IBM: A Deep Dive Through Buffett's Lens

Not every Buffett position is a winner, and IBM is the cleanest example of the Oracle of Omaha misjudging a business. In 2011, after reading fifty years of IBM's annual reports, Buffett had what he called an "aha" moment about the company's ability to retain and serve large enterprise customers. That conviction led Berkshire Hathaway to build a stake of 63.9 million shares — about 5.4% of the company — by the end of 2011, a position of roughly $10.8 billion that made Berkshire IBM's largest shareholder. At the time it was the biggest single stock investment Buffett had ever made. He even joked in his letter that if IBM's buybacks shrank the float enough, he might end up owning the whole company. For a man who had long avoided technology stocks, it was a deliberate expansion of his circle of competence.

The Moat — and the Cloud That Undermined It

IBM's apparent economic moat was its deep entanglement with corporate IT departments: mainframes, software, and services contracts that were painful for clients to rip out. Those switching-costs looked durable, and Buffett trusted them. The problem was that the ground shifted beneath them. As Amazon Web Services and Microsoft Azure redefined enterprise computing, IBM's traditional hardware and services businesses began a long, slow decline. The "cloud" was not just a new product; it was a new architecture that made parts of IBM's franchise obsolete.

The warning signs were visible early. IBM had promised $20 of earnings per share by 2015; it abandoned that target at the end of 2014 and reported just $14.92 in 2015. Revenue fell for years. Buffett had bet on a management team's ability to reinvent a giant, but the reinvention lagged the disruption. By 2017 he admitted the error publicly. On CNBC he said, "My valuation of IBM has changed... IBM is a big strong company, but they have very tough competitors." At the annual meeting he was blunter: "Six years ago I started buying IBM. Six years later, IBM has not performed as expected. I made a mistake in my judgment of IBM." Note the word "judgment" — this was not a fraud or a collapse but a misread of how fast a technology moat could be dismantled.

Capital Allocation and the Quiet Exit

Buffett had been drawn in part to IBM's enormous buyback program, believing that repurchases at a reasonable price would compound Berkshire's stake. The IBM episode became a lesson in the limits of that logic: buybacks cannot rescue a business whose owner-earnings are shrinking. Berkshire sold roughly a third of the position in the first half of 2017, then cut it by about 94.5% by year-end, leaving only about 2 million shares. By 2018 Buffett confirmed Berkshire had fully exited. The capital allocation lesson was stark — a great company returning cash is only a great investment if the underlying economics are intact. IBM itself later tried to leap into the cloud era with its $34 billion acquisition of Red Hat in 2019, a move that came only after Buffett had already moved on.

The opportunity cost was the real damage. Over the holding period IBM's shares fell about 18% (roughly flat to slightly up including dividends, a mid-single-digit total return across seven years), while the S&P 500 rose more than 100%. As Buffett likes to say, not losing money is not the same as winning. Had the same capital sat in Apple — a later, more successful tech bet that Buffett framed as a consumer company rather than a tech company — the result would have been dramatically different.

Valuation and Margin of Safety

The IBM mistake is a margin of safety story in reverse. Buffett did not overpay in the way he overpaid for Kraft Heinz; rather, he mis-estimated the durability of the moat. A circle of competence is not just knowing an industry exists — it is knowing how its advantages can be overturned. Technology moats, Buffett concluded, are fundamentally different from the century-long brand loyalty of Coca-Cola-style businesses: a tech lead can be dismantled in a decade. His later decision to treat Apple as a consumer-products company, not a tech company, was a direct correction born from the IBM experience.

Risks to Watch

• Disruption risk. Enterprise computing keeps shifting to cloud-native and AI-native architectures.
• Currency exposure. IBM's global sales make earnings sensitive to the dollar.
• Competition. Hyperscalers and niche software firms attack from above and below.
• Buyback dependence. Financial engineering can mask, not fix, weak organic growth.
• Legacy liabilities. Pension and restructuring obligations from decades of operations linger.

The broader implication is about estimating durability, not just identifying a moat. Buffett could name IBM's advantages precisely; what he underweighted was the speed at which a technology lead can be dismantled by a cheaper, better architecture. A moat assessment that ignores the replacement rate of the underlying technology is incomplete — and that blind spot, more than any single missed quarter, is what turned a flagship bet into a teaching case.

What Value Investors Should Take Away

IBM shows that even Buffett can misjudge a moat — and that the honest admission and prompt exit matter more than the original error. For ValueOS readers, the takeaway is twofold: respect the difference between a consumer franchise whose advantages compounding for generations and a technology franchise whose advantages can evaporate, and remember that compounding only works when the business itself is compounding. The IBM chapter is not an indictment of tech investing; it is a master class in knowing which moat you truly understand.