Sandvik's Tungsten Windfall Is Closing. The Price Increase Is Meant to Stay.

A record Q3 — 22.5% margin, 87% cash conversion, EPS up 65% — rests on volume and a commodity timing arbitrage that management insists was never the point.

SAND.ST · Earnings Call · 2026-10-22

The quarter that insists it wasn't a one-off

By the numbers, Sandvik's third quarter of 2026 is a blowout: orders up 20% reported and 17% organic, revenue up 29% and 25% organic, adjusted EBITA of SEK 8.5bn at a 22.5% margin versus 19% a year ago, a record SEK 7.2bn free operating cash flow at 87% conversion, and adjusted EPS 65% higher at SEK 4.64. It is the sixth consecutive quarter of double-digit order intake growth — and it landed on top of a 16% comparison. Stefan Widing wants you to know none of it is borrowed from the future:

the earnings per share is up 65% in the quarter, which I think is a quite remarkable number, especially that we should consider that there is no one time sort of effects in the comp or in the current period of any material size.
Stefan Widing, CEO · 2026-10-22

That claim is the whole debate. Because a good chunk of the call is spent explaining a one-time effect — and then explaining why it is about to stop being one.

Tungsten: an arbitrage that is closing

Sandvik's tungsten chain is unusual: a mine supplying roughly a tenth of its powder needs, about 55% recycled virgin material displacement via buyback scrap, the rest bought in. Powder Solutions reprices monthly against APT notations; cutting tools carry a nine-month cost lag. When tungsten spiked, that lag mechanically printed profit. That is the temporary timing effect — and it came in at SEK 104m this quarter against roughly SEK 200m guided, because tungsten prices slipped at quarter-end and the inventory got revalued. The forward-looking sentence matters more than the miss. Cecilia Felton: “With this price increase now, it is set so it will offset the impact of the current tungsten price levels.” — Cecilia Felton, CFO · 2026-10-22 In other words, the company has priced itself back to neutral. Q4 gets a "limited" timing impact. So the flattering line item goes to roughly zero — which throws the question onto what the 25.7% Machining margin is actually worth without it. Daniela Costa of Goldman asked precisely that: the 22–24% corridor was set before tungsten existed as a factor, so is today's level structural or flattered? Felton deferred to the November Capital Markets Day: “So it will be a cliffhanger until then.” — Cecilia Felton, CFO · 2026-10-22 The counter-argument, and it is a good one, is that high tungsten is not only a windfall but a competitive weapon. Mid-market toolmakers carry a far higher tungsten share of cost of goods, so they must raise prices harder — compressing the premium discount. Widing made the point again this quarter: “at least publicly traded competitors do not really have a margin profile that means that they can sacrifice further margin to try to gain market share” — Stefan Widing, CEO · 2026-10-22. This is not new framing; Felton said it in April as well: “if you are a mid-market player, you are -- will likely need to increase their prices more in relation to your current price which, of course, then is a favorable dynamic for us and our premium cutting tools brand.” — Cecilia Felton, CFO · 2026-04-22 The mid market squeeze is now a two-quarter-old thesis being tested by falling spot prices. And here is the uncomfortable disclosure buried in the Q&A. Strip price out of Powder and the volumes are shrinking:

Yeah, they are down double digits on both.
Stefan Widing, CEO · 2026-10-22

Orders and sales. Down double digits on volume. Powder's growth is price, and its order book has been shut for two quarters while the company reworks the pricing model — reopening in Q4 with Widing conceding “I don't know how it will land.” — Stefan Widing, CEO · 2026-10-22 The other lever nobody would size: Scrap prices. Sandvik's edge is buying scrap below virgin, with limited recycling capacity outside China. Asked whether 55% scrap could go to 75–80%, Widing hedged — “there is a limit where you need an element of virgin material to get the exact recipe and mix right” — Stefan Widing, CEO · 2026-10-22 — and promised a number later. That is a real margin option left unquantified.

What the mining cycle is actually saying

Mining is the clean part of the story. Aftermarket orders up 16%, with parts and services in the high teens; equipment up 23% excluding large orders; large orders a modest SEK 700m against a 75% comparison. Rock Processing dropped 8% organically, but on lumpiness, not demand. Crucially, this is a sector cycle rather than a Sandvik invention. Metso reported Aftermarket sales up 11% with a base order increase and a "positive cycle"; Newmont is pushing its Block Cave Project toward commercial production; Southern Copper is talking Cuajone expansion and transmission capacity. The installed-base annuity is being paid across the industry, which is good for credibility and less good for differentiation. It also repeats a pattern. Management has guided mining aftermarket to high single digits for four straight quarters while printing double digits — in July: “I think we stick to that we think long term, this is a high single-digit growth business although we had a very good quarter.” — Stefan Widing, CEO · 2026-04-22 Same again: 16%. Under-promising is a fine habit; eventually the market stops paying for it as a surprise.

The market isn't watching this tape

Worth noting what the rest of the market is arguing about this quarter. The editor-curated global themes are dominated by IRB model disclosures, Fed rate hike arithmetic, mortgage universe duration and fee earner economics. Sandvik's capex-cycle story has no macro escort — mining capex does not appear anywhere near the market's top themes. That cuts both ways: less crowding, and less oxygen. One genuine shared quirk: tariff refunds. Sandvik books a tariff refund in Machining in Q4 — already received, now handed back to customers, hitting orders and revenue but not absolute EBITA. Hasbro reported the same animal this week, a tariff refund recognised in Q3. A small, real, cross-sector artefact of the trade war quietly flattering and then distorting industrial revenue lines.

The five-week cliffhanger

Sandvik enters Q4 with net debt to EBITDA at 0.9, a closed Diemme Filtration acquisition making 24 divisions, and a margin running above corridor for two consecutive quarters. The bear case is not that tungsten collapses — it is that price increases calibrated merely to "protect margin" at spot get competed away, while Powder volumes are down double digits and the order book has been dark for two quarters. The bull case is that 17% organic growth on a 16% comparison, with a structural scrap-feed advantage and a premium mix shift, is a share-gainer the corridor has not caught up with. The Capital Markets Day on 24–25 November is where that gets settled. Until then, the market is being asked to take a 22.5% margin on faith — from a management team that just spent an hour explaining why the last two points of it were temporary, and why the next two will not be.