IBM called it liberation. But you don't free a business by handing it your decline, your losses, and a brand-new ticker to carry them on.
Pairs with the Cannibalization Decision Tree — a ready-to-use strategy tool. Get it — included with a subscription, or $1.99 →
On October 8, 2020, IBM announced it was carving out the part of itself that, for decades, had been one of its defining identities: the business that runs other companies' computers. The unit it spun out had more than 4,600 clients in 115 countries and a $60 billion backlog.2 By any normal measure of scale, this was not a side project being shed. It was a leading global business being shown the door. The company's own language was triumphant — a 'market-leading standalone company,' a sharper focus, a growth strategy accelerated.3 What the press release did not say out loud was simpler and sharper: IBM was not setting this business free. It was getting away from it.
The official story is that IBM spun off Kyndryl to unlock value for both companies. Read the filings instead of the press release and a different story shows up. The business being unlocked was already shrinking — and it would leave home deep in the red. This was not a gift to a promising orphan. It was an amputation to save the parent's valuation.
IBM had already bought a new identity — and couldn't afford the old one: once you've paid to become a cloud company, the old outsourcing business becomes a story you can't tell anymore
To understand the spinoff, start eighteen months earlier. In July 2019, IBM closed its acquisition of Red Hat for roughly $34 billion in cash — its largest deal ever, the open-source software company bought to make IBM the company of hybrid cloud.5 That is an enormous bet to plant a flag. And once you have spent $34 billion telling the market you are a cloud-and-AI company, the most dangerous thing on your own income statement is a vast, low-growth business that tells the market the opposite. IBM's Infrastructure & Cloud Services line — the core of what became Kyndryl — was $19.7 billion in 2020, and it had fallen 5.1% in a single year.4 A company trying to earn a software multiple was carrying a declining outsourcing business roughly the size of a Fortune 200 company. The two halves told two stories, and the market prices the story.
The business being 'freed' was already shrinking: the pre-spin filings show retreat, not a coiled spring waiting to be released
Here is the part the liberation narrative skips. Kyndryl's own SEC filings, prepared as it prepared to stand alone, show a business in retreat well before the split. Revenue was $20.3 billion in 2019, down 7%, then $19.4 billion in 2020, down another 5% — and the customer count slipped from roughly 4,600 to about 4,400 in a single year.7 This is the profile of managed-infrastructure services as a category: long contracts, thin margins, and a steady migration of workloads to the public cloud that nibbles the base every year. You do not spin a business like that off because it is about to bloom. You spin it off because keeping it makes the rest of you look slower than you are.
| The official framing | The numbers underneath | |
|---|---|---|
| The unit's growth | A market-leading standalone company | Revenue down 7% in 2019, down 5% in 2020 |
| The customer base | 4,600+ clients, $60B backlog | ~4,600 customers in 2019, ~4,400 by end of 2020 |
| Kyndryl's first year out | Liberated to pursue its own strategy | $2.3B net loss in 2021 |
| First full independent year | Sharper focus | $17.0B revenue, down 7%; $1.4B net loss |
When a conglomerate splits, watch which half keeps the brand and the balance sheet, and which half gets the declining revenue line. A spinoff doesn't create value out of thin air — it reallocates it. The most common version isn't 'two great companies, now unbound.' It's 'one company shedding the part that was dragging its multiple, packaged as liberation for the part being shed.' The tell is in the pre-spin filings: if the carved-out unit was already shrinking and exits loss-making, the value being unlocked was the parent's story, not the orphan's future.
The freed company immediately lost money — and kept shrinking: the decline didn't stay behind at separation — it walked out the door with the company
If Kyndryl had been a coiled spring held back by a bureaucratic parent, you would expect it to uncoil once cut loose. It did the opposite. Its 2021 results — the year of the split — showed $18.7 billion in revenue and a net loss of $2.3 billion.6 Then, in its first full year as an independent company, the fiscal year ending March 2023, revenue came in at $17.0 billion, down 7% as reported, with another net loss of $1.4 billion.6 The shrinkage did not stop at the door of independence; it walked out with the company. That is the quiet evidence that the spinoff was structurally necessary for IBM but not, in any obvious way, value-creating for Kyndryl. One company got to look like the future. The other got to keep being the thing the first one wanted to stop being — just on a different ticker.
