Record Valuations Just Sent a Rare Warning-History Says Watch Out

The warning comes from a crowded market, not one bad headline

The current warning is not about a single negative headline. It is about a market that has already priced in a lot of good news.

After the S&P 500 crashed by 20% in 2022, investors looked for fresh evidence that equities still offered compelling upside. AI provided that evidence. That helps explain the debate today: bulls see a durable productivity-driven rally, while bears see a crowded market where recent success can turn quickly into complacency.

The caution comes from valuation and positioning extremes that have historically shown up near major market turning points. The S&P 500 is near all-time-high valuations, and Ned Davis Research has warned that conditions look overbought, overowned, and overvalued. At the same time, current conditions still carry stretched positioning and a thin equity risk premium. That combination does not predict an immediate drop. It suggests the market has little room for disappointment.

Why narrow leadership can still become a liability

The main risk is not that the rally lacks a story. It is that too many investors are leaning on the same story.

AI leadership can work until expectations shift

A rally led by a narrow group of stocks can keep working for a long time. That is the trap. As long as the AI cohort keeps delivering, investors tend to reward concentration and treat narrow leadership as a feature rather than a risk.

That is why market polarization matters. J.P. Morgan describes the environment as split between AI and non-AI sectors. In that kind of setup, price can start to reflect one dominant narrative more than the full state of the market. When a trade is that popular, even a modest stumble can trigger a sharper repricing than the underlying business news alone would justify.

Breadth still matters for durability

This is the core expectation problem: when a theme is widely understood, ordinary good news may not be enough. Investors already know AI spending is real. The harder question is whether earnings can keep advancing quickly enough across the market to sustain today's expectations.

A useful counterpoint is that broader participation tends to make rallies more resilient. Even as the S&P 500 hit record highs in early June, U.S. Bank research noted that broader participation across sectors, company sizes and global stocks makes a rally more durable. Right now, that durability still looks incomplete.

What could strengthen the rally, and what could break it

The practical question is not whether AI matters. It is whether earnings and macro conditions can keep catching up to a market that is leaning heavily on an expensive leadership cluster.

Where bulls still have support

Bulls can point to real earnings momentum. J.P. Morgan expects AI investment continuing to drive market dynamics and support growth. If that investment continues to translate into profits, concentrated leadership can remain expensive for longer than skeptics expect.

Where the risk case still has substance

The more cautious case is not simply that AI is "overdone." It is that the same capex cycle supporting growth can also keep inflation sticky. U.S. Bank says energy and artificial intelligence-driven capital expenditures are adding pressure to already elevated core services inflation, while consumers are dealing with negative real wage growth, weak savings, and rising energy costs.

That matters because it changes the main vulnerability. If inflation stays sticky, the issue is no longer just earnings quality. It also becomes a policy issue, with sticky inflation likely to remain a prevailing theme and most DM central banks expected to either stay on hold or conclude their easing cycle.

What to watch next

• Earnings follow-through: whether AI-driven profits spread beyond the most obvious winners.
• Market breadth: whether more sectors, company sizes, and regions start contributing.
• Inflation and policy: whether capex-driven growth keeps supporting the case for higher-for-longer rates.

The key point is not that the AI story is broken. It is that a crowded market can punish even small gaps between expectation and delivery much more harshly than an uncrowded one would.