The Oracle's Final Trades: Why Warren Buffett Dumped $4.5 Billion in AI Stocks and Bet on The New York Times

On January 1, 2026, Warren Buffett officially retired as CEO of Berkshire Hathaway, handing the reins to his handpicked successor, Greg Abel [1]. But the 95-year-old's final moves as the head of a $1.1 trillion conglomerate — revealed in February's 13F filing with the Securities and Exchange Commission — tell a story that extends far beyond a single quarter's trades. They reveal a deliberate, years-long repositioning away from the technology sector's most celebrated names and toward a surprising new bet: a newspaper company founded before the Civil War.

The $4.5 Billion AI Exit

Berkshire Hathaway's Q4 2025 13F filing disclosed that the company sold significant portions of two stocks closely tied to the artificial intelligence revolution: Apple and Amazon [2].

Apple, which had been Berkshire's single largest equity holding since 2016, saw another trim of approximately 10.3 million shares, reducing the position to roughly 228 million shares [3]. While Apple remains Berkshire's top holding at approximately 22.6% of the portfolio, the sale continued a relentless drawdown that has seen Buffett slash the position by 74% since late 2023 [4]. The Apple stake that once peaked at over 900 million shares has been whittled down across eight consecutive quarters of selling.

The Amazon cut was even more dramatic. Berkshire offloaded approximately 77% of its Amazon position during the quarter, dumping an estimated $1.7 billion worth of shares [5]. The sale came as Amazon announced plans for a staggering $200 billion capital expenditure budget for 2026, largely directed toward data centers to meet demand for AI compute — the kind of aggressive, capital-intensive spending that has historically made Buffett uncomfortable [2].

Combined, the Apple and Amazon sales totaled approximately $4.5 billion, representing one of Buffett's most significant technology sell-offs since he began his Apple liquidation in 2024.

A 174-Year-Old Bet

While selling tech, Buffett was buying something far older. Berkshire established a brand-new position in The New York Times Company (NYT), purchasing 5,065,744 shares valued at $351.66 million — approximately a 3% stake in the company [6][7].

The move is rich with irony. Just six years earlier, Buffett had sold off all of Berkshire's newspaper holdings and predicted unending declines for most of the print media industry [8]. He sold the Omaha World-Herald, the Buffalo News, and dozens of other papers to Lee Enterprises in 2020, seemingly closing the book on his lifelong love affair with newspapers.

But The New York Times is not most newspapers. Founded in 1851 — making it 174 years old — the Times has executed what may be the most successful digital transformation in media history. Its Q4 2025 earnings revealed record digital revenue, with total digital revenues surpassing $2 billion for the first time [9]. The company added 1.4 million net new digital subscribers in 2025, bringing its total digital subscriber base to 12.8 million [10]. Digital-only subscription revenue grew 14% to $1.43 billion, fueled by a 24% increase in bundle and multiproduct subscribers [9].

The numbers that likely caught Buffett's eye: adjusted operating profit grew more than 20% to $550 million, operating margin expanded to 19.5%, and free cash flow reached approximately $551 million [9]. For a value investor who has always prized durable competitive advantages and strong cash generation, the Times presents a compelling case — a subscription flywheel powered by journalism, games like Wordle, cooking content, and The Athletic that has proven remarkably sticky.

"Buffett has always been attracted to businesses with pricing power and subscription economics," noted one Morningstar analyst. The Times, with its rising average revenue per user of $9.72 for digital-only subscriptions, fits that template [9].

Source: Berkshire Hathaway Quarterly Reports

Data as of Nov 1, 2025CSV

The Cash Mountain

The individual stock trades, dramatic as they are, are only part of a much larger pattern. Throughout 2024 and 2025, Buffett engaged in what amounted to the largest cash-building campaign in corporate history.

Berkshire's cash and short-term Treasury holdings ballooned from $168 billion at the end of 2023 to a record $382 billion by September 2025 [11][12]. The conglomerate's $305 billion in short-term Treasuries alone would rank it 10th among foreign holders of U.S. debt if counted as a country — and exceeds the Federal Reserve's own holdings of approximately $195 billion in T-bills [11].

The timeline of this build-up is striking:

Source: Berkshire Hathaway quarterly reports [12][13]

This cash accumulation was driven by more than seven consecutive quarters of being a net seller of equities — shedding more than $75 billion in stocks in Q2 2024 alone [13]. Buffett addressed the growing pile at Berkshire's 2025 annual meeting, noting that while he would prefer to deploy the capital into productive assets, he refused to overpay simply for the sake of being fully invested.

