Apollo Global Management’s chief economist Torsten Sløk reported on July 9, 2026, that an extended delay in artificial intelligence profit generation from major technology firms poses a significant risk to broader equity markets. Citing a high dependency on projected earnings from hyperscalers, Sløk’s analysis indicates that current valuations in key tech stocks embed a substantial premium for AI monetization that has yet to materialize. The assessment follows several quarters of record capital expenditures by major cloud providers with limited immediate return on investment. Sløk’s warning centers on the potential for a multi-trillion dollar market value adjustment if the anticipated revenue surge from AI services remains elusive beyond current investor timelines.
Context — why this matters now
The current market environment echoes the dot-com bubble’s infrastructure build-out phase, where massive capital investment preceded a wave of failures before eventual winners emerged. The S&P 500 Information Technology sector’s forward price-to-earnings ratio of 32.5 is 70% above the 10-year average, a disparity largely justified by AI growth narratives. This expansion has unfolded against a backdrop of elevated interest rates, with the 10-year Treasury yield holding at 4.8%, pressuring valuations for long-duration growth assets. The immediate catalyst for Sløk’s commentary is the confluence of decelerating core cloud revenue growth and sustained high capital expenditure guidance from hyperscalers for the second half of 2026, creating a precarious timeline for profitability.
Investor patience for unprofitable tech ventures is historically finite. The 2000-2002 Nasdaq correction erased approximately $5 trillion in market value as capital dried up for companies unable to convert eyeballs into earnings. Current AI infrastructure spending by Microsoft, Amazon, and Google-parent Alphabet exceeds $140 billion annually, a figure that has doubled since 2023. Unlike the dot-com era, today’s spenders are cash-rich industry incumbents, but their stock performance remains tethered to justifying these outlays with new revenue streams. The market’s tolerance for delayed returns is being tested as rate cuts remain elusive.
Data — what the numbers show
Hyperscaler capital expenditure reached a collective $152 billion over the last four reported quarters, representing a 95% year-over-year increase. The three major cloud providers—Microsoft Azure, Amazon Web Services, and Google Cloud—collectively generated $290 billion in revenue for the same period, with their growth rate decelerating to 18% from a peak of 35% in 2022. Nvidia’s data center revenue, a proxy for AI hardware demand, surged to $47.5 billion in its latest quarter, up 262% year-over-year.
| Metric | Q2 2025 | Q2 2026 | Change |
|---|
| Combined Hyperscaler Capex | $78 Bn | $152 Bn | +95% |
| Combined Cloud Revenue Growth | 22% | 18% | -4 pp |
| Nasdaq 100 P/E (Forward) | 28.5 | 32.5 | +14% |
This capex surge contrasts with the S&P 500’s aggregate capital expenditure growth of just 7% over the same period. The Nasdaq 100 index, heavily weighted toward these tech giants, has gained 24% year-to-date, significantly outperforming the S&P 500’s 11% gain. The valuation gap between AI-exposed megacaps and the rest of the market now sits at its widest level since the fourth quarter of 2021.
Analysis — what it means for markets / sectors / tickers
A delayed AI profit cycle would trigger immediate repricing in the hyperscaler complex, with secondary effects rippling through semiconductors, data center REITs, and software vendors. Direct exposure suggests Microsoft (MSFT), Amazon (AMZN), and Alphabet (GOOGL) could see 15-25% valuation corrections if growth assumptions are tempered, given their elevated multiples. Semiconductor leaders like Nvidia (NVDA) and Advanced Micro Devices (AMD), while suppliers, face a double risk of order push-outs and inventory adjustments, potentially impacting revenues by 10-15% in a downside scenario.
A counter-argument holds that AI infrastructure is a long-term necessity, not a discretionary cycle, and early spenders will capture durable competitive advantages that justify current investment. Historical precedent from cloud adoption suggests a multi-year investment lag before profitability inflected sharply upward. Positioning data shows hedge funds have increased short exposure to semiconductor equipment makers and certain software-as-a-service stocks while remaining net long the hyperscalers, a bet that end-market demand will consolidate around the largest platforms. Capital flow has recently rotated into energy and industrials, sectors with more tangible near-term earnings visibility.
Outlook — what to watch next
Key catalysts include Microsoft’s earnings report on July 24, 2026, which will provide updated Azure AI service revenue metrics and capex guidance for fiscal 2027. Amazon’s report on July 31 will detail AWS performance and any changes to its projected $60 billion annual capex budget. The Federal Open Market Committee’s decision on September 18 will influence the discount rate applied to future tech earnings.
Market technicians are watching the 17,500 level on the Nasdaq 100 as critical support, a 10% correction from recent highs that would breach its 200-day moving average. For Nvidia, the $115 per share level represents a key area of consolidation from earlier in 2026. A break below this level on heavy volume would signal weakening conviction in near-term AI hardware demand. Yield thresholds are also crucial; a sustained move in the 10-year Treasury above 5.0% would intensify pressure on long-duration tech assets irrespective of AI fundamentals.
Frequently Asked Questions
What does an AI profit delay mean for retail investors?
Retail investors concentrated in broad market index funds or technology ETFs like QQQ have significant, indirect exposure to hyperscaler profitability. A delay would pressure these funds’ performance. Direct stock owners should assess the proportion of a company’s valuation attributed to future AI earnings, a figure often detailed in analyst reports. Diversification into sectors with lower AI dependency, such as consumer staples or healthcare, can mitigate single-theme risk.
How does current AI spending compare to the cloud build-out?
The scale and speed of AI capital expenditure dwarf the initial cloud infrastructure wave. Hyperscaler capex as a percentage of revenue is now 52%, compared to a peak of 35% during the early 2010s cloud expansion. The key difference is that cloud spending served existing, known workloads, while a significant portion of current AI spending targets unproven, generative AI applications whose enterprise adoption curves and pricing power remain uncertain.
