Schneider Electric has agreed to buy PTC for $22.6bn in cash, which is the largest cheque the French group has ever written and the clearest statement yet of what the industrial giants think they are short of. PTC’s shareholders took the news as a windfall and sent the stock up around 35 per cent. Schneider’s own shareholders took it as a bill and marked their holding down. Both reactions are rational, and the gap between them is the whole story.
What Schneider has bought is the drawer where the drawings live. PTC’s business sits at the design end of the industrial stack, the computer-aided design and product lifecycle software in which a machine exists as a model long before it exists as metal, and in which every revision, part number and tolerance is recorded for the decades the physical thing will be in service. Schneider’s business sits at the other end, in the switchgear and the control systems and the electrical architecture of the factory that the drawings eventually become. Owning both ends means owning the full digital description of an industrial asset from the first sketch to the last firmware patch, and that description is the only substrate on which industrial artificial intelligence can actually be trained. Models cannot learn a factory from sensor telemetry alone. They need to know what the thing was supposed to be.
This is the move of an operator who has worked out where the barrier sits. Industrial AI has spent three years being demonstrated and barely sold, because the data that would make it valuable is locked in proprietary formats inside design tools that the automation vendors do not own. Buying the format is faster than persuading it to open.
Anybody who has bought a coffee business understands the logic of paying too much for the upstream. There comes a point in any serious roastery’s growth where the founder stops negotiating with green bean brokers and starts buying the relationship with the farm, and the price paid looks mad against the volume of beans involved. It is not a bean purchase. It is the purchase of knowing, in advance and exclusively, what is going to be available, at what quality, before anybody else gets to bid. Every roaster who made that move was briefly accused of overpaying for dirt, and every one of them stopped worrying about supply.
The share price split tells you which part of that argument the market is currently willing to fund. PTC holders are being handed cash at a premium today and have no exposure to whether the integration works. Schneider holders are being asked to carry an enormous outlay now against a thesis that pays out over a decade, in a market where the phrase industrial AI still appears more often in keynotes than in capital expenditure budgets. A fall in the acquirer’s stock on announcement day is not a verdict on the strategy, it is the arithmetic of a large cash deal meeting shareholders who did not choose it.
The thing worth watching is not the integration slide deck but the standards. If Schneider keeps PTC’s tools as open to rival automation hardware as they are today, it has bought a toll booth and will collect from everybody. If it starts tilting the design environment toward its own equipment, it has bought a moat and will have to defend it against every customer who dislikes being funnelled. The first path is worth more than $22.6bn. The second is worth a lawsuit in Brussels.
Twenty-two billion dollars for a filing cabinet full of machine drawings is either the best-timed purchase of the decade or a very expensive way to find out that the drawings were never the hard part.
Source: Silicon Republic, “Schneider Electric to Acquire PTC for $22.6bn”
