“Just buy and hold an index fund” is often described as the safe, decision-free default for long-term investors — the calm alternative to gambling on active strategies. But holding a market-cap-weighted equity index isn't the absence of a bet. It's a specific, sizeable one: that the multiple the market is willing to pay for future earnings today will hold, or expand, indefinitely.

As of August 2026, the Shiller CAPE ratio for the S&P 500 sits in the low 40s — roughly two and a half times its long-run average of about 17. In a century and a half of data, a reading this high has only been recorded once before, in December 1999, shortly before the dot-com crash erased most of the technology sector's value. The only other comparably stretched period, 1929, sits meaningfully lower.

This is not a market-timing signal or a prediction of an imminent crash — the historical record is explicit on that point. The CAPE ratio first crossed its 1929 level in 1996; the market didn't peak until early 2000, four years and one of history's great bull runs later. What elevated valuations have reliably predicted, across a century and a half of data, is weaker average real returns over the following decade — not the timing of the next twelve months.

That distinction matters enormously for what “buy and hold” actually means in practice. An investor putting new capital into the index today isn't making a neutral, riskless choice — they are implicitly underwriting continued multiple expansion from an already elevated starting point, in a way an investor deploying capital at a CAPE of 17 was not. Both investors are technically “buying and holding.” They are not taking the same risk.

The strongest objection is that the CAPE ratio's own long-run average has been drifting upward for decades, partly because the modern index is composed of asset-light, high-margin technology and platform businesses that structurally command higher multiples than the industrial-era companies in Shiller's early data. That's a legitimate point, and it means “reversion to 17” is probably not the right bar. It doesn't remove the underlying finding, though: even accounting for a higher structural baseline, a reading this far above any post-war average has never once preceded a strong decade of forward real returns.

None of this is an argument for market timing, which has its own well-documented failure rate and its own long list of investors who sat out gains waiting for a correction that came late, or not at all. It's an argument for recognizing that inaction is itself a position, sized by whatever valuation happens to prevail on the day the capital is deployed — and that a decision framed as “no decision” deserves exactly the same scrutiny as any other.