Siemens Spun Off Its Best Businesses on Purpose. The Discount Made It Worth It.

In July 2020, Siemens stood in front of its shareholders and asked permission to give away more than half of one of its largest businesses - for nothing. No buyer, no cheque, no cash in return. Just hand 55% of Siemens Energy directly to the people who already owned it, one Energy share for every two Siemens shares.⁴ The shareholders said yes by a margin that barely needs counting: 99.36% of the capital stock in the room voted for it.⁴ A company was about to subtract from itself on purpose, and almost everyone in the building agreed it was the smart thing to do.

The official story is that Siemens was streamlining - shedding a slow, troubled energy unit to focus on its future. That's true, but it buries the more interesting move. Siemens wasn't getting rid of a problem child it couldn't fix. It was running an experiment it had already run once, two years earlier, and watched succeed - and it was now applying the result with cold consistency.

The math that punishes a company for owning good businesses

Here is the thesis a Siemens board member could repeat at dinner: a conglomerate is worth less than the sum of its parts, and the only reliable way to recover the missing value is to cut the parts loose. The market does not pay a premium for breadth - it pays a discount for it. A single share of a sprawling industrial group is a bet you can't tune: you wanted exposure to medical imaging and got dragged along by gas turbines, or you wanted the energy transition and got priced as if you'd bought an industrial-automation house. Investors can't separate the engines, so they pay for the whole machine as if every engine were the weakest one. That is the conglomerate discount, and for a group reporting €86.8 billion of revenue across nearly €98 billion of assets, the discount is not a rounding error.⁵ It's real money, locked inside the structure itself.

The fix is structural, not operational. You don't earn the discount back by running the businesses better. You earn it back by letting each one be valued on its own terms. This is what Vision 2020+, announced on August 2, 2018, was built to do: it reorganized Siemens into Operating Companies and Strategic Companies, each meant to have 'significantly more entrepreneurial freedom,' with the explicit goal of dismantling the conglomerate structure.³ The language was about freedom and focus. The mechanism was about valuation.

The Healthineers dress rehearsal nobody fully spun off

Before Siemens cut Energy loose, it ran a smaller, more controlled version of the same trick - and almost nobody describes it accurately. On March 16, 2018, Siemens Healthineers debuted on the Frankfurt exchange. The popular shorthand is that Siemens 'spun off' Healthineers. It didn't. It sold roughly 15% of the shares - a secondary offering, meaning Siemens sold its own existing stock and pocketed the proceeds, around €4.2 billion. Healthineers the company received nothing from the offering. And Siemens kept about 85%.¹ This was not a divorce. It was opening a window so the market could finally see one engine running on its own - while the parent kept the keys.

Why keep 85%? Because a minority float gives you the best of both readings. The market gets a clean, separately-traded price for the medical-imaging business - a number that proves the part is worth more standing in the light than buried in the group. And Siemens still consolidates the business, still controls it, still owns the upside. The IPO wasn't a goodbye; it was a measurement device. It put a public price tag on focus. And the price tag went up.

That single sentence is the whole strategy admitting what it is. Kaeser pointed at Healthineers' rising stock and used it - openly, on the record - as the precedent for cutting Energy loose, explicitly saying the spin-off would let them 'build two focused companies, both of which will be strong players in their respective sectors.'⁶ The 2018 IPO wasn't just a financing event. It was the control group. Once it worked, the logic became mechanical: prove that focus adds value with one business, then run the same play on the next.

Cutting loose the unit that was bleeding

If the Energy separation were purely a victory lap for focus, the timing would be hard to explain. Siemens didn't spin off Energy at its peak. It spun it off into the storm. Siemens Energy was under severe pressure, and in fiscal 2020 - the year of the listing - it absorbed roughly €1.5 billion in special items: about €956 million in impairments and write-downs, €376 million in restructuring, and €195 million in one-off carve-out costs.⁷ Strip the noise out and adjusted EBITA before special items was negative €17 million - a business sitting almost exactly on the line between making money and not.⁷ This is not the profile of a unit you release because it's too strong to hold.

And here the elegance of the structure shows. Siemens didn't sell Energy and book a loss on the way out. It distributed 55% to shareholders, parked 9.9% with its pension trust, and kept 35.1% for itself.⁴ A partial spin-off lets the parent shed the volatility from its own income statement - the turbine cycles, the impairments, the energy-transition bet that nobody could yet price - while the 'focus adds value' narrative does the talking. The hard-to-fix business gets its own balance sheet and its own story. Siemens gets a cleaner one. Both can be true: this was a real strategic conviction about the conglomerate discount, and a convenient way to put distance between the parent and a unit that was, that year, losing money.

Wasn't this just dressing up a forced exit?

The honest objection is that the whole 'focus' framing is a flattering coat of paint on a company unloading a deteriorating asset at the worst possible moment - and that real conviction would have meant a clean break, not a careful, hedged retreat. There's weight to it. But the evidence cuts the other way on the central question of motive. If this were panic, Siemens would have moved fast and fully. Instead it moved slowly and partially, twice. It kept 85% of Healthineers and, as of the planned 2027 vote, was only then resolving to distribute those shares to holders - holding its controlling stake for more than eight years after the IPO.⁸ You don't hold a showcase asset for eight years if your strategy is escape. Siemens first announced plans to separate Healthineers as its own entity in November 2016,⁹ nearly two years before Vision 2020+ formalized the Energy split³ - these were different timelines with different triggers, not one panic dressed as two decisions. The conviction about the discount was genuine. The Energy timing was opportunistic on top of it. Strategy and convenience pointed the same direction, which is exactly when companies act.

That recursion is the sharpest part of the story. The logic Siemens used on itself doesn't stop at the property line. Siemens Energy - the unit cut loose to escape the conglomerate discount - is now a multi-engine business of its own, with its own sprawl, its own gap between what it earns and what the market will pay for the whole. The fix that created it can be run on it. Siemens didn't just break up a company. It demonstrated a method, and methods don't respect boundaries. The conglomerate discount isn't a problem you solve once. It's a force that keeps pulling things together, so that someone, periodically, has to choose to pull them apart again - and call the subtraction value.