Source: FRED / Federal Reserve Economic Data

Data as of Mar 8, 2026CSV

The Broader Portfolio Picture

Beyond the headline Apple and Amazon sales, Berkshire's Q4 2025 filing showed additional moves that underscore Buffett's cautious stance. The company trimmed its Bank of America position, continuing a long-running reduction of its once-massive bank bet [2]. It also sold shares of VeriSign.

On the buying side, beyond the NYT purchase, Buffett added 368,055 shares of Domino's Pizza, worth approximately $1.09 billion, representing a 12% increase from the prior quarter [14]. The pizza chain — the largest in the world with 22,000 global stores — exemplifies the kind of franchise-model, asset-light business Buffett has long favored. Domino's makes its money primarily on franchise fees rather than selling pizza directly, producing high margins with minimal capital requirements.

Berkshire also continued building its position in Alphabet (GOOG), which it had initiated in Q3 2025 with an approximately $4.3 billion purchase of nearly 18 million shares [15]. The Alphabet position, which represents about 1.7% of the portfolio, suggests Buffett wasn't abandoning technology entirely — just being highly selective about which tech companies deserved his capital.

What the Sales Signal About AI

Buffett has never been a technology enthusiast. He famously avoided tech stocks for decades, only investing in Apple after recognizing it as a consumer products company with an ecosystem moat rather than a pure technology play. His comfort with Apple was always rooted in the iPhone's dominance as a consumer device, not in the company's AI research or cloud ambitions.

The current AI investment cycle presents precisely the kind of environment that has historically made Buffett skeptical. Companies across the technology sector are committing hundreds of billions of dollars in capital expenditure to build AI infrastructure, with uncertain returns. Amazon's $200 billion capex plan for 2026 was perhaps the most vivid example of this spending spree [2]. Microsoft, Google, and Meta have announced similarly aggressive investment programs.

For Buffett, who once said his favorite holding period is "forever," the AI boom bears an uncomfortable resemblance to previous technology cycles where massive capital deployment preceded periods of oversupply and disappointing returns. His sales don't necessarily signal a belief that AI will fail — but they suggest deep skepticism about the valuations being assigned to companies that must spend aggressively for years before the technology's full economic potential is realized.

The Federal Reserve's interest rate trajectory adds another dimension to Buffett's positioning. With the fed funds rate declining from 5.33% in mid-2024 to 3.64% by early 2026, Berkshire's massive Treasury position has been earning substantial risk-free returns — a guaranteed yield that, in Buffett's calculus, may compare favorably to the uncertain returns of capital-intensive AI investments [16].

The Abel Era Begins

Greg Abel, 62, a Canadian businessman who had served as Berkshire's vice chairman of non-insurance operations, officially became CEO on January 1, 2026 [1]. Buffett, who turned 95, remains chairman of the board and has said he plans to continue visiting Berkshire's Omaha headquarters regularly — though he has promised to "go quiet ... sort of" and leave all decision-making to Abel [17].

Buffett's endorsement of his successor was characteristically blunt: "The decision to keep every share is an economic decision because I think the prospects of Berkshire will be better under Greg's management than mine" [17].

The question now facing Abel is what to do with Berkshire's extraordinary cash position. The $382 billion war chest represents both an opportunity and a burden. In a declining rate environment, the yield on short-term Treasuries will compress, creating pressure to deploy capital into productive assets. Abel inherits the largest acquisition war chest in corporate history, along with a portfolio that still counts Apple, American Express (15.7% of the portfolio), Coca-Cola (11%), Bank of America (10.1%), and Chevron (7.6%) among its core holdings [18].

Early signs suggest Abel is already putting his stamp on the portfolio. Berkshire resumed share buybacks in early 2026, a practice Buffett had paused during the cash accumulation phase, and Abel himself purchased Berkshire shares — a symbolic vote of confidence in the company's future under his leadership [19].

A Farewell in Trades

Warren Buffett's final quarter as CEO was, in many ways, a distillation of six decades of investing philosophy. The sales of Apple and Amazon reflected his lifelong aversion to overpaying for growth and his skepticism of capital-intensive business models with uncertain payoffs. The purchase of The New York Times reflected his enduring belief in the power of durable brands with loyal audiences and strong cash generation. And the mountain of Treasury bills reflected the patience that made him the most celebrated investor in history — the willingness to sit and wait, earning a modest but certain return, until the right opportunity presents itself.

Whether the AI revolution will prove Buffett right or wrong is a question that may take years to answer. But for the 174-year-old New York Times, having the Oracle of Omaha validate its digital transformation in his final act as CEO is, at the very least, a powerful endorsement — one worth $352 million.