On November 20, 2026, AI stocks experienced a significant selloff, with Nvidia falling 14.0%, the Nasdaq 100 fund (QQQ) declining 5.7%, and the S&P 500 fund (SPY) losing 3.6% [1][2]. Microsoft fell 2.8% while Meta fell 5.9%, revealing divergent financial strategies among major AI investors [1][2]. Microsoft nearly doubled its capital spending to $116 billion in the year to June 2026, yet maintained a unique position among big AI buyers by paying down $3 billion in debt rather than borrowing [1]. In contrast, Meta stopped all stock buybacks—dropping from $26 billion in the first nine months of 2025 to $0 in the same period of 2026—and instead borrowed $25 billion to fund capital spending that reached $83 billion [2].
Microsoft's financial position demonstrates exceptional strength for continued AI investment. The company brought in $183 billion in cash from operations in the year to June 2026, spent $116 billion on capital spending, paid shareholders $48 billion in buybacks and dividends, and still had $18 billion left over [1]. Microsoft's interest bill of $3.1 billion was actually less than the $3.3 billion it earned on cash holdings [1]. Analysts project Microsoft can spend approximately $39 billion more from its own cash in fiscal 2027, with potential to borrow an additional $188 billion before debt would equal a year's cash flow, given its debt was only about one-sixth of annual cash generation at September 30 [1].
Meta's situation contrasts sharply, as the company has become increasingly reliant on debt financing. Meta's business brought in $97 billion in the first nine months of 2026, with 86 cents of every dollar going to capital spending, up from 61 cents a year earlier [2]. The company's debt nearly tripled from $29 billion to $84 billion, with additional borrowing of $15 billion and €9 billion in November bonds, plus $14 billion in preferred stock sales [2]. Meta's interest bill more than tripled to $2.3 billion [2]. Despite this aggressive borrowing, Meta can still spend approximately $31 billion more from its own cash in 2027, or up to $89 billion if it borrows further, though its borrowing capacity is now the smallest among the three cash-rich buyers (Microsoft, Alphabet, and Meta) [2].
The divergent strategies have implications for Nvidia and semiconductor demand. Microsoft's $39 billion additional spending capacity from cash alone, plus potential $188 billion in borrowing capacity, positions it as a more sustainable long-term AI infrastructure investor [1]. Meta's $31 billion additional spending from cash, or up to $63 billion with borrowing after November bond sales, translates to approximately $16 billion worth of Nvidia sales based on Nvidia's historical 52-cent ratio to major buyer spending [2]. The market's negative reaction on November 20 suggests investor concerns about the sustainability of AI spending growth, particularly given Meta's shift toward debt financing and away from shareholder returns.
CONCLUSION
Microsoft's ability to simultaneously increase AI capital spending to $116 billion while paying down debt positions it as the most financially resilient major AI buyer, with $39 billion in additional spending capacity from cash alone [1]. Meta's pivot to aggressive debt financing—borrowing $25 billion while eliminating buybacks—signals a more constrained financial position despite continued AI investment capacity of $31 billion from operations [2]. The November 20 market selloff reflects investor concerns about the sustainability and returns on massive AI infrastructure spending, particularly as companies like Meta increasingly rely on debt rather than operational cash flow to fund expansion.