Berkshire's Alphabet Position Down ~$3 Billion; Stock Picks Over Past Decade Have Underperformed the Broader Market
nashnova research
Berkshire Hathaway holds roughly 106 million Alphabet shares now sitting on an estimated $3 billion unrealized loss after last week's earnings miss; *Barron's* argues the stumble is symptomatic — new stock picks over the past decade have broadly lagged the S&P 500, and incoming CEO Greg Abel faces a strategic crossroads.
Where exactly did the Alphabet bet go wrong?
Berkshire holds about 106 million Alphabet shares at an estimated cost above $300 per share, including roughly $10 billion subscribed in June at ~$350/share as part of Alphabet's $85 billion equity raise.
Last week Alphabet missed earnings expectations; the stock fell ~10% to $278, leaving Berkshire with an estimated $3 billion paper loss.
This means → a late entry at a high cost basis left no margin for error — one earnings miss wiped out all gains and flipped the position into the red.
Why has the past decade of new picks underperformed?
Since late 2019 the S&P 500 has returned roughly 150% and tech stocks roughly four-fold — yet Berkshire's new investments failed to keep pace.
Apple is the starkest "sold too early" case: Berkshire held over 900 million shares in 2023, then trimmed ~75% to 228 million shares at an estimated average sale price of ~$200, versus Apple's current ~$340.
Bank stocks were "sold entirely": JPMorgan, Wells Fargo, and Goldman Sachs were all exited between 2020 and 2023 at roughly one-third of their current prices.
In plain terms = missed the tech mega-rally + exited quality names too early — a double opportunity cost.
Occidental Petroleum — a big bet with modest returns?
Berkshire has built a ~26% stake in Occidental since 2022, currently worth about $16 billion, but up only ~20% from cost.
This reflects one of the weakest performances among energy-sector ETF constituents — a concentrated bet that has not delivered excess returns.
Are there any bright spots?
Five Japanese trading houses, acquired around 2020, have doubled in value to roughly $45 billion combined.
Chevron is up over 100% since 2020, with a position worth ~$20 billion; Chubb has gained ~50% in about three years.
Yet *Barron's* notes these winners' combined returns roughly match a simple S&P 500 index buy, while positions in ViacomCBS, HP, Paramount, GM, and Constellation Brands ended in losses or flat.
Will new CEO Abel change the stock-picking playbook?
Berkshire's real performance pillars remain decades-old positions: American Express and Coca-Cola, bought 30–40 years ago, are worth ~$48 billion and ~$38 billion respectively; Apple, despite heavy trimming, still leads the portfolio at over $75 billion; Moody's is another long-held name.
Greg Abel, who took over as CEO in early 2026, has said he will keep Apple, Coca-Cola, Moody's, and American Express as core holdings.
Whether he will focus on stock selection or lean more toward operational improvements and acquisitions remains unclear.
This means → Berkshire's stock-picking capability is in a generational transition gap — and the Alphabet loss is exposing exactly that vulnerability.
市场有风险，内容仅供研究参考，不构成投资建议。