Learn Investing from the Pros | The 'Mountain of Cash' Is Shifting! Berkshire Starts Spending: Buying Back Its Own Stock and Doubling Down on Google

Over the past two years, $Berkshire Hathaway-B (BRK.B.US)$ it has been doing something that unsettles the market: selling stocks and hoarding cash.

From Q4 2022 through Q1 2026, $Berkshire Hathaway-A (BRK.A.US)$ it was a net seller of equities for 14 consecutive quarters, with cumulative net sales of approximately $195 billion. As major holdings like Apple and Bank of America continued to be reduced, the company’s cash reserves (cash and short-term U.S. Treasuries) once approached $400 billion, setting a historical record.

The market has been wondering: Is Berkshire holding off because it’s waiting for better prices, or does it lack confidence in the overall valuation of U.S. equities?But the latest earnings finally signaled a shift—the massive cash pile is starting to loosen.

Symbols

Buying back its own shares: repurchase scale hits a new high since 2021

Berkshire Hathaway spent approximately $4.5 billion repurchasing its own stock in Q2, far exceeding the $235 million in Q1, bringing total repurchases in the first half of the year to about $4.8 billion—the largest single-quarter buyback since 2021.

Judging by the pace, management’s stance has also been gradually shifting: no repurchases in April, buying began in May, and accelerated significantly in June.In June alone, Berkshire repurchased approximately 7.14 million Class B shares at an average price of about $487.98 per share, and also bought back 413 Class A shares at an average price of roughly $733,800 per share.

This speaks louder than buying shares of other companies.Under Berkshire’s repurchase policy, the company only buys back shares when the CEO and Chairman jointly determine that the stock price is below a conservative estimate of intrinsic value. Therefore, the sudden acceleration in Q2 buybacks indicates that Abel and Buffett have reached at least one consensus: Berkshire’s own shares have once again become attractively priced.

Buying stocks again: ending 14 consecutive quarters of net selling

The shift isn’t limited to share repurchases.

In the second quarter, Berkshire bought approximately $23.5 billion of publicly traded stocks and sold about $3.7 billion, resulting in net purchases of nearly $20 billion,officially ending a streak of 14 consecutive quarters of net sales.

As of the end of June, Berkshire held $364.7 billion in cash and U.S. Treasury securities (after deducting payables related to purchases of short-term government securities), down 4% from three months earlier.This marks the first quarterly sequential decline in the company's cash holdings in four years.

In addition to publicly traded stocks, Berkshire also spent $6.8 billion in the second quarter to acquire homebuilder Taylor Morrison. However, the deal did not close until July and thus was not included in the second-quarter financial figures.

The simultaneous net stock purchases, share repurchases, and acquisitions of physical businesses indicate that Greg Abel has begun deploying more aggressively the massive cash pile accumulated during the Buffett era.

But the question is: What exactly did Berkshire buy?

Of the nearly $20 billion in net stock purchases in the second quarter, at least half went to Google.

If we break down the nearly $20 billion in net stock purchases during the second quarter,it becomes clear this was not a diversified 'shopping spree' across U.S. equities.

The clearest transaction was Berkshire’s $10 billion private placement investment in June to purchase newly issued Google common shares.

This financing is primarily intended to support Google’s expansion of AI infrastructure and global computing capacity investments. Unlike Buffett’s past crisis-era deals—which included preferred shares, high dividends, and warrants—Berkshire this time bought common stock without liquidation preference or attached warrants.

Berkshire’s $10 billion private placement was priced based on Google’s closing share price of $376 that day, with $5 billion allocated to Class A common stock $Alphabet-A (GOOGL.US)$ which received a discount of approximately 6%, and $5 billion allocated to Class C common stock $Alphabet-C (GOOG.US)$ priced at $348.20 per share, representing a discount of nearly 8%.In other words, Berkshire did not receive an exceptionally steep price discount; instead, it placed a direct bet on Google’s long-term value near prevailing market prices.

This single transaction alone accounted for roughly half of Berkshire’s net equity purchases in the second quarter.Considering Berkshire had already established a Google position in the open market and that Google entered its top-five holdings for the first time as of the end of Q2, the actual proportion of funds directed to Google may be even higher—though precise changes await confirmation from future SEC filings.

Additionally, Berkshire continued to increase its stakes in several Japanese trading companies. Therefore,the true capital allocation path in Q2 more closely resembled: repurchasing Berkshire’s own shares + concentrating additional investments in Google + further expanding exposure to Japanese assets.

It’s not a broad bullish stance, but rather a shift from 'hoarding cash only' to 'selective deployment.'

Therefore, Berkshire ending its net selling should not be simplistically interpreted as Abel having turned fully bullish on U.S. equities.

First, Berkshire still holds over $360 billion in cash and short-term U.S. Treasuries; the roughly $20 billion in net stock purchases in Q2 represents only about 5% of its cash reserves. Cash remains the absolute core of its asset allocation.

Second, the newly deployed capital is highly concentrated and has not flowed noticeably into high-valuation, high-volatility speculative stocks. Berkshire chose its own shares, Google, and Japanese trading houses—assets that share common traits: strong cash flow, clear competitive moats, long-term capital return potential, and entry prices acceptable to management.

Notably, the Google transaction shows Berkshire isn’t averse to tech stocks or AI—but what it seeks isn’t just an AI narrative. It requires mature business models generating cash flows capable of supporting massive AI-related capital expenditures.

Thus, the most important signal from Q2 isn’t that 'Buffett has started buying the dip,' but rather that Berkshire has shifted from hoarding cash across the board to selectively deploying it into a few high-conviction assets.

This move reflects neither full endorsement of overall U.S. equity valuations nor a complete pivot from defense to offense. Instead, it signals that Berkshire is now selectively deploying capital when expected returns on quality assets once again exceed those of holding cash.

The mountain of cash is indeed beginning to loosen—but Berkshire isn’t buying the entire market. It’s purchasing only a select few assets it truly understands and can confidently justify on a valuation basis.

189K Views