“Separating the managed infrastructure services business creates a market-leading standalone company and further sharpens our focus on IBM's open hybrid cloud platform and AI capabilities. This will accelerate our growth strategy and better position IBM to seize the $1 trillion hybrid cloud opportunity.”3
Wasn't this just a clean, well-run focus play?: a focus play and an escape valve can be the exact same maneuver — this one was both
The fair objection is that this is too cynical — that focus is a real and defensible strategy, and IBM executed it cleanly. There is truth in that. A managed-infrastructure business and a hybrid-cloud platform business have genuinely different cost structures, different growth curves, and different kinds of investors; running both under one roof forces compromises that serve neither. Splitting them lets each be valued on its own terms and managed by leaders who aren't fighting for the same capital. That is a legitimate case, and IBM's announced focus on Red Hat and AI was consistent with a direction it had been building toward for years. But notice what the steelman quietly concedes. 'Letting each be valued on its own terms' is exactly the point — IBM's terms were favorable and Kyndryl's were not, and everyone involved knew which was which. A focus play and an escape valve can be the same maneuver. The honest read is that this one was both, and the 'unlock value for both' language papered over the asymmetry. The parent kept the story it had paid $34 billion for. The spun-off business kept the decline.
Whenever a company announces a spinoff in the language of mutual liberation, do one thing: pull the carved-out unit's own pre-spin filing and read its revenue trend and its first independent results. The parent's narrative will always sound like focus. The orphan's numbers will tell you whether anything was actually unlocked. If the unit was already shrinking and exits loss-making, the spinoff was a balance-sheet decision dressed as a growth story — useful to know before you treat either stock as a 'freed' growth bet.
IBM did not free Kyndryl. It set down a weight. The $34 billion it spent on Red Hat had already committed it to a story — cloud, open source, AI — and you cannot tell that story convincingly with a $19.7 billion declining outsourcing business hanging off the same income statement. So the declining business got its own name, its own ticker, and its own losses to carry. The maneuver was real, and for IBM it was probably right. But strip the word 'liberation' away and what remains is colder and more useful: a company that bought a new identity, then quietly amputated the old one so the market would finally believe the purchase. The spinoff didn't create value out of thin air. It just moved the part that was dragging the multiple to where it couldn't drag it anymore.
When companies cut off a piece of themselves on purpose
Cannibalization Decision Tree
A decision tree for the moment the new thing threatens the cash cow: is the disruption real, will someone else do it if you don't, and can you afford to bleed your own margin to own the future? Blank to run on your own line; filled as the worked example tracing how the story's incumbent chose to cannibalize — or flinched and got cannibalized.
Included with any subscription, or unlock this tool for $1.99. Get it → · See plans →
Sources
Where this comes from — the filings, records, and reporting behind it.
- 1On November 3, 2021 (the Separation Date), IBM completed the separation of its managed infrastructure services business into a separate, independent publicly traded company, Kyndryl Holdings, Inc. The separation was structured as a pro rata distribution to IBM stockholders of 80.1% of the outstanding shares of Kyndryl; IBM stockholders received one share of Kyndryl for every five shares of IBM held on October 25, 2021.
- 2IBM announced the separation of its managed infrastructure services business (initially called 'NewCo') on October 8, 2020. The announcement stated that NewCo would immediately be the world's leading managed infrastructure services provider with more than 4,600 clients in 115 countries and a backlog of $60 billion.
- 3In the October 19, 2020 Q3 earnings press release, IBM CEO Arvind Krishna stated: 'Separating the managed infrastructure services business creates a market-leading standalone company and further sharpens our focus on IBM's open hybrid cloud platform and AI capabilities. This will accelerate our growth strategy and better position IBM to seize the $1 trillion hybrid cloud opportunity.'
- 4IBM's Infrastructure & Cloud Services revenue (the primary component subsequently carved into Kyndryl) was $19,669 million in full-year 2020, a decrease of 5.1% as reported versus prior year.
- 5IBM closed its acquisition of Red Hat on July 9, 2019, acquiring all outstanding common shares at $190.00 per share in cash, representing a total equity value of approximately $34 billion. The merger agreement had been signed October 28, 2018.
- 6Kyndryl's full-year 2021 revenues (the partial year including its pre-spin period) totaled $18.7 billion and its net loss was $2.3 billion. In its first full independent fiscal year ended March 31, 2023, Kyndryl reported revenues of $17.0 billion (a decline of 7% reported, flat in constant currency) and a net loss of $1.4 billion.
- 7Kyndryl's pre-spin revenue was already in decline: $20.3 billion in 2019 (down 7%) and $19.4 billion in 2020 (down 5%), with approximately 4,400 customers at end of 2020 versus 4,600 at end of 2019, per Kyndryl's own SEC Form 10 filing.
- 8Kyndryl began trading on the New York Stock Exchange under the symbol 'KD' on November 4, 2021 — the first trading day after the November 3 distribution date. The NYSE opening price reference date for tax-basis purposes was November 4, 2021, when IBM closed at $120.85 per share and Kyndryl closed at $26.38 per share.