Berkshire Bet $31 Billion on Alphabet Because the Capex Scare Is a Phase, Not a Pattern

The market fixates on Alphabet's $132 billion in annual capital spending and the resulting 20% decline in free cash flow. Investors want to know when AI infrastructure costs will end or whether they'll ever pay off. It's the same question that keeps them off the sidelines.

Berkshire Hathaway isn't waiting for clarity. Under Greg Abel, it has poured roughly $31 billion into Alphabet - making the Google parent a top-10 holding and about 6% of the portfolio. Warren Buffett admits he "made a mistake" not getting in sooner. But the person expanding the bet is Abel, who tripled the stake in his first quarter as CEO and then added another $10 billion via a private placement in June at a discount to market price.

The numbers behind the conviction are harder to ignore once you look past the capex headline.

The proof point is operating cash flow, not free cash flow

Free cash flow fell 20% year-over-year to $53.3 billion. That's the metric driving nervousness. But free cash flow is operating cash flow minus capital expenditures. Alphabet generated $185.7 billion in operating cash flow over the trailing twelve months. The gap between those two numbers - the $132 billion in capex - is concentrated, not chronic. It's a multi-year buildout of AI compute infrastructure that Alphabet itself says is responding to "unprecedented customer demand."

The mechanism matters: when the infrastructure buildout reaches a plateau - and it must, because spending can't outpace revenue growth forever - operating cash flow continues while capex normalizes. That would return free cash flow to a level closer to what Alphabet's margins already support.

Operating margin sits at 32.7%. Gross profit growth is 24%. Return on invested capital is 16.5%. Return on equity is 50%. These aren't the numbers of a company burning cash on a speculative bet. They're the numbers of a company investing heavily for one year while its core business continues to expand revenue at 20%.

The market is still pricing Alphabet like a company whose cash generation is under structural threat. The operating path says the threat is temporary and the earnings machine is already running.

Earnings are blowing through the consensus

The earnings data makes the gap between market fear and business reality even starker. Alphabet reported $9.11 per share in Q2 2026 against a consensus estimate of $2.88. Q1 2026 came in at $5.11 versus $2.63 expected.

Those beats aren't accounting quirks. They reflect revenue that keeps accelerating - $119.8 billion in Q2 against a $117.1 billion estimate - while margins hold at 38.2% EBITDA. The consensus was simply not tracking how fast AI-driven demand is converting into actual profit.

Buffett told CNBC in a recent interview that Alphabet is "more likely to be a winner based on their record than probably 90% or 95% of what gets merchandised through Wall Street." He noted that companies "don't have any choice" in the AI spending race - they're "playing a game they don't want to play." That's the inflection insight: the capex spending is competitive defense, not optionality. Alphabet has to spend to stay ahead, and it's already winning while spending.

Valuation tells the old story

Alphabet trades at 17.8 times trailing earnings and 34.7 times forward earnings. That forward multiple is elevated, yes - but the PEG ratio (price-to-earnings-growth) sits at 0.16. A PEG below 1.0 means the stock's earnings multiple is small relative to its growth rate. At 0.16, Alphabet's earnings growth is priced at a fraction of what the multiple suggests.

The math is simple. Alphabet's revenue grew 20% year-over-year. If earnings continue to beat consensus at the pace Q1 and Q2 2026 have set, forward earnings will compress that 34.7x multiple much faster than the market expects. Berkshire's $10 billion private placement bought shares at $351.81 and $348.20 - roughly 5% to 6.5% below market at the time. Even at today's $356, that entry discipline is intact.

The stock has gained 13.8% year-to-date and 87% over the trailing 12 months. Some of that move reflects growing conviction, not just momentum. The 20-day window shows the stock was essentially flat (-1%) before a 12% five-day surge. That pattern - a pause followed by a reacceleration - suggests the thesis is being repriced in chunks, not in a straight line.

What this means

Abel isn't just maintaining Buffett's legacy here. He's aggressively building a position that Buffett started but didn't fully commit to. The Taylor Morrison acquisition and the exits from Amazon, Visa, Mastercard, and UnitedHealth show Abel is willing to make decisive moves and rotate capital into where he sees the clearest path. Alphabet's AI infrastructure buildout, its earnings power, and the valuation gap between forward multiples and actual growth - that's the setup he's betting on.