Global pension funds are cutting exposure to US equities because a handful of artificial intelligence stocks now dominate index returns, according to the Financial Times. The move comes as the S&P 500 trades at 7780.02, up 0.74% on the day, and the Shiller CAPE ratio sits at 41.38, a level rarely seen outside the dot-com era.
Pension Giants Retreat From Concentrated US Equity Bets
The Financial Times reports that large pension funds are reassessing how much capital they hold in US equity markets, specifically citing concentration risk tied to artificial intelligence stocks. Index funds built around the S&P 500 and Nasdaq 100 have become increasingly dependent on a small group of technology names, meaning a downturn in AI-linked earnings could drag down diversified-looking portfolios far more than their labels suggest.
This is a structural concern for institutions managing liabilities measured in decades. Pension funds answer to retirees, not quarterly benchmarks. When a market-cap-weighted index becomes a concentrated bet on one theme, the fund's actual risk profile diverges from its stated mandate. The Shiller CAPE reading of 41.38 reinforces the caution: valuations at this level have historically preceded extended periods of below-average forward returns, even when short-term momentum stays positive, as it has with the Nasdaq 100 up 0.82% and the Dow up 0.28% today.
The AI Boom's Rate Problem
The pension retreat lands alongside a separate concern flagged by the New York Times: high interest rates are not slowing capital spending on artificial intelligence infrastructure. That is unusual. Rate-sensitive sectors typically pull back when borrowing costs rise. With the Fed Funds Rate at 3.75% and the US 10-year yield at 5.24%, AI-related capital expenditure has kept expanding regardless, according to the report.
For the Federal Reserve, this complicates the usual transmission mechanism between policy rates and corporate investment. If a narrow set of companies can fund expansion through cash reserves or captive financing rather than bank credit, rate policy loses some of its bite on that segment of the market, even as it continues to weigh on smaller firms and consumers. That divergence helps explain why equity benchmarks keep climbing, with the Russell 2000 up 0.67% and the VIX still low at 15.62, despite a policy rate environment that would historically be expected to cool risk appetite.
The Shiller CAPE at 41.38 represents a level exceeded only during the dot-com peak of 1999-2000, a period also marked by capital spending that kept growing well after valuations had detached from near-term earnings.
Diversification Strategy in a Top-Heavy Market
The practical question for portfolio construction is straightforward: how much AI-linked concentration does an investor actually want, and is it arriving by design or by default through passive index exposure? A standard S&P 500 tracker today carries a weighting toward a small number of mega-cap technology firms far above historical norms. Investors who believe they hold a diversified US equity allocation may instead hold a leveraged bet on one sector's earnings trajectory continuing uninterrupted.
Rebalancing is the most direct tool available. Trimming winners back toward target weights, even mechanically, forces discipline that pure buy-and-hold does not provide. Readers can review the mechanics in our guide to Portfolio Rebalancing. Geographic diversification offers a second lever. The Nikkei 225 is up 2.4% today, the Shanghai Composite has gained 0.31%, and the FTSE 100 sits at 10497.94, up 0.34%. None of these markets carry the same AI concentration as the S&P 500, and each responds to a different domestic growth and rate cycle.
History offers a sobering parallel for investors tempted to treat current valuations as the new normal. For broader context on how market cycles have handled extreme valuations before, see Bubblepedia's analysis of historical market bubbles, which traces how concentration risk has played out across past cycles. The pattern is consistent: narrow leadership eventually broadens or breaks, and portfolios built around a handful of winners carry outsized drawdown risk when sentiment shifts.
Background: Building Resilience Through Systematic Investing
One response to concentration risk and valuation uncertainty is to remove timing decisions from the equation entirely. Dollar Cost Averaging involves investing fixed amounts on a regular schedule regardless of where valuations sit, which smooths the average purchase price over a full cycle rather than betting on a single entry point. It will not protect against sustained overvaluation, but it reduces the behavioural risk of chasing momentum near a peak, a pattern discussed in our guide to Behavioural Finance.
Pension funds operate on longer horizons than retail investors, but the underlying logic of their retreat from concentrated US equity exposure applies broadly. Position sizing, rebalancing frequency, and geographic spread are the controllable variables when valuation metrics like the Shiller CAPE sit well above historical medians. Our guide to Risk Management covers how to size these decisions without abandoning growth exposure altogether.
Investors reviewing year-end allocations should treat the pension fund shift as a signal worth weighing, not a directive to exit US equities outright. The sensible response is checking actual sector and single-stock weightings inside index holdings, then deciding deliberately how much AI concentration the portfolio should carry going into 2026.