Friday, August 28, 2026
DarkSubscribe
AI Infrastructure · News & Analysis
HomeCapital MarketsReport
Capital Markets · Report

24/7 Wall St. identifies Amazon and Oracle as hyperscalers facing the highest AI capital expenditure risks relative to cash flow generation.

If these two struggle to monetize massive infrastructure deployments, it could trigger sector-wide reassessments of hyperscaler ROI and slow downstream GPU/cloud procurement.
Trade pressSlicast · August 19, 2026 · US · Source: Google News
importance 70

Oracle (NYSE:ORCL) and Amazon (NASDAQ:AMZN) both closed their latest quarters with the same uncomfortable admission: artificial intelligence infrastructure is now consuming cash faster than either business can generate it. Oracle reported fourth-quarter results for fiscal year 2026 on June 10, 2026, followed by Amazon’s second-quarter fiscal 2026 report on July 30, 2026. When measured against peers Microsoft (NASDAQ:MSFT) and Alphabet (NASDAQ:GOOGL), both companies appear the most exposed to current spending pressures.

Oracle’s cloud infrastructure segment grew 93% year over year, while remaining performance obligations (RPO) ballooned to $638 billion, a 363% increase. The trade-off is steep: fiscal 2026 capital expenditures reached $55.66 billion against operating cash flow of just $31.98 billion, resulting in negative free cash flow of $23.69 billion. Chief Financial Officer Hilary Maxson advised analysts to expect a “net cash outlay for capital expenditures of around $70 billion” in fiscal 2027, to be financed by approximately $40 billion in debt and equity issuances. Non-current debt has already climbed to $124.7 billion. Investors have reacted accordingly: ORCL shares are down 26.88% since the earnings release.

At Amazon, AWS revenue grew 36.7% year over year to $42.2 billion—the fastest pace in 18 quarters—supported by a $496 billion backlog. However, second-quarter capital expenditures alone hit $54.21 billion, pushing first-half 2026 free cash flow into negative territory at $27 billion. CEO Andy Jassy framed the situation plainly: “we’ll spend a lot of capex and encounter free cash flow headwinds until these data centers come online.” Large-scale Trainium commitments from OpenAI (approximately 2 gigawatts starting in 2027) and Anthropic (up to 5 gigawatts) justify the buildout, but they also concentrate payback risk around two primary counterparties.

Microsoft still generated $67.99 billion of fiscal 2026 free cash flow despite recording $115.95 billion in capital expenditures. Alphabet’s second-quarter free cash flow slipped to negative $5.86 billion—an uncomfortable result that remains cushioned by its core search advertising business.

I will monitor whether Oracle’s BBB-rated balance sheet can absorb another debt cycle if its concentrated AI backlog slips even a couple of quarters. For Amazon, the central question is margin sustainability: retail operations cannot indefinitely subsidize an annual capital expenditure run rate of approximately $220 billion. All of this infrastructure requires substantial power and cooling, which is why we have compiled seven key suppliers driving the AI infrastructure buildout in a complimentary research report.

From an exposure-with-safety-net perspective, Amazon’s profile appears sturdier. AWS continues to deliver 39% operating margins, and its retail division provides reliable funding for the expansion. Oracle presents a more compelling case for turnaround investors comfortable with a 40.26% one-year drawdown and the possibility that RPO conversion materializes on schedule. A single clean quarter of positive free cash flow at Oracle would serve as a meaningful signal to track. Both companies remain ownable, but Oracle carries the thinnest margin for error among the hyperscalers.

Contact [email protected] for any questions or corrections.

Read the original
24/7 Wall St. identifies Amazon and Oracle as… · Slicast