Wall Street analysts warns the AI boom is on ‘borrowed time'
Key Points
- The cyclically adjusted price-to-earnings (CAPE) ratio has risen by more than 12 points since early 2023 and now exceeds 40, matching levels seen before the dot-com crash
- If the S&P 500 reaches around 8,000 by end of 2026, Capital Economics estimates a subsequent decline of at least 30% could become increasingly likely
- Market concentration in AI stocks now exceeds the technology sector's weighting during the late-1990s bubble peak, creating heightened vulnerability to sentiment reversals
AI Summary
Summary: Wall Street Warns AI Rally Nearing Its Peak
Capital Economics has issued a warning that the AI-driven market rally may be entering its final stage, despite expecting continued gains through 2026. The firm projects the S&P 500 will reach 8,250 by year-end 2026 but forecasts a significant retreat to approximately 6,500 by the end of 2027—representing a potential 21% decline.
Key Concerns
Valuation Metrics: The cyclically adjusted price-to-earnings (CAPE) ratio has surged by more than 12 points since early 2023, now exceeding 40—a level last seen before the dot-com bubble collapse. Most S&P 500 gains over the past three years have come from valuation expansion rather than fundamental improvements.
Market Concentration: A small group of AI and technology companies accounts for a disproportionate share of market value, with concentration levels now surpassing the tech sector's weighting during the late-1990s bubble peak.
Additional Risk Factors:
- Technology-related capital spending relative to economic output has reached elevated levels
- Stock market value to net worth of non-financial corporations approaches record highs
- Strong foreign inflows into U.S. equities increase vulnerability
Market Implications
Capital Economics expects tech-heavy markets in the U.S., Korea, and Taiwan to continue outperforming through 2026, driven by strong AI-related earnings momentum. However, the firm warns that weaker revenue growth or reassessment of AI spending assumptions could quickly undermine investor sentiment.
The analysts estimate that if the S&P 500 reaches around 8,000 by end-2026, a subsequent decline of at least 30% becomes increasingly likely, though this would still be less severe than the nearly 50% dot-com crash.
Model Analysis Breakdown
| Model | Sentiment | Confidence |
|---|---|---|
| GPT-5-mini | Bearish | 75% |
| Claude 4.5 Haiku | Bearish | 85% |
| Gemini 2.5 Flash | Bearish | 85% |
| Consensus | Bearish | 81% |