The group announced on August 4 that it has acquired the compressed air business of 3C Industrial LLC, headquartered in Corpus Christi with branches in Houston, Fort Worth, San Antonio and Austin.

The 26-employee operation generated roughly SEK 108 million (USD 11 million) in revenue during 2025 and joins Compressor Technique’s service division. Terms weren’t disclosed.

It’s a small deal, but a familiar one: Compressor Technique alone completed nine similar bolt-on acquisitions in the first half of 2026, steadily widening its U.S. service footprint rather than chasing scale.

And yet the single number that moved this stock to an all-time high last month has almost nothing to do with any of that. It’s chip fabs.

Vacuum Technique — the division that makes the pumps semiconductor plants use to create the near-perfect vacuums their manufacturing processes require — posted order growth of 60 percent in the second quarter.

That’s not a typo, and it’s not just a weak comparison base either, though the base helps.

It’s the clearest sign yet that the AI infrastructure buildout has started showing up in the order books of companies most investors would never think to associate with it.

The stock has responded accordingly, and the question worth asking is whether it’s responded too much.

A 150-Year-Old Compressor Maker With a New Growth Engine

Atlas Copco is a Swedish industrial group with a market cap around SEK 995 billion, listed on Nasdaq Stockholm.

It runs four divisions:

• Compressor Technique, the mature cash-generating core (roughly 46 percent of revenue)
• Vacuum Technique, which serves semiconductor and scientific customers (about 22 percent)
• Industrial Technique, precision tools for automakers and electronics makers (around 16 percent)
• Power Technique, portable compressors and generators for construction (about 17 percent)

It’s Vacuum Technique carrying the current story, and carrying most of the stock’s recent move.

How the Installed Base Actually Makes Money

The business model that ties the four divisions together is the installed base.

Atlas Copco doesn’t just sell a compressor and walk away — it sells decades of service contracts, spare parts and upgrades on equipment that’s expensive to switch away from once it’s running a customer’s production line.

Service now makes up roughly 45 percent of Compressor Technique’s revenue and closer to 30 percent in Vacuum Technique, and that recurring tail is a big part of why the group can post 20-percent-plus operating margins even in a down cycle.

Vacuum Technique runs on a similar logic but a different customer base: semiconductor fabs, scientific instrument makers, food packaging lines.

When a fab commits to a vacuum pump supplier, it’s committing to years of service revenue alongside the initial sale.

That’s the part of the business currently in a boom, and it’s why a company most people picture as a maker of steel boxes has quietly become one of the more resilient industrial businesses in Europe.

The Quarter Where Vacuum Technique Flipped Overnight

Second-quarter orders came in at SEK 50.95 billion, up 27 percent year on year and 26 percent organically — a record for the group.

Revenue grew 9 percent to SEK 44.97 billion, with adjusted operating margin climbing to 21.0 percent from 20.4 percent a year earlier.

But look at where that came from.

Vacuum Technique’s operating margin jumped to 21.0 percent from 18.9 percent, on order growth of 59 percent organically, driven by what management called record demand from semiconductor and flat-panel customers.

Compare that to Power Technique, where margin merely held roughly flat at 17.5 percent versus 17.1 percent a year ago, after five straight quarters of decline through the first quarter of this year — a stabilization, not yet a recovery.

Industrial Technique’s margin did jump to 20.0 percent from 17.1 percent, which is genuinely encouraging, but off a smaller revenue base and a weak prior-year comparison.

The quarterly trend inside Vacuum Technique tells the story better than any single quarter can.

The division’s revenue fell year over year in every quarter of 2025 — down 11 percent in Q2, 12 percent in Q3, 11 percent in Q4 — before flipping to plus 17 percent in the second quarter of this year.

That’s not gradual improvement. That’s a segment going from contraction to acceleration in two quarters, and it’s the number I watch most closely, because it’s the cleanest read on whether the semiconductor capex cycle is actually turning or just had one strong quarter.

A Record High Built on One Segment’s Reversal

The stock closed at roughly SEK 178 the day before the July 16 report and has since climbed to just under SEK 204, clearing the previous all-time high set in June 2024 around SEK 206.

That’s a move of nearly 12 percent in ten trading days, on top of a stock that was already up over 30 percent from its 2025 lows before the report even landed.

Multiple sell-side desks raised price targets in the days after the print — Citigroup to SEK 245, SEB to SEK 230, Morgan Stanley to SEK 215 — while Deutsche Bank, more cautious, moved its target to just SEK 181 and kept a hold rating.

That roughly SEK 64 spread between the most bullish and most cautious targets tells you the sell side itself hasn’t agreed on how much of this quarter is durable.

When a stock breaks a two-year-old record high on the back of one segment’s order book, the technical picture isn’t really separate from the fundamental question — it’s the same question, expressed in price.

The Market Has Decided This Is Structural, Not Cyclical

The prevailing read on Atlas Copco right now is straightforward: the semiconductor investment cycle is real and multi-year, Vacuum Technique is the purest way to play it inside a diversified industrial, and the rest of the group provides ballast rather than drag.

I don’t think that’s wrong, exactly. But I think it’s doing more work than the data has earned it yet.

Semiconductor capex has a long history of moving in sharp, unpredictable bursts rather than smooth, structural curves — that’s precisely why Vacuum Technique’s own margin swung nearly nine percentage points over the last two years.

The market is pricing this quarter as the start of a new, durable growth trajectory. It could just as easily be the sharpest quarter of a cycle that mean-reverts the way every semiconductor cycle before it has.

The Concentration Risk No Press Release Will Mention

Three risks are worth sitting with here, and none of them are the ones a press release would flag.

First, concentration. Nearly all of the recent acceleration is coming from one division exposed to one end-market that’s historically been Atlas Copco’s most volatile.

If semiconductor capex intentions cool even modestly heading into 2027 planning season, the group-level growth story loses its main engine overnight, and the stock has priced in continuation.

Second, the currency drag hasn’t gone away. A strong krona shaved 3 percent off Q2 revenue and 8 percent off first-half order growth.

If SEK keeps strengthening, it will keep eating into reported numbers even if underlying demand holds — a headwind that’s easy to overlook when the order growth headline is this loud.

Third, and most overlooked: return on capital employed has fallen for three straight years, from 30 percent in 2023 to 24 percent now, as the capital base grew faster than earnings during the softer 2024–2025 period.

A stock trading at close to 37 times trailing earnings needs that trend to reverse, not just stabilize, to justify the multiple on a forward-looking basis.

Great Business, Fully Priced Quarter

Atlas Copco is a genuinely strong business having a genuinely strong quarter, and I don’t think the semiconductor demand behind it is fake.

What I don’t buy is that the price today still leaves much room for anything other than continued good news. For the thesis to keep working from here, I’d want to see:

• Vacuum Technique orders holding double digits organically into the October report, rather than this quarter proving to be the peak
• Power Technique’s margin actually inflecting upward, not just stabilizing
• Return on capital employed arresting its three-year decline

This is a stock for investors who want quality industrial exposure to the AI infrastructure buildout without owning a pure-play chip supplier, and who are comfortable paying up for that diversification.

It’s the wrong entry point for anyone underwriting a re-rating from here — with the multiple already reflecting a best-case read on one segment, there’s more room for disappointment than for surprise.