders. Group order intake of DKK 4.5bn rose 36% nominally; the order intake jump in Products was better than 200% year over year, lifting the full-year Products order book into positive territory after three quarters of drought. Three large orders were booked and separately announced. The base-order run rate — the DKK 400m-700m per quarter management has defended for a year — is unchanged. That distinction matters, and management was unusually disciplined about not over-selling it. “Q3 was clearly a good quarter, a strong quarter and that will be lumpy from here. There will be more of those orders, but it won't be a new normal.” — Roland Andersen, CFO · 2026-11-11 Toni Laaksonen described these as similar orders that had been sitting in the funnel for a long time, waiting on a final investment decision that nobody could date. That is the honest read: this is a pipeline maturing, not a demand shock. What genuinely changed is the composition of the pipeline. “the pipeline is a bit stronger than in the past and there are more projects closer to the FID phase.” — Toni Laaksonen, CEO · 2026-11-11 Behind it sit copper and gold at levels that let miners write cheques, and governments that have turned friendly. Laaksonen's answer on permitting is the most informative sentence of the call: “the permitting process takes still long. So it hasn't changed, but what has changed is that many governments are positive towards the mining investments.” — Toni Laaksonen, CEO · 2026-11-11 Permits have not got faster; political will has got better. Note the company's own keyword ledger now carries "permitted" as a fresh riser for the first time in the series — a tell that FIDs, not commodity prices alone, are the live variable.

technically we have received an order in '26 where the last piece of revenue comes in '30. So it can be everything from 12 or 18 months up to 36 plus.

Roland Andersen, CFO · 2026-11-11

The Deferred Decade

That quote is the whole catch, and it is why the Products order bonanza did not move the revenue guide. Products organic revenue is still guided down 9%-13% for the year. The backlog is real; the P&L recognition is not. Bigger tickets mean longer lead times, and the ones booked in Q3 land in 2027 through 2030. So even if the mining capex cycle that management has forecast for four quarters is now genuinely arriving, it will not rescue 2027 revenue. Investors buying the order headline are buying an out-year option, not next year's earnings. The second, quieter change is balance sheet discipline. The top keyword in FLSmidth's own Q3 lexicon is milestone payment — a term that barely registered a quarter earlier. Andersen was blunt that older projects were signed with payment structures that did not pay fast enough, and that this is being fixed: "It's 1 or 2 years ago they were booked and the structure has changed as we speak." Net working capital is up 45%-50% from the clean end-2025 base, and he explicitly refused to promise normalisation: “For '27, we are still expanding inventories. We will still be delivering the projects under execution. So it's a little bit of a transition year. We'll talk a bit more about that on Tuesday. So don't expect it to come all the way back in 2027.” — Roland Andersen, CFO · 2026-11-11 That is a lower-quality earnings mix than the headline margin implies, and a reason to discount the 16.3%. Competitive behaviour also shifted. On the Jefferies question about OE pricing versus aftermarket capture, Laaksonen described bundling a full five-year lifecycle service agreement into an Indian equipment order at ramp-up — selling the flow sheet and the service contract as one package, something he claims "not too many peers" can match. If that approach scales, it converts today's lumpy order into tomorrow's annuity, which is the only structural answer to the deferred-revenue problem.

What Tuesday Has to Answer

Nearly everything of strategic weight was deferred to the Capital Markets Day on 17 November: midterm growth targets for Service (pushed back twice on this call), the M&A agenda, and a fourth business line that becomes effective 1 January but whose contents remain undisclosed. Toni had previously committed to "1 or 2" bolt-on deals in 2026 and on Wednesday could only say “there are good cases in the funnel that hopefully we can announce something in the coming months.” — Toni Laaksonen, CEO · 2026-11-11 Twelve months of pipeline-building without a signed deal is a soft data point. The Products segment, meanwhile, has gone from structural loss-maker to roughly breakeven year-to-date — a milestone Andersen had previously pinned to "Q4 next year" at a breakeven revenue level of DKK 2.8bn-3bn. Getting there a year early, on volume, is real progress. Set that against the wider tape and the contrast is stark. The market's 2026 leadership has been an AI-infrastructure complex that is now visibly cracking — the global keyword cluster around AI data centres and high-bandwidth memory sits deep in 30-day drawdown. FLSmidth is the opposite trade: a physically constrained, permitting-gated, multi-year copper and gold capex cycle whose commodity prices tailwind is real but whose revenue arrives slowly. The orders are a genuine inflection. The earnings are not. Management, to its credit, said so out loud rather than dressing a lumpy quarter as a trend.