As reported by CNBC, Fortune and others (Berkshire Q2 2026).
Berkshire’s cash pile barely moved in percentage terms — but buried inside the Q2 numbers is a $10 billion primary placement that directly funds Alphabet’s AI infrastructure, and a total stake that now ranks among Berkshire’s largest positions.
Reporting from The Motley Fool, alongside Berkshire’s Q2 2026 filings covered by CNBC and Fortune, shows that Berkshire Hathaway ended the second quarter with $365.5 billion in cash and equivalents — down from a record $397.4 billion at the end of Q1, a roughly $32 billion decline and the first drawdown in more than three years. The right-sized read: $365.5 billion is still a near-record hoard, Berkshire remains overwhelmingly positioned in cash and Treasury bills, and a single quarter’s movement partly reflects buybacks and Treasury-bill timing as much as any strategic repositioning. This is not a spending spree.
Inside the quarter, net profit roughly doubled to approximately $25.67 billion, Berkshire repurchased around $4.5 billion of its own shares, and — for the first time in over three years, after approximately fourteen consecutive quarters of net selling — it was a net buyer of public equities, with roughly $20 billion more bought than sold across several names. That net buying figure spans a portfolio of positions; it is not all Alphabet.
The centerpiece is Alphabet. Berkshire now holds approximately 68.46 million Class A shares and 17.94 million Class C shares, a combined position worth roughly $31.2 billion at current prices — approximately 9% of its entire investment portfolio, making it one of Berkshire’s largest single holdings. The composition matters: roughly $21 billion sits in ordinary public-market shares accumulated over time, while $10 billion was wired directly to Alphabet in June 2026 via a private placement. Warren Buffett has said he personally initiated the position in Q3 2025 (approximately 17.8 million shares, roughly $4.3 billion at entry); Greg Abel, who now steers capital allocation, has been topping it up since.
HOW THE POSITION WAS BUILT
Q3 2025
Warren Buffett personally initiates Alphabet position — ~17.8M shares, ~$4.3B. Long-held conviction, not a sudden AI pivot.
Q4 2025 – Q1 2026
Greg Abel expands the position across public shares. Berkshire’s cash remains near record; Alphabet stake grows to ~$21B in public shares.
June 2026
Berkshire wires $10B into Alphabet via private placement — part of Alphabet’s ~$80B equity raise to fund AI-infrastructure capex. Primary capital: new money at the balance sheet.
Q2 2026 Close
Total Alphabet stake: ~$31.2B (~9% of portfolio). Berkshire’s cash: $365.5B — still historically vast. First net stock buyer in 3+ years.
The key insight: Berkshire’s $10 billion June tranche was not a secondary-market purchase — it was a primary placement, meaning the cash went directly onto Alphabet’s balance sheet to fund its AI infrastructure program. The most hype-averse institutional investor in the world is not merely betting on AI; it is financing the buildout.
Berkshire Hathaway’s cash and equivalents fell to $365.5 billion at the end of Q2 2026 from a record $397.4 billion a quarter earlier – a roughly $32 billion decline and the first drawdown in more than three years, as it bought back ~$4.5B of stock and turned a net buyer of equities for the first time in over three years. The pile remains near-record, so this is a modest dip, not a spending spree. Sources: Berkshire Q2 2026 filing; CNBC.
The Structural Read
The easy version of this story is “Buffett finally buys AI.” That framing misses the structural point on two counts. First, Buffett started the position in Q3 2025 — this is a long-held conviction being scaled, not a sudden conversion. Second, and more important, the mode of investment distinguishes what Berkshire is actually doing from any normal equity allocation.
Alphabet is running one of the largest capital expenditure programs in corporate history. Its full-year 2026 capex budget sits at roughly $175–185 billion; Q1 2026 alone came in at approximately $35.67 billion, more than double year-over-year, the bulk of it directed at AI infrastructure — data centers, custom TPUs, and the model stack behind Gemini. Alphabet did not raise ~$80 billion in new equity to pad a treasury; it raised it to fund that buildout. When Berkshire participated in that raise with $10 billion, it was acting as infrastructure financier, not as a public-market stock picker.
