In the spring of 1996, economist Robert Shiller walked into a Federal Reserve boardroom and delivered a presentation that made Alan Greenspan uncomfortable enough to coin the phrase “irrational exuberance” shortly thereafter. Shiller’s tool was a valuation metric he had developed with fellow economist John Campbell: the Cyclically-Adjusted Price-to-Earnings ratio, now universally known as CAPE.

Nearly thirty years later, the CAPE ratio remains the most rigorously validated long-horizon valuation metric in empirical finance. Understanding it is essential for anyone who wants to think clearly about multi-decade equity returns.

What CAPE Measures

The conventional price-to-earnings ratio divides a stock index’s current price by its most recent annual earnings. This ratio is volatile — earnings fluctuate substantially with the business cycle, making the raw P/E ratio a noisy and often misleading signal. At the bottom of a recession, when earnings collapse, the conventional P/E inflates to apparently extreme levels, falsely signaling that stocks are expensive precisely when they are often cheapest.

Shiller and Campbell’s solution was elegant: instead of using one year of earnings in the denominator, use ten years — specifically, the average of the prior decade’s real (inflation-adjusted) earnings. The result is a valuation ratio that smooths through business cycles and provides a far more stable signal of fundamental value.

The mathematical expression is straightforward:

CAPE = Current Price ÷ Average of Past 10 Years of Real Earnings

What makes CAPE remarkable is not its construction but its empirical track record. Drawing on data extending back to 1871, the metric shows a robust negative correlation with subsequent 10-year returns. When CAPE is low, forward returns have historically been high. When CAPE is elevated, forward returns have tended to be modest or negative in real terms.

The Historical Record

The predictive relationship is most clearly illustrated at market extremes.

At the depth of the Great Depression in June 1932, the CAPE ratio fell to 5.6 — among the lowest readings in the dataset’s 150-year history. An investor who purchased a broad index at that moment would have earned approximately 13.8% per year over the subsequent decade. The valuation signal was accurate: assets priced for catastrophe delivered exceptional returns.

At the opposite extreme, the peak of the technology bubble in December 1999 saw the CAPE reach 44.2 — a figure that had never previously been observed in the entire dataset. The implication of that reading was grimly straightforward: over the following decade, the S&P 500 would produce an annualized real return of approximately negative 1.4%. Not only did investors not make money in real terms over ten years; they actually lost purchasing power. Again, the signal proved accurate.

The pattern holds across other market turning points:

At the secular bear market trough of August 1982, the CAPE stood at 6.6. Subsequent 10-year annualized returns: approximately 17.5%. At the post-war peak in January 1966, the CAPE registered 24.1. Subsequent 10-year returns: approximately 3.6% per year — painful in a decade of high inflation.

The correlation is not perfect. No single metric can be. But across 150 years of data, CAPE has demonstrated greater predictive power for 10-year forward returns than any other commonly used valuation measure.

Why CAPE Works — and Why It Works Best Over Long Horizons

The theoretical underpinning for CAPE’s predictive power is straightforward. In the long run, stock prices cannot diverge infinitely from the earnings they represent claims upon. Mean reversion is not a mystical force — it is an economic inevitability. Companies that earn $X in profits can only sustain stock prices far above the present value of those earnings for a limited time before investors demand either higher returns to justify the risk, or prices must fall to restore fair value.

Over short horizons — one year, two years — this mean reversion mechanism operates weakly. Sentiment, momentum, and changes in risk appetite swamp fundamental valuation signals. This is why CAPE is a poor market-timing tool over short periods. It says nothing reliable about what the market will do next quarter or next year.

Over 10-year horizons, however, the fundamentals reassert themselves with considerable force. This is the domain where CAPE earns its keep.

A study by Shiller, Campbell, and various subsequent researchers consistently finds that CAPE explains somewhere between 40% and 60% of the variance in 10-year forward returns — an extraordinarily high figure for any single predictor in financial markets.

Interpreting CAPE: The Framework for Long-Term Investors

A practical framework for interpreting CAPE, calibrated to historical data:

CAPE below 10 has historically been associated with subsequent 10-year annualized returns in the range of 14–18%. These are “once in a generation” buying opportunities, last seen in the early 1980s and briefly during the worst of the 2009 financial crisis.

CAPE between 10 and 20 represents the historical average valuation range. Expected 10-year returns in this zone have clustered around 9–13% annually — consistent with the long-run average.

CAPE between 20 and 30 has typically produced more modest subsequent returns: 5–9% per year on average. Not poor, but below historical norms.

CAPE above 30 has historically preceded some of the most challenging periods for long-term equity investors. The 10-year returns from CAPE levels above 30 have often been in the 3–6% nominal range — and have been negative in real terms in several historical instances.

The Limitations

Intellectual honesty requires acknowledging CAPE’s limitations. Critics, most prominently practitioners at investment banks, have argued that accounting changes since the 1990s have structurally lowered reported earnings, inflating the CAPE ratio and making today’s readings appear more extreme than they truly are.

Others note that the relationship between interest rates and equity valuations means that the “fair” level of CAPE is not fixed across history — when bond yields are very low, higher equity valuations may be justified.

Both arguments have merit. Neither invalidates the metric entirely. The reasonable conclusion is that CAPE is best used as a directional guide — indicating the broad range of likely outcomes rather than a precise point forecast. An investor who uses CAPE to calibrate expectations (rather than to time market entry and exit) is using the tool appropriately.

The Current Moment

As of early 2024, the Shiller CAPE ratio stands in the range of 32–34 — a level that places the U.S. equity market in the top 10% of historical valuations by this measure.

This does not mean a crash is imminent. It does not mean equities should be avoided. What it means, if history is a reliable guide, is that investors buying U.S. equities at current prices should calibrate their expectations for the next decade accordingly: nominal returns in the 5–8% range are plausible; the exceptional 15%+ annual returns of previous decades from lower valuations are unlikely.

For long-horizon investors, this is useful information — not for timing the market, but for setting realistic expectations and ensuring that financial plans are not predicated on a continuation of the exceptional post-2009 return environment.

The CAPE ratio does not eliminate uncertainty. Nothing does. But for investors willing to think in decades, it remains the most rigorous empirical signal available about the range of probable long-run outcomes.

CAPE data sourced from Robert Shiller’s publicly available dataset at Yale University (www.econ.yale.edu/~shiller). Subsequent return calculations based on S&P 500 total return index. Past performance is not indicative of future results.