Shiller CAPE Ratio Explained: How Cyclically Adjusted P/E Works

The Shiller CAPE ratio, or cyclically adjusted price-to-earnings ratio, divides a company’s or market’s current price by its average inflation-adjusted earnings over the preceding 10 years. Rather than comparing today’s price to this quarter’s earnings, CAPE stretches the lens wide, smoothing out earnings booms and busts to expose whether the market is historically cheap or rich. Robert Shiller won a Nobel Prize partly for developing this framework.

Why Shiller Built CAPE

Robert Shiller, a Yale economist, noticed that the traditional price-to-earnings ratio swings wildly depending on where you are in the business cycle. During a recession, earnings collapse and P/E balloons even though stocks may be cheap. During a boom, earnings soar, P/E compresses, and the market looks “cheap” even at bubble prices.

A one-year P/E is like measuring a person’s health by their weight on a single bad day. If they’ve just recovered from the flu, they look emaciated; if they’ve eaten a huge meal, they look bloated. Shiller wanted a healthier measure.

He proposed averaging earnings over 10 years, then adjusting that average for inflation to remove the distortion of nominal growth and deflation. The result: a ratio that captures whether the market is priced expensively or cheaply relative to its true earning power, smoothed across the entire business cycle.

How CAPE Is Calculated

The numerator is straightforward: the current price (usually of an index like the S&P 500 Index).

The denominator is the average of the previous 10 years of real (inflation-adjusted) earnings. This is typically earnings per share, adjusted for CPI or another inflation index to put all years on a comparable basis.

For a company that earned $1 (real) in 2014, $2 (real) in 2015, and so on through $5 (real) in 2024, the 10-year real average is the sum divided by 10. If the stock trades at $150 today, CAPE is $150 divided by that 10-year average.

For the S&P 500, the same logic applies: take the price of an SPX-tracking fund or the index value, divide by the 10-year real average of index earnings, and you have the market’s CAPE.

What CAPE Reveals

A CAPE of 15 or below historically preceded 10-year returns that averaged 8–10% annually. These were periods when investors were pessimistic and prices were depressed—the 1980s, the 2008–2012 period, and a few other pockets.

A CAPE above 25 preceded periods of subpar returns. The late 1990s, 2007, and early 2022 all saw CAPE in that range, and subsequent 10-year returns were meager or negative.

Critically, CAPE does not predict the next quarter or year. It captures a longer regime: if the market is wildly overvalued by historical standards, the next 7–10 years are likely to be disappointing. If it’s deeply undervalued, they’re likely to be generous. This is useful for asset allocators and long-term value investors, not for traders.

The 10-Year Earnings Window

Why 10 years, not 5 or 20? Shiller’s research showed that 10 years captures most of a typical business cycle (usually 7–11 years from recession to recession in large developed economies). A shorter window would still be too noisy; a longer window would exclude recent-enough data to be relevant.

This means CAPE is slow. It takes a full decade for a valuation extreme to “age out” of the calculation. If the 2007 peak (CAPE near 26) was the last data point added, and it finally drops below the threshold in 2017, that’s only after 10 years—well after the bubble had burst and most of the damage was done.

Limitations of CAPE

Rate environment blindness: CAPE doesn’t account for interest rates. When yields on Treasury bonds are 5%, a P/E of 20 may be reasonable because the discount rate investors apply to future earnings has fallen. When Treasuries yield 1%, the same P/E looks cheap. Shiller’s framework was born in a high-rate era (the 1980s) and can misprice risk during low-rate regimes.

Structural growth shifts: If corporate profit margins expand structurally due to technology or globalization, historical CAPE thresholds are less predictive. The 2010s saw persistently higher margins; CAPE looked expensive by historical standards but worked out fine for buy-and-hold investors because the underlying earnings power had truly improved.

Sector composition: The S&P 500 of 2024 is dominated by software and artificial intelligence; the S&P 500 of 1974 was utilities and industrials. A CAPE comparison ignores that the cost of capital and earning stability differ across these regimes.

Mean reversion not guaranteed: CAPE assumes that valuation reverts to the long-term average. If structural conditions change—emerging-market growth accelerates, AI productivity compounds—the “average” itself may shift higher and stay there. CAPE would flag overvaluation for years while the market continued higher.

CAPE and Dividend Yield

A complementary approach is to look at the earnings yield—the inverse of P/E—and compare it to the risk-free rate. When earnings yield is well below Treasury yield, stocks are dear. When it’s well above, stocks are cheap. This is rate-sensitive in a way CAPE is not, and it’s useful to watch both.

Practical Use

Asset allocators use CAPE to set strategic allocations to equities. When CAPE is at historical lows, they may be willing to hold a higher equity weight for the next decade. When it’s at historical highs, they may de-risk. Pension funds and endowments that care about 10-year outcomes find this particularly useful.

Stock pickers sometimes use CAPE to filter: it’s more forgivable to buy a single stock at a 25 P/E if the market’s CAPE is 15 (suggesting broad undervaluation) than if it’s 30 (broad overvaluation). The tide lifts all boats; swimming against the tide is harder.

Retirees and near-retirees use CAPE to sense the investment climate. A CAPE below 18 suggests the market is not overheated, and a conservative withdrawal rate from equities is justified. A CAPE above 25 suggests a period of caution, and higher cash holdings make sense.

See also