Apollo warns: The credit logic of hyperscale cloud vendors is a “big gamble” of $2 trillion in cash flow

The Zhitong Finance App learned that Apollo Global Management Chief Economist Thorsten Slocke recently released a report warning about the credit prospects of hyperscale cloud service providers. He pointed out that the credit narrative of these five tech giants — Alphabet, Meta, Amazon, Microsoft, and Oracle — is based on a widely accepted forecast in the market: their combined operating cash flow will triple from about $600 billion to $2 trillion. However, whether this forecast can be fulfilled is becoming a core issue hanging over the global credit market.

The hyperscale cloud service provider groups Alphabet, Meta, Amazon, Microsoft, and Oracle mainly supported an unprecedented capital expenditure cycle with debt financing at a time when free cash flow was negative. The reason the market is accepting this leverage is that it is expected that AI-related revenue will eventually drive a sharp rebound in cash generation.

In the past, tech giants relied on abundant cash flow from software, advertising, and cloud services to maintain an asset-light image with strong cash flow, low leverage, and high profit margins, and enjoyed extremely low financing costs for a long time. But competition for AI infrastructure has completely changed this landscape. Data centers, high-end GPU chips, power systems, and network facilities all mean huge capital expenses, long payback cycles, and rapid depreciation pressure.

Capital expenditure is expanding at a rate that far exceeds the growth in operating cash flow. The market expects that the total capital expenditure of the five companies Alphabet, Microsoft, Amazon, Meta, and Oracle will reach 769.2 billion US dollars in 2026, almost double that of 2025, and may further rise to about 1 trillion US dollars in 2027.

Financial data confirms this pressure: In the second quarter of 2026, Alphabet's operating activities generated cash flow of about US$39.1 billion, but capital expenditure reached about US$44.9 billion, and free cash flow fell to about US$5.9 billion — this is the first time since Alphabet went public that the quarterly free cash flow was negative.

Sloke believes that if this rebound fails to materialize, the consequences will not be limited to the balance sheets of a few tech companies. Investors' interest in broader AI transactions may fade, credit spreads on hyperscale corporate debt may widen, and companies may be forced to cut back on spending plans. Since these expenses have become an important contributor to US economic growth, once they fall back, they will eventually drag down GDP.

The report framed this assumption as the only point of agreement to sustain the credit story. It did not predict failure, but clearly lays out a downward path: weakening cash conversion will first impact valuation and financing costs, then affect investment, and eventually spread to the real economy.

S&P Global Ratings also assumes an inflection point in 2028 in its model — revenue growth accelerates and capital expenditure growth slows — but whether this assumption holds true depends on whether AI commercialization can beat the rate of capital expenditure expansion.

For credit investors, the meaning is that the bonds of hyperscale enterprises are not so much a traditional investment-level story as a concentrated bet on the timing and scale of AI monetization.