A sharp decline in the S&P 500’s forward price-to-earnings ratio is raising concerns about the durability of the stock market rally, with Ned Davis Research warning that a valuation signal historically associated with recessions and bear markets has emerged at a time of growing uncertainty over artificial intelligence spending.
The research firm said in a note on Wednesday that the index’s forward price-to-earnings ratio fell by 4.4 points in the 12 months through October, an unusually large decline that has occurred only six times since 1983. The ratio compares the S&P 500’s market value with analysts’ expectations for companies’ earnings over the next 12 months.
The multiple now stands at 18.8 times forward earnings, down from 23.2 times last year. That earlier level represented the index’s most expensive valuation on this measure since the pandemic-era stock market boom.
Register for the next Tekedia Mini-MBA.
Register for Tekedia AI in Business Masterclass.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Ed Clissold, chief US strategist at Ned Davis Research, said five of the six previous episodes occurred when the US economy was already in recession or during an “echo bear market”, a downturn that follows within a few years of an earlier recession.
The historical pattern comes with a buzz because it indicates that a steep contraction in market valuations can accompany deteriorating economic conditions, rather than simply represent a routine adjustment in investor sentiment. However, the signal does not establish that a recession is imminent. Its historical record is limited, and the circumstances surrounding each episode differ.
The only exception identified by Clissold occurred in 1998, when the US economy was expanding. He attributed the decline in the valuation multiple to the market fallout from the 1987 stock crash. The US economy subsequently entered a recession about two years after the signal appeared in 1988, according to his account.
This year’s episode is also unusual because it is the only instance in the historical sample in which the forward P/E ratio has fallen so sharply while the S&P 500 remained higher over the preceding 12 months.
That combination points to investors becoming less willing to pay elevated prices for expected corporate earnings even as the index maintains gains. In other words, the market’s headline performance has not fully captured the change in how investors value future profits.
AI Spending, Earnings Expectations Face a Tougher Test
The decline in the forward P/E ratio also carries implications for the artificial intelligence investment cycle, which has supported valuations across major technology companies and the broader US equity market.
Clissold said investors have begun pricing in the possibility that the AI boom may not continue at its previous pace, pushing down the S&P 500’s valuation multiple by an amount more commonly associated with recessions or bear markets. That adjustment could provide some protection if corporate earnings growth slows next year. When investors reduce the prices they are willing to pay for expected profits ahead of an earnings downturn, some of the potential disappointment may already be reflected in share prices.
However, the protection is limited. If a slowdown in AI investment contributes to weaker economic activity and a recession, lower valuations may not be enough to prevent substantial losses.
“The skepticism could cushion the blow if earnings growth slows next year, but history suggests that if an AI slowdown leads to a recession, the recent drop in the P/E may not prevent significant drawdowns,” Clissold said.
The distinction between an earnings slowdown and a broader economic contraction is important. A moderation in AI-related spending could hurt technology companies, semiconductor suppliers and data center operators without necessarily pushing the wider economy into recession. A more serious downturn would affect corporate investment, consumer demand, employment and earnings across multiple sectors.
The historical performance following previous valuation declines offers another reason for caution. Although the S&P 500 was higher over the subsequent 12 months in five of the six instances, its average median drawdown during those periods was approximately 18%. That means the index often recovered or ended the following year higher, but investors experienced substantial losses along the way. The pattern points to the possibility of sharp corrections even when the longer-term market direction remains positive.
A market can deliver positive annual returns while still imposing significant losses on portfolios during periods of volatility. The historical record therefore offers little reassurance to investors expecting the current rally to continue without substantial interruptions.
The pressure is emerging as technology shares face renewed scrutiny over the economics and practical execution of AI infrastructure investment. Hyperscalers have committed substantial capital to data centers, computing capacity and supporting infrastructure, but the returns on those investments remain a major question for equity investors.
The latest midterm election results have added another source of uncertainty. Investors sold some technology stocks on Wednesday as they assessed the possibility that Democratic lawmakers could slow or restrict data center construction. Such facilities are key to the expansion plans of major cloud providers and AI companies, making permitting, electricity supply, water availability and local opposition important considerations for future investment.
This creates a more complicated environment for technology valuations. Companies may continue to report strong demand for AI services while facing higher infrastructure costs, regulatory obstacles and uncertainty over how quickly spending will translate into sustainable earnings.
The S&P 500’s declining forward multiple suggests investors are already reassessing the price they are prepared to pay for that growth. But a lower valuation does not automatically mean stocks are cheap, particularly if earnings forecasts have yet to account fully for weaker demand or slower investment.
The more consequential test will be whether AI-related revenue and productivity gains can justify the scale of capital expenditure committed by the largest technology companies. If earnings continue to expand, the recent valuation adjustment could leave room for the market to stabilise. If investment slows sharply or projected returns disappoint, the combination of weaker earnings expectations and further multiple compression could deepen the correction.
Ned Davis Research’s historical signal is therefore less a prediction of an immediate recession than a warning about the risks embedded in the current market. The S&P 500 has retained its gains, but investors are paying less for each dollar of expected earnings. If the AI investment cycle weakens at the same time as economic growth deteriorates, the valuation reset may prove to be an early stage of a much broader market downturn.