S&P 500 Shiller P/E Ratio Signals Rare Risk

The S&P 500 Shiller P/E ratio now exceeds 42, a valuation zone seen in only three periods since 1871. The prior two completed episodes preceded major losses, although the measure cannot tell investors when prices will turn or what might trigger the change.

This warning arrives while the market appears strong. From early June 2026, the Dow Jones Industrial Average, the S&P 500, and the Nasdaq Composite each moved to new records. Several developments reinforced that advance:

• Continued excitement surrounding artificial intelligence

• A record level of S&P 500 corporate repurchases during 2025

• Company earnings that came in ahead of expectations

• Strong demand for initial public offerings, encouraged in part by interest in SpaceX

Those forces help explain why prices climbed. They do not answer whether current prices remain reasonable relative to normalized corporate profits.

Record prices and extreme valuations now coexist

Risk appetite is visible beyond the indexes. Margin borrowing has risen at a rapid rate for the fourth time since 1999. Earlier bursts of that behavior were followed by trouble for equities, but margin debt is not the primary concern raised by the source analysis.

The more unusual signal comes from valuation.

Valuation is never completely objective. One investor may consider a company expensive while another believes its growth prospects justify the same price. Analysts still need consistent measures that allow current conditions to be compared with earlier market periods.

The ordinary price-to-earnings multiple is one such measure. It relates a company’s market price to its earnings per share from the most recent 12 months. Paying a lower multiple generally means paying less for each dollar of profit.

Its weakness appears when annual earnings become unusually volatile or fall below zero. Negative profits during a recession can make the standard multiple unusable precisely when economic conditions require additional context.

CAPE averages inflation-adjusted profits across a decade to reduce short-term earnings distortion.

CAPE smooths a decade of corporate earnings

The cyclically adjusted price-to-earnings measure addresses that limitation by averaging a decade of corporate profits after accounting for inflation. It is commonly called either the CAPE ratio or the Shiller P/E.

Although economists developed this framework in the 1980s, historical records support calculations extending to January 1871. Across the resulting 155-year series, the average reading is about 17.4.

The level recorded in July 2026 is far removed from that baseline. On July 10, the measure rose beyond 42. During the current episode, it has reached 42.84. That is the second-most expensive multiple documented during an ongoing bull market.

The significance lies in rarity rather than precision. A high CAPE reading does not prove that every company is overvalued, and it does not establish the date of a market reversal. It shows that broad equity prices stand at an exceptional level when measured against smoothed earnings.

Only two earlier episodes offer a completed outcome

Since the historical series began, CAPE has crossed the 40 threshold during three distinct periods:

• The first began in January 1999 and continued through September 2000. The measure established its record of 44.19 in December 1999.

• The second occurred briefly during the opening week of January 2022.

• The third started in May 2026 and remained active when the source was published in July.

Two completed examples cannot support a precise forecast. Their outcomes are still relevant because both were followed by severe index declines.

The technology bubble started to break three months after the December 1999 valuation peak. Across the decline that followed, the S&P 500 gave up 49%, while the Nasdaq Composite surrendered 78%.

The January 2022 episode also preceded a bear market. The S&P 500 fell by roughly one-quarter, and the Nasdaq Composite declined by approximately one-third.

Both periods demonstrate that a major technological shift can reshape the economy without protecting every price attached to it. The internet’s long-term importance did not prevent the dot-com correction. In the same way, the potential impact of artificial intelligence does not remove the risk created when expectations and valuations become unusually elevated.

The two completed moves beyond 40 preceded substantial losses for major U.S. indexes.

The signal warns about conditions, not a deadline

CAPE supplies context, not a market clock. It cannot identify the month a correction might begin, calculate the eventual loss, or reveal the event that could change sentiment. An expensive market may remain expensive while earnings, capital flows, buybacks, and investor confidence continue supporting demand.

That limitation matters when interpreting the current reading. The historical record above 40 supports concern about a future bear-market decline across the major U.S. indexes. It does not establish that such a decline must begin immediately.

This is where the signal stops working as a precise guide: using an extreme valuation to call an exact market top asks it to answer a timing question it was not designed to resolve.

Investor emotion can also shape the path of a downturn. Fiscal and monetary decisions influence financial conditions, but neither can permanently remove corrections from the market cycle. Periods of weakness remain part of equity-market history because expectations, positioning, and confidence change.

Market-cycle duration provides necessary context

Valuation history looks unfavorable at the current level. The duration of past market cycles adds a different perspective.

Bespoke Investment Group reviewed the calendar length of S&P 500 bull and bear phases beginning in September 1929, when the Great Depression started. Its figures reveal that advances have generally persisted much longer than contractions:

• A bear market averaged 286 calendar days, equal to approximately 9.5 months.

• During the 97-year period examined, none of the bear phases lasted beyond 630 calendar days.

• Bull markets averaged 1,023 calendar days, about 3.6 times the typical bearish period.

• Of the 27 bull markets in the data, 14 continued longer than the most prolonged bear market.

The advance that began on October 12, 2022, had become the ninth-longest S&P 500 bull market by May 30, 2026. It had also exceeded the 1,324-day run that concluded on February 9, 1966.

These duration figures do not make valuation risk disappear. A relatively brief decline can still produce substantial losses, as the earlier CAPE episodes demonstrate. The figures instead show that market peaks and troughs have not formed a symmetrical pattern. Historically, expansion has occupied more time than contraction.

The warning does not erase the longer record

The historical evidence supports two conclusions that may initially appear to conflict.

First, a CAPE level above 42 is extraordinary. Both previous moves through 40 were followed by bear markets, so the current valuation deserves attention. Prices at this level may leave less room for disappointing earnings, weaker economic conditions, or a shift in sentiment.

Second, downturns have generally been shorter than advances. Over long periods, growth in the U.S. economy and rising values among its most influential companies have allowed the major indexes to recover from declines and eventually establish higher records.

Perspective changes how these facts are understood, but it does not change the facts themselves. Treating the current reading as harmless would ignore two difficult historical precedents. Treating it as an exact forecast would ignore the indicator’s inability to identify timing or catalysts.

The S&P 500 Shiller P/E ratio therefore carries a clear but limited message. Valuations are in a range with almost no historical precedent, and earlier examples ended badly. The warning concerns vulnerability and reduced tolerance for disappointment. It does not provide certainty about when the market cycle will turn.