The S&P 500's Shiller P/E reached 41.67 on Oct. 5, near its dot-com-era record. Past peaks came before steep losses, while long holding periods have historically rewarded investors who stayed invested.
Spending on artificial-intelligence infrastructure has helped support the rally, alongside corporate earnings that came in stronger than expected. The Dow Jones Industrial Average and S&P 500 have recently set records. The Nasdaq Composite has too.
On Oct. 5, the Nasdaq closed at a record, while the S&P 500 sat about 0.3% below its Aug. 13 closing record, according to Reuters' Oct. 5 market report. Momentum and valuation are separate questions. The second now deserves a hard look: the S&P 500's Shiller P/E is nearing a level reached only once before, in December 1999.
Investors are paying unusually high prices relative to a decade of inflation-adjusted earnings. That does not set a date for a decline.
On Oct. 7, the U.S. Treasury sold $39 billion in 10-year notes at a 5.3% yield, below the level expected by the market, indicating sufficient demand despite volatility.
A rare valuation reading
The Shiller P/E, also called the Cyclically Adjusted P/E Ratio or CAPE, compares stock prices with average inflation-adjusted earnings over the previous 10 years. That longer window can be more useful than a standard price-to-earnings ratio when a recession pushes recent earnings sharply lower. It cannot tell investors when prices will turn.
Researchers have backtested the measure to January 1871, giving them about 156 years of monthly data. Its historical average is 17.42. At the Oct. 5 close cited in the source material, the S&P 500's reading was 41.67, not far below the December 1999 record of 44.19.
Since early June, the Dow has gained 0.83. The S&P 500 is up 0.59%, and the Nasdaq Composite has added 0.64%. Each has reached a new high.
High valuations do not operate alone. Rising margin debt, persistent inflation and higher long-duration bond yields could weigh on stocks. On Oct. 8, the 10-year Treasury yield reached 5.305%, while the 30-year yield hit 5.666%. Reuters reporting carried by Fidelity said the 10-year yield was near a 24-year high amid rising oil prices and inflation concerns.
On Oct. 7, the S&P 500 closed below its record. Nine stocks were at 52-week highs and 12 at lows. The Nasdaq recorded 32 highs and 250 lows, according to Reuters' market-breadth report. Treasury yields had also put pressure on a record-setting rally, as an earlier market report described.
U.S. bond funds attracted a record $19.78 billion in net inflows in the week ending Oct. 9, including $6.76 billion for short- and intermediate-term government bond and Treasury funds.
History warns, but cannot time
Past CAPE readings above 30 came before sharply different market declines. The first move above 30 came in August and September 1929, before the Great Depression. The Dow eventually fell 89% from peak to trough.
After the CAPE topped 30, the S&P 500 lost about 20% in the fourth quarter of 2018. In February and March 2020, the COVID-19 crash erased 34% of the index's value in 33 calendar days.
The first move above 40 came during the internet boom. After the CAPE reached 44.19 in December 1999, the dot-com bubble burst. By October 2002, the S&P 500 had fallen 49%, and the Nasdaq Composite was down 78%. These episodes make a high CAPE a warning about risk over time. They do not show that the next decline will match an earlier crash or arrive soon.
Valuation is a poor short-term clock. The ratio offers no reliable signal for the catalyst or timing of a selloff. The source data does not establish that today's market will follow the path of 1929. It does not establish that the market will repeat 2000, either, or that 2020 is a guide. Treating a historical comparison as a forecast would ask more of the indicator than it can support.
Time changes the picture
Longer holding periods show a more reassuring record. Bespoke Investment Group reviewed 27 S&P 500 bull and bear market periods since September 1929. The average bear market reached its bottom in 286 calendar days, or about 9.5 months. None lasted more than 630 days.
The average bull market lasted 1,023 days, around 3.6 times as long. Fourteen of the 27 bull markets outlasted the longest bear market in that data set.
Crestmont Research examined 107 rolling 20-year periods from 1900-1919 through 2006-2025, counting dividends. Every period had a positive average annual total return. Because the S&P 500 began in 1923, the analysis used the performance of its components in other major indexes for 1900-1923. That record describes those historical windows. It does not promise gains in every future 20-year investment.
The practical distinction is between a risk signal and an investment timetable. A high CAPE argues for realistic expectations and attention to how much of a portfolio loss a household can absorb, not an automatic decision to sell. For investors with long horizons, historical evidence supports resisting panic. The possibility of a sharp downturn remains a reason not to treat recent records as proof that stocks are safe.
CAPE smooths earnings across a full decade, so it answers a different question from a short-term trading indicator: how expensive the broad market looks against a long earnings record. That makes it useful for judging starting conditions, but less useful for predicting next month's direction. A market can remain expensive for an extended period, and a low reading does not prevent further losses.
