Stock bull market nears 4-year anniversary driven by AI spending — but there are looming risks

The S&P 500 bull market is nearing its four-year anniversary after gaining 117%, with earnings expected to rise more than 35% this year as AI-related capital spending supports profits and growth. The rally is increasingly concentrated: the top 10 S&P 500 companies now account for about 40% of the index, up from about 28% in October 2022, while Nvidia’s market value has climbed to $5.8 trillion from $286 billion. Risks include Fed rate hikes, a 10-year Treasury yield around 5.2%—its highest level in 24 years—and possible volatility around next month’s US midterm elections.

Analysis

The key vulnerability is a timing mismatch: chip and infrastructure suppliers can book demand as AI capacity is built, while the hyperscalers must demonstrate that usage and incremental revenue justify the full cost of deployment. If returns disappoint, capex cuts could transmit quickly to suppliers such as NVIDIA; weakness in Alphabet or Meta could instead reflect rising infrastructure costs before it shows up as lower demand for AI. These are different earnings channels, so treating all AI exposure as one trade obscures the risk.

Higher yields make that proof burden more consequential: they raise the hurdle rate for long-lived projects and offer investors a more credible alternative to expensive growth equities. That argues for monitoring revisions to capex plans and monetization—not the bull market’s age—as the key 1–3 month signal. A further yield rise or weaker guidance could pressure concentrated index exposure; an easing in yields, broadening earnings growth, or sustained monetization would undermine the cautionary case.

Contrarian point: concentration is not itself a sell signal when earnings leadership remains intact, and a broad market hedge could be costly if that leadership persists. The more attractive response is measured exposure reduction and a relative-value expression, rather than an outright short of AI leaders without valuation and positioning evidence. Over 6–18 months, the structural question is whether AI productivity gains spread beyond infrastructure beneficiaries enough to broaden profit growth.

Key Decisions for Investors

• Consider a modest, staged long RSP / short SPY relative-value position to reduce dependence on the largest index constituents without making a broad bearish call. Reassess if market breadth narrows further and mega-cap earnings revisions continue to lead; exit if the relative underperformance of equal-weight exposure accelerates rather than stabilizes.
• Avoid adding to NVIDIA solely on continued infrastructure spending. Track hyperscaler capex guidance alongside evidence of AI monetization; a capex slowdown or weaker returns on AI investment is the downside catalyst, while resilient spending and improving monetization would falsify the caution.
• For Alphabet and Meta, separate advertising or other core-business performance from AI-related infrastructure expense in upcoming results. Do not assume AI capex is either pure upside or a near-term earnings problem without updated guidance and expense detail.
• Treat Treasury yields and the Fed path as portfolio-level triggers: a renewed rise in yields alongside downward earnings revisions would favor additional equity-risk reduction; falling yields with broadening earnings growth would argue against extending the hedge.