Buffett's final portfolio moves reveal a radical thesis about AI, media, and the end of American exceptionalism—as Greg Abel inherits the most powerful war chest in corporate history
Executive Summary
- Warren Buffett's final quarter as CEO saw Berkshire dump 77% of Amazon, trim Apple further, and make a surprise $352M bet on The New York Times—marking 13 consecutive quarters of net equity sales and the most decisive risk-off positioning in the conglomerate's 60-year history.
- Greg Abel officially took the helm on January 1, 2026, inheriting a $382 billion cash fortress that earns more in Treasury interest than most Fortune 500 companies generate in total revenue—raising existential questions about whether Berkshire can deploy capital at scale or must finally pay its first regular dividend.
- The February 28 annual letter—Abel's first—arrives in a week that also features Trump's State of the Union, Nvidia's earnings, and the Ukraine war's fourth anniversary, making it a potential inflection point for markets already rattled by the SCOTUS tariff ruling and $175 billion in refund uncertainty.
Chapter 1: The Final Act — Buffett's Q4 2025 Portfolio Revolution
On February 17, 2026, Berkshire Hathaway's Q4 2025 13F filing landed like a quiet earthquake. It was Warren Buffett's last quarterly snapshot as CEO—and it told a story of systematic retreat from the market he'd dominated for six decades.
The headline numbers were striking. Berkshire sold 77% of its Amazon position in a single quarter, reducing a $2.1 billion stake to roughly $457 million. Apple, once half the equity portfolio at $175 billion, was trimmed another 4.3%—continuing a two-year liquidation that had already slashed the position by 70%. Bank of America saw further reductions. The pattern was unmistakable: 13 consecutive quarters of net selling, the longest such streak in Berkshire's history.
But the most revealing move wasn't a sale. It was a purchase.
Berkshire took a $352 million stake in The New York Times—5.07 million shares, roughly 3% of the company. This from the man who in 2020 sold his entire portfolio of 31 daily newspapers, calling the industry "toast." The reversal wasn't nostalgia. It was a thesis.
| Position | Action | Details |
|---|---|---|
| Amazon (AMZN) | Sold 77% | 7.7M shares sold, 2.3M remaining ($457M) |
| Apple (AAPL) | Trimmed 4.3% | 10.3M shares sold, $62B position remains |
| New York Times (NYT) | New position | 5.07M shares, $352M (3% stake) |
| Bank of America (BAC) | Trimmed | Continued multi-quarter reduction |
| Alphabet (GOOGL) | Added | Built to $5.4B stake (14th-largest holding) |
| Chevron (CVX) | Added | Increased energy exposure |
The portfolio concentration tells its own story: the top five holdings—Apple, American Express, Bank of America, Coca-Cola, and Chevron—account for 70.9% of the $274.2 billion equity portfolio. The top ten represent 88%. Buffett wasn't diversifying. He was simplifying—stripping the portfolio down to a core of businesses he understood intimately, while letting the cash pile grow to extraordinary proportions.
Todd Combs' departure to JPMorgan in December 2025 added another dimension. Combs was widely credited with championing Berkshire's technology positions, including Amazon. His exit may have accelerated the tech liquidation. Ted Weschler, the remaining investment manager, has a different temperament—more value-oriented, less growth-tilted.
Chapter 2: The $382 Billion Question — Cash as Conviction
The cash pile is the elephant in every Berkshire conversation. At $381.7 billion in cash and short-term Treasuries as of year-end 2025, Berkshire now holds more liquid assets than the GDP of Finland, Portugal, or New Zealand.
To put this in perspective: at a 4.5% Treasury yield, Berkshire's cash pile generates approximately $17 billion annually in interest income alone. That's more than the entire annual revenue of companies like Marriott International or Starbucks. The conglomerate is, in effect, running one of the world's largest government bond funds as a side business.
Why so much cash? Buffett's public explanations have been characteristically sparse. At the May 2025 shareholder meeting—his final as CEO—he offered only that he saw "nothing attractive at current prices" and that he preferred "the certainty of Treasuries to the uncertainty of equities at these valuations."
But the timing tells a deeper story. Berkshire's selling accelerated precisely as the AI bubble inflated asset prices across technology, pushed the Buffett Indicator (total market cap to GDP) above 220%, and drove the S&P 500's CAPE ratio above 40. Buffett, who famously avoided the dot-com bubble, appeared to be making the same call about AI-era valuations.
