The hotly anticipated artificial intelligence (AI) company, which has risen 2,750% over the past five years, is poised to complete its first-ever stock split.
For the past 30 years, Wall Street and investors have been constantly on the lookout for the next big thing, technology, innovation, or trend. What’s unusual is when two trends capture attention at the same time. That’s what’s happening right now.
The rise of artificial intelligence (AI) has fascinated Wall Street as we enter 2023. The ability of software and systems to learn without human intervention will exponentially expand the market size of AI in the long term.
But we’ve also seen the stock split craze return.
A stock split is when a publicly listed company changes its share price and the number of shares outstanding. These are purely cosmetic measures designed to make the shares nominally more affordable to public investors, like a stock split, or to raise a company’s share price, like a stock split, to meet the minimum continued listing standards of a major stock exchange. Stock splits do not affect a company’s market capitalization or its underlying financial performance.
In the past six months, more than a dozen well-known, established companies have announced or completed stock splits, all but one of which were forward-looking splits. Investors tend to be attracted to companies that undertake forward-looking splits because they are undertaken from a strong position of management.
No stock split was as expected this year as AI giant Nvidia. (NVDA 2.83%) and Broadcom (AVGO 0.14%)Another fast-growing artificial intelligence stock is vying for attention.
Nvidia and Broadcom completed historic stock splits this year.
While big-name brands Walmart and Chipotle Mexican Grill kicked off the stock split party in 2024, Nvidia’s announcement on May 22 that it would implement its largest stock split in history, a 1-for-10 split, got current and prospective investors excited.
About two weeks after the stock split completed after the market closed on June 7, Nvidia’s stock briefly surpassed $140 per share, and the company’s market capitalization peaked at $3.46 trillion. The euphoria of the stock split, combined with the company’s hardware leading the AI revolution, made Nvidia briefly the largest publicly traded company in the world.
Nvidia has a near monopoly on graphic processing units (GPUs), the split-second decision-making that powers AI-driven data centers: In 2023, Nvidia will account for 3.76 million of the 3.85 million GPUs shipped to data centers, and current demand for the company’s superstar H100 GPU far outstrips supply.
History has shown that every big next-generation innovation is destined to burst, which would undoubtedly deal a huge blow to Nvidia’s stock price, but Nvidia has used its computing dominance to generate huge operating and stock price profits.
On June 12, AI networking specialist Broadcom followed suit, announcing a 10-for-1 stock split, marking Broadcom’s first stock split since Avago and Broadcom formally merged in early 2016. Shares began trading at the split-adjusted price on July 15.
Broadcom has been an early adopter of interest and demand for AI-related networking solutions. Last year, the company unveiled its Jericho3-AI Fabric, which can connect up to 32,000 high-performance GPUs. Its solutions are designed to reduce tail latency and help businesses get the most computing power out of their GPUs.
But Broadcom also generates a lot of revenue and operating cash flow outside of AI. Its wireless chips and accessories are staples in next-generation smartphones. In addition, Broadcom offers a range of products for industrial automation, next-generation automobiles, and cybersecurity solutions.
While Nvidia and Broadcom have been grabbing the spotlight, Wall Street’s newest AI stock split is about to have its moment of glory, with its share price rising 2,750% over the past five years.
See Wall Street’s latest AI-driven stock splits
The AI stock with the momentum to displace Nvidia and Broadcom is none other than Super Micro Computer, a specialist in customizable rack servers and storage. (SMCI 2.60%).
After the close of trading on August 6, Supermicro announced that its board of directors approved a historic 10-for-1 stock split. Like Broadcom, this will be Supermicro’s first stock split. It will be effective on September 30, 2024, and the stock is expected to open at nearly $51 (based on the closing price on August 8) when trading begins on October 1.
Just as Nvidia chips have become the “brains” of AI-accelerated data centers, Super Micro Computer appears to be the go-to choice for AI infrastructure companies. Net sales soared to $5.31 billion in the June quarter, up 144% from the same period a year ago, and demand for the physical infrastructure needed to power AI-accelerated data centers is widely expected not to slow down anytime soon.
At the same time, however, it is also necessary to debate whether supermicrocomputers have advanced too quickly.
As I mentioned earlier, over the past 30 years, every potentially game-changing innovation, technology, or trend has had an early stage bubble burst. This is a euphemism for the tendency of investors to always overestimate the adoption and usefulness of new innovations. Because most companies have no real game plan for AI, we are perhaps witnessing the next in a long line of early stage bubbles.
Supermicro also has a history of high-growth trends not playing out as expected: In the mid-2010s, the company’s stock price soared on hopes that its infrastructure would be essential to the new cloud computing boom. But demand for its customizable rack servers ultimately failed to live up to lofty expectations, and the stock price fell by about 75% over the course of several years.
Another thing to consider about Super Micro Computer is that its success is heavily dependent on Nvidia. Its rack servers are equipped with Nvidia’s H100 GPUs. If H100 production capacity remains limited, Super Micro may not be able to fulfill all of its infrastructure orders.
While Supermicro’s stock looks cheap (it trades at just 12 times forward earnings), that valuation is based on near-perfect execution going forward. History teaches us that the rallies that accompany new technologies and trends are rarely perfect.