According to The New York Times, artificial intelligence is simultaneously lifting equity markets and the broader economy while introducing material risks to both — a dynamic playing out in real time as the S&P 500 sits at 7,509 and the Shiller CAPE ratio stands at 41.0, a valuation level exceeded only during the peak of the dot-com bubble and briefly in 2021. The tension between genuine productivity gains from AI and concentration risk in a handful of mega-cap technology stocks is now a central concern for portfolio construction. Investors who ignore the risk side of this equation do so at measurable cost.
AI as Both Catalyst and Concentration Risk
The NYT's framing captures a structural dilemma that has no easy resolution. AI adoption is producing real economic output — higher corporate earnings, improved productivity metrics, and accelerating capital investment in compute infrastructure. These are genuine fundamental supports for equity valuations, not mere speculation. But the gains are concentrated. A small cluster of technology companies commands a disproportionate share of index weight, meaning passive investors in broad benchmarks carry far more single-sector exposure than the diversification label implies.
This concentration creates asymmetric drawdown risk. When a handful of stocks account for an outsize portion of index returns, a sentiment shift in that cluster — triggered by earnings misses, regulatory action, or a change in rate expectations — can produce index-level losses that feel far larger than the underlying economic fundamentals would justify. The NASDAQ 100, currently at 29,065 and down 0.31% on the day, illustrates this dynamic: it trades at a significant premium to the broader market, amplifying both upside and downside for investors who hold it as a core position. For portfolio managers focused on drawdown protection, this concentration is the primary variable to manage.
What a Shiller CAPE of 41 Means for Drawdown Probability
Valuation does not predict timing, but it does predict magnitude. At.0, the historical base rate for severe drawdowns over a five-to-seven year horizon increases materially. The ratio compares current price to the ten-year average of inflation-adjusted earnings, smoothing out cyclical noise. At current levels, it signals that the market is pricing in a long runway of strong earnings growth — precisely the narrative AI is currently supplying.
The problem is that AI-driven earnings optimism can shift faster than the underlying business fundamentals. If the productivity gains from AI prove slower to materialise in corporate earnings than the market currently expects, or if interest rates remain elevated — the US 10-year yield is at 4.60% and the Fed Funds Rate sits at 3.63% — the multiple compression from current levels could be severe. A reversion to even the 30x range would imply a price decline exceeding 25% from current levels, without any deterioration in underlying earnings. That is not a tail risk; it is a plausible central scenario if AI monetisation disappoints.
A Shiller CAPE of 41.0 has historically been associated with below-average real returns over the subsequent decade and above-average probability of a drawdown exceeding 30% at some point within five years.
VIX at 16.93 Signals Complacency — Not Safety
The VIX closed at 16.93, down 0.70% on the day. Low volatility readings are frequently misread as confirmation that risk has been reduced. In practice, a suppressed VIX reflects current market complacency rather than diminished future risk. Options are cheap because few investors are buying protection. This is precisely the environment in which tail hedges are most affordable and most necessary.
Investors at Bubblepedia have documented repeatedly how periods of suppressed volatility preceding large drawdowns share a common feature: broad consensus that the prevailing rally has a durable structural foundation. In the late 1990s it was the internet; today it is AI. That structural foundation may be real — but real structural change and sustainable equity valuations are not the same thing. Buying portfolio protection when the VIX is below 18 costs less than buying it after a shock pushes it above 30. The current window is an opportunity, not a signal to reduce hedges.
Gold at $4,129.85, up 1.31% on the day, and silver at $60.13, up 1.74%, are sending a separate signal. Both metals are rising alongside equities — a pattern that typically reflects either currency debasement concerns or a broader search for assets that hold value through volatility. Allocating a portion of portfolio risk budget to gold provides a genuine diversifier that does not rely on AI earnings growth to maintain its value. For more on this, see the Precious Metals Hub.
Background: Crowd Psychology and the AI Rally
Understanding the behavioural dimension of the current market is as important as the quantitative one. Behavioural Finance - Part 2: Market Manias, Crowd Psychology, and Investor Traps details how recency bias and FOMO systematically cause investors to extrapolate recent strong performance into the future, increasing position sizes at exactly the point when risk is highest. The AI rally has many of the hallmarks of a crowd-driven momentum trade layered on top of genuine fundamental change — a combination that historically produces the most damaging drawdowns because the unwind surprises investors who believed the structural case made valuations irrelevant.
This is not a call to exit equities. It is a call to assess whether current portfolio construction reflects the actual risk being taken, not the risk that felt appropriate twelve months ago when valuations were lower and the AI investment thesis was less widely owned. Portfolio rebalancing at current levels — trimming outperforming technology positions and reallocating to less correlated assets — is one of the most direct tools available. Combining rebalancing with a dollar cost averaging approach for redeployment reduces the risk of poorly timed exit decisions driven by short-term market noise.
Actionable Positioning for Drawdown Protection
Three practical steps follow from this analysis. First, review technology and AI-related concentration. If the NASDAQ 100 or a small group of mega-cap AI stocks accounts for more than 30% of total portfolio value, that is a concentration risk that valuation history suggests warrants reduction. Second, consider the cost of protection while the VIX remains below 17. Put options on broad indices or a modest allocation to gold provide asymmetric payoffs if the AI earnings narrative disappoints. Third, assess fixed income duration. With the US 10-year yield at 4.60%, short-to-medium duration bonds now offer real yield buffers that were absent two years ago, providing ballast without the full duration risk of long-dated Treasuries.
Investors who build drawdown protection now — while markets are orderly, volatility is cheap, and consensus is optimistic — will be better positioned than those who wait for the catalyst that makes protection both necessary and expensive. See the Retirement Planning hub for guidance on matching risk tolerance to time horizon in the current cycle.