Siemens Spent a Decade Spinning Off Its Crown Jewels. It Kept the Keys to Every One.

In January 2013, Siemens shareholders voted to give away a piece of one of the most famous names in light bulbs. For every ten Siemens shares they held, they would receive one share of OSRAM — the lighting business Siemens had nurtured for the better part of a century, now set loose to trade on its own.¹ It looked like a clean break. It wasn't. After the spin-off, Siemens quietly kept 19.5% of the company it had just let go.¹ That small retained stake is the tell. It would repeat, with bigger numbers and bigger businesses, for the next twelve years.

The official story is that Siemens spent a decade shedding its sprawling conglomerate skin to become a sharp, focused industrial-software company. The real story is more interesting: Siemens kept giving things away while holding on to a controlling or anchor stake in nearly every one — and is still doing it today. The focus is real. The letting-go is theater.

Three crown jewels, none fully released

Walk the timeline and a pattern hardens into a habit. OSRAM left in 2013, with Siemens holding 19.5%.¹ Then came the bigger one. In March 2018, Siemens floated Siemens Healthineers — its medical-imaging crown jewel, a €13.8 billion revenue business running an 18% margin — in what became the second-largest German IPO since 2001, raising €4.2 billion.²³ But this was no clean spin-off. Siemens sold only existing shares to a 'meaningful minority' of new investors and walked away still owning roughly 85% of the company, which stayed fully consolidated on its balance sheet.²³ Healthineers was 'public' the way a tenant is independent while the landlord still owns the building.

Energy followed in 2020 — and again, the structure tells you who was really in charge. Siemens Energy was not an IPO at all; it was a spin-off under the German Transformation Act. Existing Siemens shareholders received one Energy share for every two Siemens shares they held, the gas, power, and Gamesa wind operations were legally carved out, and 55% of the new company went to the market.⁴⁵ Siemens kept a 35.1% anchor stake.⁴ No new capital was raised at listing; this was a redistribution, not a sale. Three businesses, three different legal mechanisms, one consistent reflex: never quite hand over the keys.

Why prune at all — and why buy software with the proceeds

The logic of the pruning is sound, and the numbers prove it worked. A conglomerate is valued at the average of its parts; the market applies a discount to a basket it can't cleanly model. Lighting, medical scanners, gas turbines, and factory automation share almost nothing — not customers, not cycles, not the engineering that makes one excellent. Holding them together meant every business was dragged toward the median. By fiscal 2024, the trimmed-down core was posting comparable revenue of €75.9 billion, record net income of €9.0 billion, an Industrial Business margin of 15.5%, and €9.5 billion of free cash flow.⁸ Discipline shows up in the margin.

But pruning is only half the play. The other half is where the proceeds go — and Siemens spent them buying its way deeper into a single adjacency: industrial software and AI. The capstone was Altair Engineering, agreed in October 2024 at $113 per share, a 19% premium, for an enterprise value of approximately $10 billion.⁶ The deal closed on March 26, 2025; Altair left NASDAQ and became a wholly owned subsidiary.⁷ This is the textbook adjacency move: don't diversify into anything that grows — concentrate into the layer that makes your existing business worth more. Simulation software sits on top of the same factories and machines Siemens already automates. The discount-magnet sprawl went out one door; depth in one adjacency came in the other.

The anchor stake the focus story can't explain

Here is where the clean narrative cracks. If the goal was focus, why keep 19.5% of OSRAM, 85% of Healthineers, and 35.1% of Siemens Energy?¹²⁴ A truly focused company sells outright and walks away. Siemens didn't — it kept anchor stakes for years, continued consolidating Healthineers on its balance sheet, and retained enough of each to remain entangled in businesses it claimed to be exiting. And the habit hasn't broken. In 2025 Siemens announced it would finally deconsolidate its remaining ~67% Healthineers stake — but again not by selling, rather by transferring just 30% to its own shareholders via a direct spin-off.⁹ Seven years after the IPO, Siemens is still working out how to let go of the company it 'sold' in 2018.

Isn't the slow exit just prudent capital management?

The fair objection is that retaining stakes isn't conglomerate nostalgia — it's discipline. Dumping 85% of a business onto the market in one day crushes the price you receive; phasing the exit protects shareholder value, and the records show it. Each step was approved overwhelmingly — the Energy spin-off carried 99.36% of votes.⁴ And the results vindicate the patience: record fiscal 2024 net income, a strengthening margin, and a €2.0 billion gain on the later Innomotics sale.⁸ By that read, the anchor stakes were a controlled descent, not a failure to jump.

It's a strong defense, and partly true — but it doesn't explain the duration. A controlled exit takes quarters, not a dozen years. The honest counter is that prudence and reluctance produce identical short-term behavior; you can only tell them apart by how the story ends. Twelve years after OSRAM and seven after Healthineers, Siemens is still mid-exit on a business it once described as having a 'meaningful minority' float.³⁹ At some point a controlled descent that never lands is just a company that doesn't want to let go. The focus is real where it shows up in the margin. The discipline is still a work in progress where it shows up in the cap table.

Siemens did the hard thing most conglomerates only talk about: it let go of its most famous businesses, one by one, and poured the proceeds into the one layer where it could compound — buying $10 billion of industrial software while the basket-discount went out the door. The margin proves the strategy worked. But the cap table tells the rest of the truth. A company that keeps 19.5%, then 85%, then 35.1%, and is still negotiating its goodbyes a decade later isn't a conglomerate that became a focused technology company. It's a conglomerate learning, very slowly and at great expense, what focus actually costs — and discovering that the hardest portfolio decision was never which business to keep. It was finding the nerve to fully let one go.