That reframes what “value investing meets AI” actually looks like in practice. It is not a punt on a frontier lab with no revenue. It is infrastructure financing routed through the profitable incumbent that owns its own chips, its own models, and its own distribution — a structure that the Map of AI locates at multiple layers of the stack simultaneously. Alphabet is not a single-layer bet; it is a vertically integrated AI system, and Berkshire’s capital is now part of what keeps that system expanding.
Map of AI — Infrastructure Layer
The Profitable Incumbent as Infrastructure Vehicle
Berkshire’s entry into Alphabet’s primary raise illustrates a pattern now visible across the AI buildout: the safest way to finance frontier compute is through the incumbents who already generate the cash flows to service the capex. Alphabet’s $175–185B annual capex program is underwritten by search, cloud, and YouTube revenues that have compounded for two decades. Berkshire is not financing a speculative model; it is financing the infrastructure arm of a profitable conglomerate. That is the compute-securitization logic in its most conservative possible form.
The Financial Clock
“The same week Berkshire put $10 billion of primary capital into Alphabet’s AI-infrastructure raise, a leveraged AI-native hedge fund saw its assets collapse from roughly $45 billion to $10 billion on public-market AI bets. Two opposite answers to the same question: how do you get AI exposure without being destroyed by the cycle?”
The contrast is the sharpest possible illustration of what the AI financial clock looks like when it runs at two different speeds simultaneously. Leverage plus public-market AI exposure produced a rapid, catastrophic drawdown. Zero leverage plus primary-market placement in the most solvent AI infrastructure operator produced a quiet $31 billion position that barely moved the needle on Berkshire’s balance sheet. Same underlying bet on AI’s continued buildout; entirely different risk topology.
The honest caveats run in a specific direction. “Safe” is relative: Alphabet is itself making one of the largest capital bets in corporate history, and Berkshire is now exposed to the AI-capex cycle, just through its most creditworthy participant. If AI infrastructure spending compresses sharply — delayed returns, oversupply of compute, a demand shortfall — Alphabet’s stock is not immune, and neither is a $31 billion position in it. The $10 billion primary placement ties Berkshire directly to the capex cycle, not merely to Alphabet’s earnings. That is a meaningful exposure, even if it is structured as conservatively as possible.
1. THE PRIMARY MARKET IS WHERE THE REAL AI BETS ARE BEING PLACED
Berkshire’s $10 billion was not a secondary-market allocation that recycled existing shares — it was new capital that went directly onto Alphabet’s balance sheet. As the AI infrastructure buildout scales toward $175–185 billion annually at Alphabet alone, the meaningful financial signal is not who is buying the stock; it is who is participating in primary raises. Berkshire just joined that list, alongside the Apollo- and Blackstone-structured compute-securitization vehicles that have been quietly financing the same cycle from the credit side.
2. GREG ABEL’S CAPITAL ALLOCATION PHILOSOPHY IS NOW VISIBLE
Buffett started the Alphabet position in Q3 2025, but Abel has been steering it since. The $10 billion primary placement in June 2026 is the clearest evidence yet of how Abel approaches large-scale technology exposure: not through frontier labs with uncertain revenue, not through leveraged public-market plays, but through the profitable incumbent whose AI infrastructure ambitions are underwritten by decades of durable cash flows. That is a legible, consistent philosophy — and it differs from what most institutional allocators are doing in AI.
3. THE CAPEX-CYCLE RISK IS REAL, EVEN FOR THE MOST CONSERVATIVE ENTRY POINT
Berkshire’s entry through a primary placement at a profitable incumbent is the most de-risked structure available in AI infrastructure finance. It is still AI-capex-cycle exposure. If compute demand disappoints relative to the $175–185 billion annual build, Alphabet’s returns compress and the $31 billion position moves accordingly. The structural conservatism of the vehicle does not neutralize the underlying cycle risk — it just means Berkshire entered at the strongest possible point on the credit and earnings quality spectrum. The risk is managed; it is not absent.
Gennaro is the creator of FourWeekMBA, which reached about four million business people, comprising C-level executives, investors, analysts, product managers, and aspiring digital entrepreneurs in 2022 alone | He is also Director of Sales for a high-tech scaleup in the AI Industry | In 2012, Gennaro earned an International MBA with emphasis on Corporate Finance and Business Strategy.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
