Billionaire Ken Griffin Just Increased His Position in This Data Center Stock by 704% (Hint: It's Not Nvidia)
Billionaire Ken Griffin's Citadel Advisors increased its position in data center stock Equinix by 704%, now holding approximately 564,000 shares. This move highlights the growing demand for data center services driven by the rise of AI technologies, particularly large language models. Equinix's CEO noted that AI training workloads are significantly boosting demand, similar to the cloud computing boom. Despite a high forward P/E ratio of nearly 80, data center stocks are gaining attention as a lucrative investment opportunity in the AI sector.
Semiconductor stocks have been some of the biggest beneficiaries in the ongoing artificial intelligence (AI) revolution. Chipsets known as graphics processing units (GPUs) are important for developing generative AI, and companies including Nvidia, Advanced Micro Devices, and Taiwan Semiconductor have emerged as early winners in the GPU realm so far.
IT infrastructure is an area tangential to the GPU landscape, and I continue to think it's going overlooked. GPUs are stored in data centers, so wouldn't it make sense that as demand for these chips rise, so will the need for data center services?
Well, apparently billionaire investor Ken Griffin of Citadel Advisors might think so. According to Citadel's most recent 13F filing, the fund increased its position in data center stock Equinix (EQIX 0.86%) by 704% during the second quarter -- bringing its position to roughly 564,000 shares.
I'll outline why I see Equinix as an under-the-radar opportunity in the AI space, and will assess if I think the stock is a good buy right now.
How is AI demand impacting Equinix?
One of the most popular applications in AI right now is the large language model (LLM). LLMs such as ChatGPT, Claude, and Gemini have loads of features -- from image creation, generating software code, or generic search functionality, these models are changing the way people interact in the workplace and online.
And while LLMs have the ability to generate answers to your queries almost instantly, the underlying build supporting these models is far more complex than you might realize. Generative AI that can process questions and perform tasks quickly takes an incredibly long time to develop. The reason is that these machine learning (ML) applications undergo ongoing training and inferencing testing. Another way of looking at it is that AI models are constantly processing vast amounts of data through complex algorithms -- indeed, a big tailwind for the data center market.
During Equinix's last earnings call, CEO Adaire Fox-Martin made an interesting analogy when comparing the rise of AI to that of cloud computing a decade ago. He stated that "in the near term, AI training workloads are driving significant demand" while inference demand is also "beginning to take shape."
Image source: Getty Images.
Why am I bullish on the long-term outlook?
Today, cloud services have become a multibillion-dollar opportunity for tech stalwarts such as Amazon, Alphabet, Microsoft, and Oracle. One of the reasons for this is that demand for digital infrastructure has risen in parallel with businesses investing more heavily into data to make more informed, efficient decisions.
But with that said, the rise of the cloud did not happen overnight. Over the course of many years, Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform have evolved into more sophisticated products, spanning across database management, cybersecurity, and much more.
Similar to cloud infrastructure, AI platforms should become more advanced in the coming years. Equinix's core data center services, including its xScale suite, stand to benefit from secular tailwinds fueling AI demand -- particularly the $1 trillion IT infrastructure opportunity.
Fox-Martin gave investors a preview of the long-run impacts AI could have on the business when he shared, "As with cloud, Equinix continues to be the preferred location for network nodes as customers seek the right connectivity solutions for data ingestion and distribution."
Is Equinix stock a good buy right now?
The chart benchmarks Equinix against a small, competitive cohort on a forward price-to-earnings (P/E) basis. Although Equinix is right in the middle between Digital Realty and Iron Mountain, a forward P/E of nearly 80 is by no means a bargain. The average forward P/E of the S&P 500 is only around 23.
EQIX PE Ratio (Forward) data by YCharts
There are a couple of important takeaways from these trends. First, even though Equinix and its peers are a bit pricey, investors appear to be placing a premium on the data center market compared to the broader market. This could suggest that data center stocks remain a compelling choice among AI investors.
Moreover, all three companies have experienced notable valuation expansion over the last several months. This trend could support the idea that the data center pocket of the AI realm is beginning to fetch more attention and is becoming an increasingly lucrative opportunity.
In the coming weeks, institutional investors such as Citadel will be publishing updated 13F filings for the third quarter. I would keep a keen eye on whether Citadel and others on Wall Street are adding or reducing positions in data center stocks.
As Fox-Martin alluded to, the AI story is still early, and demand for data processing and storage protocols should continue rising as AI needs become more robust. For now, I think Equinix stock is worth monitoring but I see opportunities to invest at more reasonable valuations down the road.