Historical comparison: Berkshire's cash as percentage of total assets
| Year | Cash ($B) | % of Total Assets | Market Context |
|---|---|---|---|
| 1999 | $36 | 20% | Dot-com peak |
| 2005 | $43 | 22% | Post-bust, pre-housing bubble |
| 2008 Q3 | $25 | 10% | Deployed into crisis (GS, GE) |
| 2020 Q1 | $137 | 17% | COVID crash—limited deployment |
| 2024 Q4 | $325 | 28% | AI bubble inflation |
| 2025 Q4 | $382 | 31% | Record—systematic selling |
The 2008 comparison is instructive. When Lehman Brothers collapsed, Buffett had $25 billion in cash and deployed it aggressively—investing $5 billion in Goldman Sachs at terms so favorable they became Wall Street legend. When COVID crashed markets in March 2020, Buffett had $137 billion but barely deployed any of it, later admitting he "didn't see anything that attractive."
Now Abel inherits $382 billion—fifteen times the 2008 war chest—in a market environment characterized by AI-inflated valuations, SCOTUS tariff chaos, geopolitical fragmentation, and the SaaSpocalypse decimating software sector valuations. The question isn't whether Berkshire will deploy. It's whether any single acquisition or series of investments can move the needle for a company this large.
Chapter 3: The Abel Doctrine — From Stock Picker to Operational Machine
Greg Abel is not Warren Buffett. This is not a criticism—it's a strategic reality that will reshape the world's most famous conglomerate.
Where Buffett was fundamentally a capital allocator who happened to run an operating company, Abel is an operations executive who must learn to allocate capital at a scale unprecedented in corporate history. His career trajectory—rising through MidAmerican Energy (now Berkshire Hathaway Energy) to lead all non-insurance operations—speaks to a different skill set: infrastructure management, regulatory navigation, utility-scale project execution.
The early signals of the Abel era are already visible:
Leadership changes: Abel appointed Charles Chang as CFO and Nancy Pierce to lead GEICO. These aren't cosmetic moves. Pierce's mandate at GEICO is to close the technology gap with Progressive, which had pulled ahead during Buffett's era of tech-skepticism. Under her leadership, GEICO's combined ratio has stabilized at 84.3%—a stellar number that suggests the digital transformation is working.
The Alphabet bet: Perhaps the most Abel-coded move in the Q4 filing was the continued building of a $5.4 billion stake in Alphabet. Where Buffett famously avoided tech he didn't understand, Abel appears comfortable with "infrastructure tech"—companies whose moats are built on physical assets (data centers, fiber networks) rather than pure software. Alphabet's Google Cloud, which grew 48% in Q4 2025, fits this framework perfectly.
The dividend question: For the first time in Berkshire's history, serious analysts are whispering the "D-word." With interest rates softening and the cash pile growing faster than deployment opportunities, the mathematical case for a regular dividend is becoming harder to ignore. Abel may initiate Berkshire's first-ever regular dividend by end of 2026—a move that would represent a philosophical break from the Buffett doctrine of perpetual reinvestment.
Projected 2026 operating performance:
| Segment | Projected Earnings | Key Driver |
|---|---|---|
| Insurance (GEICO, Gen Re, BHRG) | $12-14B | GEICO turnaround, hard market |
| BNSF Railway | $6-7B | Rail volume recovery |
| Berkshire Hathaway Energy | $4-5B | Renewables + regulated returns |
| Manufacturing/Services | $14-16B | Precision Castparts, Lubrizol |
| Investment Income | $17-19B | $382B at ~4.5% |
| Total Operating Earnings | $42-47B |
Chapter 4: Scenario Analysis — What Does the Cash Fortress Buy?
Scenario A: The Great Deployment (30%)
Thesis: A market dislocation—triggered by the SCOTUS tariff chaos, a SaaSpocalypse credit event, or an AI capex cycle reversal—creates Buffett-style "fat pitch" opportunities. Abel deploys $100-200 billion in a combination of whole-company acquisitions and distressed equity positions.
Historical precedent: In 2008-2009, Berkshire invested approximately $25 billion at distressed prices. Adjusted for Berkshire's current scale, a proportional deployment would be $150-200 billion. The private credit crisis (now $3 trillion with rising defaults), the software sector implosion, and potential hyperscaler debt distress all present plausible catalysts.
Trigger conditions: VIX above 40, investment-grade credit spreads widening beyond 200bps, a major AI-funded company failing to refinance debt.
Probability reasoning: At 30%, this reflects the genuine possibility of a 2026 market dislocation. The SCOTUS IEEPA ruling has created $175 billion in refund uncertainty. Morgan Stanley has flagged $400 billion in AI-related leveraged lending. UBS warns of 13% default rates in private credit. The ingredients exist—but timing remains uncertain.
Scenario B: The Slow Transformation (45%)
Thesis: Abel gradually pivots Berkshire from a Buffett-era holding company into a more institutional conglomerate. He initiates a regular dividend, makes 2-3 mid-sized acquisitions ($10-30 billion each), and allows the cash pile to stabilize at $200-250 billion. The equity portfolio continues to de-risk, shifting from concentrated bets to a broader, infrastructure-heavy allocation.
