AI semiconductor stocks have taken a big dip from their July highs. But is the drop an opportunity, or is there an AI bubble deflating?
AI skeptics abound over the technology's ultimate usefulness, or whether the strong spending can continue if there is a slowdown in the economy.
However, the one place you won't find evidence of an AI slowdown is in the earnings calls and commentary from technology executives themselves. This week, Oracle (NYSE: ORCL) founder and Chairman Larry Ellison gave an incredibly bullish outlook on AI spending through this decade, which should allay any fears for long-term tech investors .
Why have AI semiconductor stocks dropped?
It may be shocking to some that the likes of Nvidia (NASDAQ: NVDA) , Broadcom (NASDAQ: AVGO) , and Advanced Micro Devices (NASDAQ: AMD) have dropped so much over the summer, with these stocks now residing 25% to 41% below their all-time July highs.
There could be several reasons behind the drops, despite each company posting strong AI-related growth currently.
One, these stocks have had quite a run since the semiconductor stock bottom in late 2022, with each posting large gains over the past 18 months or so. Some investors are likely nervous about those big gains increasing the concentration in their portfolios, and may have looked for any excuse to book profits or rebalance their portfolios.
Furthermore, recession fears came to the fore after a weak July jobs report, even as the Federal Reserve chose to keep interest rates high that month. While the August jobs figure showed sequential increase in job additions, the August figures were also below expectations . So some investors may be growing nervous over the economy.
While they have outsized long-term growth prospects, semiconductor stocks are traditionally known to be cyclical . That could again spur older investors near retirement or short-term traders who can't afford large near-term declines to take profits.
Finally, some investors are skeptical that the enormous revenue gains from Nvidia and others aren't sustainable. While basically all of the Magnificent Seven stocks said they will continue spending into next year, some analysts are worried their near-term revenue growth and profits may not match all that spending. If AI spending doesn't lead to a material increase in growth for these large tech stocks, these CEOs may find all this AI infrastructure spending hard to justify.
"An ongoing battle for technical supremacy"
On Monday night, database and cloud giant Oracle held its first fiscal quarter earnings release, with both results and guidance impressing investors, sending the stock up double-digits the next day.
Since the topic of the day is AI, management was asked on the conference call with analysts about whether the huge spending on training AI models was a bubble that pulled forward chip demand. One skeptical analyst posed that once the models are trained, the bulk of spending would be over.
Chairman Larry Ellison negated that assertion pretty handily with an interesting metaphor:
A lot of people think that, my God, I send a kid to college and then I'm done. They're training's over. I got four years of training, and then I can put the kid to work and they'll be doing inferencing. And that's not true. This race goes on forever, to build a better and better neural network."
Ellison pointed out that AI investments should continue to grow for at least five, and probably 10 years. This is for a couple reasons. First, there are a handful of companies that can make the largest frontier models. These companies have a lot of cash, and can't afford to cede the market to someone else. Ellison called it an, "ongoing battle for technical supremacy," that will be fought by a handful of companies and "maybe one nation-state" over the next decade.
Image source: Getty Images.
Ellison noted the entry price for one frontier model was probably $100 billion. If one looks at the capital spending plans for the Mag 7 stocks, even their increased spending plans for the year don't add up to that. For instance, Meta Platforms (NASDAQ: META) , which is building out its own Llama frontier models, just upped its capital expenditures , mostly for data center infrastructure for AI, to $37 billion to $40 billion for 2024. Microsoft (NASDAQ: MSFT) , which has invested in and has an exclusive cloud arrangement with OpenAI , spent $19 billion on capex last quarter, including finance leases, but also noted only about half was for data center infrastructure.
In addition to just the large frontier models, Ellison pointed out that there will be a lot of smaller but specialized models in every field. He gave the example of the medical industry, where millions of CAT scans, biopsies, or blood samples could be processed by industry-specific AI models to detect cancer. Those models will be built in addition to the large general models. The same goes for sovereign nations who want their own AI.
Not sold separately
Meanwhile, Ellison stressed that AI isn't some separate thing that is sold to customers, it will be embedded in Oracle's current products, such as database administration or automated records retrieval in Oracle's Cerner hospital software.
It's not something you sell separately. It's the diagnostic system. It's the electronic health record system. The pharmacy system, this prescription system, the user authentication, the log-in system, it's all AI. And I know people think it's a separate thing that, "Oh my God." And I hear a bunch of applications companies say, "Oh, we've now got AI agents. We'll charge for it separately." I mean, our applications are going to be primarily AI applications, everything. How do you charge separately for everything?
And this gets to a key point. While nervous investors may have pointed to a lack of specific AI-related growth, technology companies need to invest in AI going forward in order to keep their existing products competitive. As such, there's a big risk to present and future revenues if they don't invest. And that should keep AI spend high for years.
"Enormous" and "immense" demand
CEO Safra Catz added that Oracle is seeing "enormous" and "immense" demand for Oracle's cloud infrastructure for AI processing. That's in line with what other tech companies are saying this earnings season.
And yet, AI-related chip stocks are having a significant pullback even in spite of these comments. But that leads me to believe the current downturn in AI stocks is a mere pullback preceding larger gains, and not some imminent crash or longer downturn.
As such, now is a great time to load up on your favorite AI chip stocks for the long haul. Over the coming years, the growth should continue and they should bounce back.
Should you invest $1,000 in Oracle right now?
Before you buy stock in Oracle, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Oracle wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $716,375 !*
Stock Advisor provides investors with an easy-to-follow blueprint for success, including guidance on building a portfolio, regular updates from analysts, and two new stock picks each month. The Stock Advisor service has more than quadrupled the return of S&P 500 since 2002*.
See the 10 stocks »
*Stock Advisor returns as of September 9, 2024
Randi Zuckerberg, a former director of market development and spokeswoman for Facebook and sister to Meta Platforms CEO Mark Zuckerberg, is a member of The Motley Fool's board of directors. Billy Duberstein and/or his clients have positions in Broadcom, Meta Platforms, and Microsoft. The Motley Fool has positions in and recommends Advanced Micro Devices, Meta Platforms, Microsoft, Nvidia, and Oracle. The Motley Fool recommends Broadcom and recommends the following options: long January 2026 $395 calls on Microsoft and short January 2026 $405 calls on Microsoft. The Motley Fool has a disclosure policy .
Local News
Comments / 0