Wall Street's AI Divide: Inside the Bull and Bear Cases for What Comes Next for the Stock Market

Dow Jones04:55

The capex trade will eventually end, likely in a destructive way, but economic tailwinds will prevail for now: BCA Research

Is the AI trade the greatest earnings story in decades or the most concentrated bubble since 2000?

Three years into the artificial-intelligence trade, the U.S. market has delivered some of the best returns, and the most extreme concentration, in decades. Yet Wall Street is no closer to agreeing about whether it's a bubble, how much longer it can keep going or what might ultimately end it.

That divide was on full display in a recent internal debate among BCA Research's top macroeconomic experts. A closer look at the discussion shows just how far apart different views remain, and why it's a debate every investor is taking part in whether they realize it or not.

Peter Berezin, chief economist at BCA Research, and Arthur Budaghyan, head of core macroeconomics and chief emerging-markets strategist, said their bearish view rests on the idea that the U.S. stock market is in the midst of an earnings bubble.

"Although profits are certainly quite strong, margins are less durable than they appear, and valuations have already priced in a best-case outcome with little room left for surprise," Berezin and Budaghyan said during a bulls vs. bears debate webcast last week.

Stock valuations look stretched by almost every measure, they said. Adjusted back to 2019 profit margins, the S&P 500 SPX would trade at 27 times forward earnings today, above the peak of roughly 26.5 times reached in March 2000.

Stripping out technology XX:SP500.45 and financial stocks XX:SP500.40, the rest of the stock market is not cheap either, trading at 26 times trailing earnings with only about 3% profit growth over the past two to three years, according to data compiled by BCA Research (see charts below).

SOURCE: BCA RESEARCH, FACTSET

In terms of the massive capital expenditures in AI - with Big Tech hyperscalers collectively pouring hundreds of billions of dollars into data centers, specialized chips, power grids and energy upgrades - Berezin and Budaghyan said the huge amount of capex is generating earnings but eroding cash flow.

"Every chip sold registers as revenue and profit for the seller, and the hyperscalers buying those chips treat the purchase as capex rather than an operating cost, so reported earnings rise without a matching increase in cash flow," they said.

"I think AI will end up being a lot like electricity. It's going to make companies more efficient, but it's not necessarily going to make them more profitable if everyone has access to it," Berezin said.

Meanwhile, the AI bulls include BCA's Juan Correa, head of portfolio construction, and Noah Weisberger, head of equities, who think the compute market remains "scarce rather than oversupplied." They say the return on capex is being underestimated rather than overstated, and the earnings strength is broader than the handful of companies usually associated with it.

According to BCA Research, hyperscaler cloud backlogs - orders received but not yet fulfilled - have grown by roughly $750 billion over just the past two quarters. That's evidence of demand still running ahead of what the industry can deliver.

Despite market concerns about lofty stock valuations, Correa and Weisberger argue that hyperscalers are actually trading at their cheapest valuations in a decade, even as returns continue to improve (see chart below).

SOURCE: BCA RESEARCH, FACTSET

Rental rates for the newest graphics processing unit generations have risen throughout the year, cloud contract values have roughly doubled on a per-megawatt basis from a year ago, earnings before interest and taxes per employee have climbed sharply and profits outside the core cloud business, such as Meta's $(META)$ AI-driven advertising gains and the resurgence of Google Search $(GOOGL)$ $(GOOG)$, are also growing rapidly.

"The expectation is that the market gets surprised by how much money this capex is actually generating within the next two quarters," they said.

The debate is not about whether AI is real or whether it's reshaping the U.S. economy by fundamentally altering corporate productivity, capital allocation and the labor market. Both sides take that as given. The argument is about timing: how much longer AI spending can outrun the doubts, and how violently the market will react when the reckoning finally arrives.

"The capex trade will eventually end, likely in a destructive way," said Correa and Weisberger from the bull camp. "But for now, economic and fundamental tailwinds will prevail. Multiple compression, not recession or an earnings collapse, is the bulls' chief worry."

U.S. stocks finished mostly lower on Monday. The S&P 500 was off 0.2%, while the Dow Jones Industrial Average DJIA slid 0.6% and the Nasdaq composite COMP was flat, according to FactSet data.

-Isabel Wang

This content was created by MarketWatch, which is operated by Dow Jones & Co. MarketWatch is published independently from Dow Jones Newswires and The Wall Street Journal.

 

(END) Dow Jones Newswires

July 20, 2026 16:55 ET (20:55 GMT)

Copyright (c) 2026 Dow Jones & Company, Inc.

At the request of the copyright holder, you need to log in to view this content

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Comments

We need your insight to fill this gap
Leave a comment