Whether stocks are expensive depends on how you look at things.
With a weighted forward price-to-earnings multiple of 22.4, the S&P 500 SPX is trading at an 11% premium to its five-year average of 20.1 and a 19% premium to its 10-year average of 18.8.
But, since the end of September, the index's forward P/E ratio has actually decreased from 22.9. And more than half of stocks in the S&P 500 now have lower P/E ratios than they did as of Sept. 30.
For the index, the forward P/E is based on current stock prices and rolling 12-month earnings-per-share estimates for its constituent stocks among analysts polled by LSEG, weighted by market capitalization.
Some analysts and money managers believe the recent declines in P/E valuations are highlighting opportunities to buy into quality businesses at more attractive prices, and LSEG's data shows that some of the stocks with particularly robust projected upside lie within the Big Tech universe.
Oracle Corp. $(ORCL)$ has become the poster child for AI-debt fears, and its P/E has fallen to 23.9 from 39.2 at the end of September. That contraction has been driven mainly by the "price" part of the P/E equation: Oracle shares are down 34% since Sept. 30, while its rolling 12-month earnings-per-share figure has increased by 55 cents.
Microsoft Corp. $(MSFT)$ has seen a less-dramatic contraction, with its forward P/E declining to 27 from 31.9 as of the end of September. That has resulted from an 8.3% decline in its share prices as its rolling forward EPS estimate has increased by 9%.
The consensus price target for Oracle shares among analysts polled by LSEG is 60% above Monday's closing price, while the consensus price target for Microsoft's stock implies 31% upside over the next 12 months.
Screening for a combination of declining P/E and increasing EPS estimates
The trend isn't limited to AI-focused stocks. We screened the entire S&P 500 and found that 290 of the stocks were showing lower forward P/E valuations at Monday's close than they did as of Sept. 30. And 249 of those companies had seen their rolling 12-month EPS estimates increase over the same period.
Here are the 20 largest companies by market capitalization that passed the screen, with forward P/E valuations declining and rolling 12-month EPS estimates increasing since Sept. 30. This first table shows the changes in forward P/E ratios and consensus 12-month EPS estimates. You might need to scroll the table or flip your device to landscape to see all of the data.
Nvidia (NVDA) tops the list with the largest market capitalization and a remarkable decline in its forward P/E ratio. In fact, one can call Nvidia a cheap stock, based on its forward P/E valuation relative to the S&P 500 and its projected growth rates for revenue and earnings over the next two years.
Here's the list with projected compound annual growth rates (CAGR) for revenue and earnings from calendar 2025 through calendar 2027, with those for the S&P 500 at the bottom. This table also includes price changes (excluding dividends) from Sept. 30 through Monday.
Considering how much higher the projected sales and EPS CAGR are for Nvidia than they are for the S&P 500, Nvidia’s stock now seems to be a bargain.
Here is a summary of analysts’ opinions about the 20 stocks:
Amazon and Broadcom are tied for the highest percentages of “buy” or equivalent ratings. Oracle’s stock has the most aggressive consensus price target relative to Monday’s closing price.
Oracle has emerged as a barometer for risk in the AI trade because of its high debt levels and dependence on OpenAI. While Oracle’s stock has been hit hard due to concerns about high AI spending, many Wall Street analysts believe these risks are overblown. Instead, they are optimistic that the company will be able to grow revenue and improve margins from its infrastructure business to make all the spending worthwhile.
Oracle is “paying the price for the abnormal speed in which investment is required to meet current AI demand trends,” Bank of America analyst Brad Sills wrote in a note last week. He attributed Oracle’s quarterly revenue miss to a “timing mismatch of build-out spend to revenue conversion” instead of a deterioration in fundamentals.
Jefferies analyst Brent Thill and Guggenheim analyst John DiFucci also highlighted in notes last week that Oracle’s data-center capacity is fungible. This means that even if OpenAI failed to pay for its contracted services, Oracle could reallocate its data-center capacity to other customers. Additionally, management shared on the earnings call that funding requirements will be below $100 billion, whereas some investors had been fearing a steeper amount.
Oracle’s risk/reward profile is skewed positively, according to Thill, whose analysis suggests 15% downside in a bear case but over 100% upside if Oracle delivers on Jefferies’ fiscal 2027 target of $8.39 in earnings per share.
Broadcom’s stock also fell after it reported earnings last week and announced a mystery fifth customer that had ordered $1 billion of chips for delivery toward the end of 2026. Bank of America’s Vivek Arya speculated in a note last week that the customer could be Apple.
Investors are concerned that Broadcom’s margins are compressing because of increased costs from the company’s fast-growing custom AI-chip business, which produces Google’s tensor processing units. It is a fair concern, according to Arya, but he believes Broadcom will be able to maintain its overall operating expenses. Meanwhile, the top-line trajectory looks optimistic, and Arya sees a path to $50 billion of AI revenue in fiscal 2026, up from $37 billion for the fiscal year that ended this October.
Unlike Oracle and Broadcom, Amazon’s stock hasn’t experienced much price action in recent months. The stock is up only 1.4% this year, as investors have perceived the company as an AI laggard because of slower-than-expected growth in Amazon Web Services.
However, TD Cowen analyst John Blackledge highlighted Amazon as his top internet pick in a note last week, pointing to the company’s moves to increase data-center capacity and the signing of a $38 billion deal with OpenAI. He said AWS can reaccelerate and reach a 23% annual growth rate from 2025 to 2030.
Additionally, Amazon’s “underappreciated” advertising business could comprise up to 35% of the company’s overall operating income for 2025, thanks to Prime Video’s massive reach, Blackledge wrote. These factors could lead earnings per share to grow at a 22% annualized rate from 2025 to 2030, he believes.
Comments