Where Will the S&P 500 Be in 10 Years? Nobel Laureate Robert Shiller Weighs In

Robert Shiller's CAPE-based model warns the S&P 500 could deliver just 1.3% total annual returns over the next decade, or -0.7% annually excluding dividends, implying an index level of 6,381 by end-2035. The article argues today's market is highly concentrated in AI stocks and resembles prior bubble-like periods, while highlighting better relative value in Europe and Japan, where Shiller forecasts 7.8% and 6.2% annual total returns, respectively. It is a valuation and long-term allocation piece rather than a near-term catalyst for individual stocks.

Analysis

The useful signal here is not the broad “valuation is high” message; it is that return dispersion is likely to widen materially as passive-cap-weighted exposure becomes more hostage to a small cluster of AI leaders. That creates a second-order setup where index-level returns can stagnate even if earnings growth remains healthy, because multiple compression in the mega-cap cohort would mathematically overwhelm the rest of the market. In that environment, factor dispersion—not directionality—becomes the main edge.

The best risk-adjusted opportunity is likely outside the obvious U.S. AI complex: cheaper international equities and neglected domestic value/quality names with balance-sheet support and cash-return capacity. If the next decade looks like a lower-beta, lower-multiple regime, businesses with durable free cash flow and visible buybacks should outperform because their return profile is less dependent on terminal multiple expansion. That favors companies where capital returns can offset muted top-line growth.

On the losers’ side, the biggest vulnerability is to AI infrastructure winners whose current valuations embed multiple years of flawless execution. Any slowdown in hyperscaler capex, product-cycle disappointment, or supply normalization could trigger an abrupt de-rating, and the longer-dated risk is similar to the late-1990s pattern where a few winners stayed dominant for years before the market realized the valuation was doing too much work. The point is not that AI is a bad theme; it is that the payoff distribution is skewed and crowded, making forward returns fragile from here.

The contrarian read is that investors may be over-anchored to the index and under-allocated to non-U.S. markets where earnings are improving but sentiment remains discounted. If the dollar weakens or U.S. growth broadens out, the relative-return gap could close faster than valuation models imply. That creates an asymmetric setup for diversified value and international exposure versus concentrated exposure to the same U.S. mega-cap beneficiaries everyone already owns.

Key Decisions for Investors

• Overweight quality value and capital-return names versus cap-weighted U.S. index exposure over the next 12 months; prefer businesses with net cash, >5% FCF yield, and active buybacks because they should outperform in a low-multiple regime.
• Initiate a pair trade: long EFA or VGK / short QQQ or SPY, sized for 6-12 months, to express the view that non-U.S. equities can rerate while U.S. mega-cap multiple risk compresses index returns.
• Trim or hedge crowded AI leadership exposure via NVDA downside puts or call spreads into strength over the next 1-3 quarters; the payoff is asymmetric if capex growth or guidance normalizes even modestly.
• Add selectively to INTC only as a funded turnaround/value basket position, not a standalone secular growth bet; treat it as a long-duration execution trade with high dispersion and binary operating risk.
• Use pullbacks to accumulate dividend growers and buyback compounders instead of broad market beta; the risk/reward is superior if the next decade delivers flat index multiples rather than broad multiple expansion.