Historical precedent: This mirrors the trajectory of Berkshire itself during the 1990s-2000s transition from a textile company holding insurance businesses to a diversified conglomerate. The shift was gradual, spanning a decade. Also comparable: the post-Jack Welch transition at GE, though Abel's starting position is incomparably stronger.
Trigger conditions: No major market dislocation in 2026. Federal Reserve maintains rates above 3.5%, keeping Treasury income attractive. GEICO turnaround and BNSF recovery provide organic earnings growth.
Probability reasoning: At 45%, this is the base case because it requires no extraordinary events—just competent management of an extraordinarily well-positioned enterprise.
Scenario C: The Buffett Overhang (25%)
Thesis: Abel struggles to escape Buffett's shadow. Every capital allocation decision is compared to the Oracle's track record. The cash pile grows beyond $400 billion as Abel proves too cautious to deploy at scale. Berkshire begins trading at a persistent conglomerate discount, and activist investors—or even Berkshire's own board—push for a breakup.
Historical precedent: The most relevant parallel is Berkshire's own history during the late 2010s, when critics argued Buffett was hoarding cash unnecessarily. The "conglomerate discount" phenomenon afflicted companies like General Electric and Honeywell after leadership transitions. More ominously, the post-Welch GE destroyed $300 billion in shareholder value over two decades.
Trigger conditions: Abel's first letter disappoints on February 28. Berkshire misses a major acquisition opportunity that later proves transformative. Apple or another major holding declines sharply, creating mark-to-market losses that overshadow operating performance.
Probability reasoning: At 25%, this reflects the genuine difficulty of CEO succession at iconic companies. Research by Harvard Business School's Joseph Bower shows that roughly 30% of "iconic CEO" successions result in sustained underperformance within five years.
Chapter 5: Investment Implications — The Berkshire Signal
Berkshire's positioning isn't just about Berkshire. When the most patient, most successful investor in history spends two years systematically converting equities to cash, the signal is worth parsing.
What Buffett's selling tells us:
-
Valuation warning. The Buffett Indicator at 220% is historically associated with poor forward returns. Buffett sold into the 1999 dot-com peak and was vindicated. His 2024-2025 selling has the same flavor.
-
AI skepticism. The Amazon dump—selling into AI hype—suggests Buffett sees AI capex as potentially destructive of returns for the companies making the investments, even if the technology itself proves transformative. This mirrors his dot-com insight: the internet changed the world but destroyed most internet company investors.
-
Dollar skepticism. The shift toward Chevron and energy exposure, combined with the massive Treasury position (which benefits from rate certainty, not rate direction), suggests hedging against fiscal instability. With US debt at 120% of GDP and DOGE failing to deliver meaningful savings, Buffett may be positioning for a currency event.
What the NYT bet tells us:
The New York Times purchase is Buffett's most contrarian final move. In a world of AI-generated content, algorithm-driven news, and collapsing media business models, buying a 172-year-old newspaper seems anachronistic. But the Times has 12 million digital subscribers, $550 million in annual free cash flow, and a brand moat that AI cannot easily replicate. If AI destroys commodity content (as seems likely), premium trusted sources become more valuable—not less. It's the anti-AI bet hiding inside a media trade.
Asset class implications:
| Asset | Berkshire Signal | Implication |
|---|---|---|
| US Equities | Persistent net seller | Overvalued at current levels |
| Short-term Treasuries | Record allocation | Cash preservation paramount |
| Energy/Commodities | Adding Chevron | Inflation/real asset hedge |
| Insurance | Core holding | Hard market cycle continues |
| Premium Media | NYT new position | Trust premium in AI era |
| Tech/AI | Amazon dump, Apple trim | Capex intensity destroys returns |
Conclusion: The Week That Decides Everything
The February 28 annual letter arrives at an extraordinary confluence. Trump delivers his State of the Union on February 24, framing the tariff chaos and DHS shutdown as strength. Nvidia reports earnings on February 25, potentially validating or puncturing the AI capex thesis. Ukraine marks four years of war as Russia hammers its power grid. The SCOTUS tariff ruling's aftershocks ripple through global trade.
Into this maelstrom steps Greg Abel with his first public communication as CEO of a $1 trillion company sitting on $382 billion in cash. What he says—and more importantly, what he signals about deployment, dividends, and direction—could move markets more than any earnings report this quarter.
Buffett spent 60 years building the ultimate financial fortress. Now, as the 95-year-old Oracle watches from the chairman's seat, the question is whether his successor will use it as a launchpad or let it become a mausoleum.
The cash is patient. The market may not be.
Sources: Berkshire Hathaway 13F filing (Feb 17, 2026), Simply Wall St, Fintool, SEC filings, Apple Q1 FY2026 earnings, Benzinga, The Street
Leave a Reply