Concentrated capital can mark an industrial-software company down seventeen percent in nine months and still write a twenty-two-billion-dollar all-cash check for it — and that, not any machine and not any line of code, is what now decides who will own the software American factories run on. The Wall Street Journal reports that Schneider Electric agreed to buy Boston-based PTC for $22.6 billion, offering $205 a share in cash, a 42% premium to its last close, with PTC’s board recommending the deal and closing expected in the third quarter of 2027. The stated purpose is portfolio-building: one of the largest industrial-software collections assembled at a moment when artificial intelligence is reshaping what customers want.

The honest case deserves stating before anything else. Hardware makers do need software if they intend to sell machines that design, monitor, and repair themselves; industrial customers genuinely do want one thread running from product design to the shop floor; and Olivier Blum, Schneider’s chief executive, is not wrong that the physical and the digital increasingly meet inside the same product. The premium is real — 42% over Friday’s close is a genuine sum to the people who held PTC, and Neil Barua, PTC’s chief executive, is telling his shareholders the truth when he calls it certain and compelling value. Schneider’s own shares have nearly doubled over three years on the data-center build-out. If you believe the next industrial age will be decided by whoever controls product design and machine data, moving to the front of that line is not madness. It is exactly what a well-advised conglomerate would do.

But the sequence is the part that will not leave me. PTC had been accelerating its own share repurchases because it considered its stock “compressed” — down 17% since the first of the year. Read that plainly: the people closest to the asset looked at the surplus their work produced and could not find a productive home for it at a price the market would respect, so they handed the money back to shareholders. Months later a French conglomerate paid roughly eight and a quarter times annual revenue for the same asset. That is the anatomy of the thing. The surplus travels to whoever holds the claim, not to whoever writes the software. The engineers and salespeople in Boston who make the product go to work every morning under one set of rules, and a twenty-two-billion-dollar decision about the next ten years of their labor is made under another — by owners in Paris and Boston who will never set foot in the building. I used to trade the corn before it was planted, and I know the difference between a claim on future value and the work that creates it. What decides ownership here is the claim.

And it is the workers on both sides of the Atlantic who absorb the risk of that decision. When relevance is bought with borrowed urgency, the first things mortgaged are the things that cannot be sold to a shareholder: the training, the slow institutional knowledge, the plant town’s expectation that the firm will still be there in a decade. Schneider’s shares fell more than 8% in European trading while PTC’s rose 35% in the premarket. The sellers were made whole in a single morning. The buyers now wait until the third quarter of 2027 to find out whether they were, and the people who work at either company never got a vote in the matter at all.

The language is doing work the arithmetic will not support. Schneider describes the result as “a software and AI powerhouse with a portfolio bridging the physical and digital worlds.” What the company’s own figures describe is nearly a quarter of revenue coming from software and services, up from less than a fifth — a one-fifth-to-one-quarter mix shift, not a transformation. In the same breath, Schneider says it will finish this year’s €600 million buyback, pause stock repurchases entirely in 2027 and 2028, and reaffirm a program of €2.5 to €3.5 billion through 2030. The same outfit writing $22.6 billion in cash for a Boston software company is telling its own owners it cannot buy back their stock for two years. The announced synergies — around €800 million in revenue and €250 million in cost savings by the third year — are the ordinary promises that attend every merger of this size, and they are being asked to carry a check written at more than eight times revenue against an asset the market had just finished discounting.

The German comparison is quieter and more instructive. Siemens has assembled its software portfolio through multibillion-dollar acquisitions of Altair Engineering and Dotmatics without a 40% premium, without an 8% drop in its own shares, and without announcing a bridge between the physical and the digital. Siemens bought software the way an industrial operator buys software. Schneider bought it the way a two-hundred-year-old conglomerate buys relevance. That a French industrial house feels compelled to follow a German one into an American software shopping spree tells you the deal is the price of admission rather than an edge — the cost of staying in a game whose rules someone else wrote.

So here is the question the coverage will not ask, and the only one that matters: what would it look like to own industrial software for the commonwealth instead of for the multiple? There is an answer, and it is not a footnote. It is a stewardship operator — permanent capital with no exit clock, chartered to hold an asset for decades rather than to dress it for sale, run by people whose compensation rises with the durability of the enterprise rather than with the premium they extract at the door. It is an employee-ownership trust standing between the workforce and any future buyer, so that the engineers in Boston and the plant crews in Grenoble share in the surplus their own labor produces, and so that no board can sell the future of a company they built without asking them. It is a firm answerable to the town it sits in — the subsidiary thing kept small and local and governed by the people who live with the consequences, which is what subsidiarity has always meant. This is distributism in modern dress: property spread widely enough that no single block of interests can dictate the terms of other people’s work, the earth’s goods understood as given for all rather than as a portfolio to be optimized. We know it works because it has worked — Mondragon’s worker federation, Land O’Lakes, the credit union, the mutual insurer, the electric cooperative on the highway north of Friendship where the members elect the board and the dividend comes back to the people who pay the bill. I manage one of those. It is not a theory.

Note what that answer refuses as well. The concentrated state is the same disease in a different coat; a bureau in Washington deciding which industrial software Americans may own would only move the unaccountable power from Paris to the Potomac and leave the engineers out of the room anyway. The cure for concentrated capital is distributed ownership, not concentrated administration.

Schneider’s deal closes, if it closes, in the third quarter of 2027. Two boards have already agreed; what remains is procedure. What has already happened is smaller and harder to undo: an American industrial asset, marked down for nine months by the only people permitted to price it, was sold in an afternoon to the one bidder willing to pay above the price. The stock will be integrated, the synergies will be reported, and the surplus will flow to holders on schedule. Meanwhile the people who write the software will keep writing it, and the people who buy the machines will keep running them, and neither group will have been asked whether the commonwealth they serve is better off being owned by a conglomerate that just told its shareholders it cannot afford to buy back its own stock. That is what has been lost. It was never ours to lose, which is exactly the grievance.