Key Takeaways
- • Pershing Square argues Wall Street is penalizing hyperscalers (Microsoft, Meta, Amazon, Alphabet) for AI build-out costs while underweighting future revenue and earnings growth from those investments.
- • The four hyperscalers could spend a combined $740 billion in 2026, up from $140 billion in 2023, with analysts expecting more than $1 trillion in 2027, per the firm's analysis.
- • Semiconductor stocks rose 44% and technology hardware stocks gained 36% this year, while the hyperscalers' shares rose just 8% on average despite roughly 27% earnings growth.
- • Pershing Square's Q3 letter said the typical S&P 500 company fell 4% in Q3 and the average constituent gained just 1% through Sept. 30, versus a 13% gain for the index.
- • Pershing Square says Amazon's AWS revenue rose 39% in Q3 and that AI spending is driving growth across ad targeting, Meta's Muse assistant, and Azure and Microsoft 365 Copilot.
Summary
The firm highlights a widening gap between index performance and underlying fundamentals. The S&P 500 advanced slightly more than 2% in Q3, yet the typical constituent fell 4%, and more than 60% of index members ended the quarter lower. Through Sept. 30, the benchmark was up 13% for the year while the average constituent gained only 1%. Estimates for the median S&P 500 company's earnings per share rose 4% in the quarter and nearly 10% over the first nine months, a disconnect Pershing Square says creates opportunities in businesses whose fundamentals have improved despite weaker share prices.
AI infrastructure suppliers have led the rally. The S&P 500 Semiconductor Index has advanced 44% this year and the technology hardware index 36%. By contrast, Microsoft (MSFT), Meta Platforms (META), Amazon (AMZN) and Alphabet (GOOGL) have averaged an 8% share price increase despite roughly 27% earnings growth. Pershing Square attributes the compression in hyperscaler price-to-earnings multiples to investor concern about the returns these companies will earn on elevated capital expenditure.
Pershing Square characterizes hyperscaler capex as "offensive rather than defensive," arguing the companies are investing to accelerate growth from positions of advantage in distribution, customer relationships, proprietary data, security and balance sheet strength. The firm cites Amazon's AI investment as boosting AWS, where revenue rose 39% in Q3, and says AI is improving Meta's ad targeting and supporting faster revenue growth. It also points to Meta's Muse assistant, which passed 10 million downloads within weeks, and to AI-driven opportunities at Microsoft through Azure, Microsoft 365 Copilot and GitHub. The firm notes the four companies could spend a combined $740 billion in 2026, versus $140 billion in 2023.
Pershing Square's central argument is that analyst estimates penalize hyperscalers for current build-out costs while assigning limited value to future revenue and earnings growth. Analysts expect heavy AI spending to continue for several years but do not foresee revenue growth accelerating much beyond this year and next. The firm says much of 2026 revenue growth stems from AI investments made in 2024 and 2025, and expects current higher spending to drive stronger growth that could exceed forecasts. On Stocktwits, retail sentiment on PS stock was reported as neutral, and the stock has gained 92% year-to-date.
Go deeper than the headline
Research every firm in this story, track the ones that matter and get the daily briefing — free for 7 days.