The artificial intelligence boom has created plenty of winners, but one company sits at the center of it all. Every major AI player — Nvidia (NASDAQ:NVDA | NVDA Price Prediction), Advanced Micro Devices (NASDAQ:AMD), Broadcom (NASDAQ:AVGO), Amazon (NASDAQ:AMZN), Microsoft (NASDAQ:MSFT), and Alphabet (NASDAQ:GOOG) — depends on a single manufacturer to turn advanced chip designs into physical products: Taiwan Semiconductor Manufacturing (NYSE:TSM).
Today, Taiwan Semi is worth more than $2 trillion and manufactures roughly 90% of the world’s most advanced semiconductors. In many ways, it is the ultimate AI gatekeeper. As investor Gavin Baker recently noted, Nvidia could potentially sell “trillions” of dollars worth of AI chips annually if it weren’t constrained by Taiwan Semi’s manufacturing capacity. The foundry is the chokepoint that determines how many advanced AI processors can actually reach customers.
Which makes one investing decision by Warren Buffett all the more fascinating.
Buffett’s Rare $4.1 Billion U-Turn
Buffett’s Berkshire Hathaway (NYSE:BRK-A)(NYSE:BRK-B) revealed in November 2022 that it had purchased approximately $4.1 billion worth of TSM stock during the third quarter.
For most investors, that would have looked like a classic Buffett move. The foundry possessed dominant market share, generated substantial free cash flow, earned attractive margins, and operated a business protected by enormous barriers to entry.
Yet what happened next was anything but typical. Just one quarter later, Berkshire disclosed it had sold 86% of the position during the fourth quarter of 2022. Then, by May 15, 2023, Berkshire exited the remaining stake entirely.
The complete round trip — from a $4.1 billion position to zero — took roughly two quarters. For a man whose favorite holding period is famously “forever,” that is about as close as Buffett gets to day trading.
© 24/7 Wall St.
The Problem Was Never the Business
Buffett never criticized Taiwan Semiconductor’s operations. In an April 2023 interview with Nikkei Asia, Buffett described TSM as a well-managed company and indicated geopolitical concerns played a major role in his decision to sell.
A month later, during Berkshire Hathaway’s annual shareholder meeting, he expanded on that view. Buffett called Taiwan Semi “one of the best-managed companies and important companies in the world” and said investors would likely be able to say the same thing five, 10, and 20 years from now.
His concern was Taiwan. Specifically, he said, “I don’t like its location.”
That was a direct reference to escalating tensions between the U.S. and China and the possibility of a future conflict involving Taiwan.
Surprisingly, Buffett also stated there was nobody in the semiconductor industry operating in Taiwan Semiconductor’s league. In other words, he was selling despite believing the company remained the industry’s undisputed leader.
The AI Boom Proves Buffett Was Right — And Wrong
With hindsight, Buffett’s decision looks both understandable and costly.
On one hand, the geopolitical risks he identified have not disappeared. If anything, they remain one of the largest risks facing global technology supply chains.
On the other hand, TSM has become arguably the most important company in the AI economy.
Without the foundry, Nvidia’s AI accelerators don’t get built, AMD’s AI chips sit on the drawing board, and custom processors designed by cloud giants evaporate.
That’s because Taiwan Semiconductor occupies a position few companies in history have ever enjoyed. It isn’t merely participating in the AI revolution — it is enabling it.
That said, Buffett’s decision also highlights an important investing lesson: Sometimes great businesses face risks that have nothing to do with profits, margins, or market share.
Key Takeaway
In short, Warren Buffett didn’t sell Taiwan Semiconductor Manufacturing because he doubted its business. He sold because he reassessed the geopolitical risks surrounding Taiwan.
The irony is that TSM has since become one of the world’s most dominant AI companies, sitting at the center of an industry expected to generate trillions of dollars in economic value over the coming decades.
Although Buffett may have missed one of the biggest winners of the AI era, his decision also demonstrates a principle that savvy investors shouldn’t ignore: even the best company in the world can become a difficult investment if the risks surrounding it change.
For investors evaluating the foundry today, that’s still the central question. The business case has rarely been stronger. The geopolitical debate also remains as relevant as ever.
