HomeIndustriesThe AI ​​boom is frightening

The AI ​​boom is frightening

This article is an onsite version of our Unhedged newsletter. Premium subscribers can enroll Here to receive the newsletter every working day. Standard subscribers can upgrade to Premium here or explore all FT newsletters

Good morning. The big event this week is the primary presidential debate of 2024, which, like every presidential debate, can be terrible. But Unhedged feels compelled to listen as to if either candidate is serious about reducing the gaping budget deficit that has been a serious support for stock prices. What will you be anticipating? Email me: robert.armstrong@ft.com.

Insights into the AI ​​Rally

Last week, Unhedged checked out how virtually the entire gains within the S&P 500 since late March have come from AI and AI-related stocks. Below is a chart showing the dollar and percentage gains of Nvidia, the “Fab Five” Big Tech stocks (Alphabet, Amazon, Apple, Microsoft, Meta), and a basket of semiconductor stocks which have soared in anticipation of an AI-driven digital investment boom:

I find this a bit of concerning. A rising, historically expensive market that may easily fall if it weren't for a single investment theme could, in my view, turn right into a bear market if sentiment on that theme turns sour. For example, imagine if Nvidia were to lower its revenue forecast or certainly one of the Fab Five were to chop its capital spending budget.

Expectations are clearly high for AI stocks, however it's probably value getting a bit of more specific here. Consensus estimates for revenue growth next yr and for 2026 actually don't seem overly ambitious. Analysts expect a 23 percent annual growth rate for Nvidia over that period. This would represent some moderation in the speed; over the past five years, Nvidia's revenue has grown 50 percent per yr. Likewise, the two-year revenue growth rates expected for the Fab Five are at or below growth rates in recent history. Only a handful of chip stocks — Micron, Texas Instruments, Analog, and Lam — are expected to see significant acceleration in revenue.

Is the AI ​​group's recent surge because of earnings forecast increases? Looking on the 2025 estimates, probably not. Since the tip of March, earnings forecasts for all the group have increased by a low single-digit percentage. Apple, Amazon and Micron are the one ones which have received more significant increases.

What has radically modified within the AI ​​rally is, not surprisingly, valuations. Over the past three months, the price-to-earnings ratios of Nvidia, Apple, Broadcom and Qualcomm have all risen by over 20 percent. Looking back to last October, when the rally began, the typical (harmonic mean) P/E of the AI ​​group has risen by almost 50 percent.

That's lots. How should we interpret that? Maybe it just reflects momentum and animal spirits. Put more charitably, it would reflect an expectation that the AI ​​business will proceed to generate increasing profits for a few years to return. In other words, it's a bet on the competitive dynamics inside the AI ​​industry: that there can be no hyper-competition and that the winners in the long term can be the identical as today's winners – the Fab Five and today's market leaders within the semiconductor industry.

The second half of the bet, that incumbents will proceed to win, seems reasonable to me. Incumbents within the technology sector are very powerful, within the sense that corporations can leverage their strong market position in a single technology to construct a powerful position in one other (consider Microsoft's shift from PC operating systems to cloud computing). The first half of the bet, that AI won’t change into a capital-intensive knife fight where nobody makes big profits, is something I cannot comment on.

To assess the steadiness of the rally, we also must take a more in-depth take a look at how the non-AI stocks are performing. Here is how the S&P 500 sector performed after removing AI stocks:

Bar chart of market-weighted sector performance of the S&P 500, excluding AI stocks*, 3/28/24-6/21/24, showing Defensives > Cyclicals” data-image-type=”graphic” width=”2100″ height=”1500″ loading=”lazy”></picture></figure>
<p>Since March, non-AI stocks are down 2 percent overall (again, market-weighted) and 9 of the 11 sectors are down. And there’s a really clear pattern: defensive sectors (utilities, staples, healthcare) are doing relatively well and cyclical sectors (energy, materials, industrials, consumer goods) are doing relatively poorly. The non-AI market doesn’t seem very optimistic in regards to the economy.  </p>
<p>Unhedged recently identified that utilities' outperformance could also be due partially to them underperforming and becoming low cost for a protracted period starting in 2022, and that the prospect of rate of interest cuts has also made their yield more attractive. Additionally, I've read an argument or two that the AI ​​boom, being incredibly power hungry, will increase demand for power generation. In the present environment, I wouldn't rule out the potential of the AI ​​hype reaching utilities as well.</p>
<p>Nevertheless, the defensive bias of the market that doesn’t bet on artificial intelligence is clearly visible. Another indicator of this could possibly be the wonderful recent performance of Apple shares, which have accounted for half of the Fab Five's increase in value since March. With its high margins, recurring revenue and high customer loyalty, Apple appears to be a protected bet.</p>
<p>An internal tension inside the AI ​​rally is that the revenues of the leader Nvidia represent a value to a few of its biggest beneficiaries, the Fab Five. In the short term, Nvidia's success is a drain on the money flows of the Big Tech corporations that buy the vast majority of Nvidia's chips. Charles Cara of Absolute Strategy Research recently made a provocative remark about this. He points out that 40 percent of Nvidia's revenues come from Microsoft, Meta, Amazon and Google, and that even the very large expected increase in capital spending at these corporations just isn’t very large relative to the expected increase in Nvidia's revenues. His chart:</p>
<figure class=

The increase in capital spending by the 4 corporations between last fiscal yr and 2025 is $54 billion, which is greater than 40 percent of Nvidia's expected $100 billion increase in revenue. However, only a fraction of Big Tech's capital spending is prone to go toward Nvidia's GPUs. So either Big Tech will spend more or Nvidia will earn less. “We fear that either the technology sector's money flows or Nvidia's revenues are overestimated,” Cara writes.

It's possible that spending by other tech corporations at Nvidia will increase rapidly next yr, offsetting any shortfalls at the massive techs, but that may show how competitive the AI ​​business could change into.

A superb read

Social media is a tricky business.

FT Unhedged Podcast

Can't get enough of Unhedged? Listen our latest podcastto immerse yourself in the most recent market news and financial headlines for quarter-hour twice per week. Read previous editions of the newsletter Here.

Recommended newsletters for you

Swamp Notes — Expert insights into the intersection of cash and power in US politics. Sign up Here

Chris Giles on central banks — Important news and views on central bank considering, inflation, rates of interest and money. Sign up Here


Please enter your comment!
Please enter your name here

Must Read