Since the proliferation of the internet roughly 30 years ago, there hasn't been a next-big-thing technology, innovation, or trend that's come remotely close to rivaling it... until now.
The arrival of artificial intelligence (AI) is forecast to add $15.7 trillion to the global economy by 2030, according to analysts at PwC. With AI, software and systems are given autonomy over tasks that would normally be overseen or undertaken by humans. The catch is that these systems have the ability to learn and evolve over time without human intervention. The capacity to become more proficient over time gives AI utility in pretty much every sector and industry.
Although most AI stocks have been unstoppable over the last 18 months, it's Nvidia (NVDA -1.91%) that unquestionably sits on a pedestal above all others.
Image source: Getty Images.
Since 2023 began, shares of Nvidia have gained 828%, as of June 19, 2024, with the company adding nearly $3 trillion in market value and undergoing a recent 10-for-1 stock split. In fact, Nvidia unseated Microsoft and Apple this week to become the world's largest publicly traded company.
But while short-term catalysts help explain the euphoria surrounding AI and Nvidia, tangible long-term headwinds are mounting that suggest the world's hottest artificial intelligence stock is in an irrational bubble that could, eventually, push it out of the trillion-dollar market cap club.
Euphoria surrounding Nvidia may be nearing a crescendo
No company has more directly benefited from the AI revolution than Nvidia. The company's H100 graphics processing units (GPUs) have, in short order, become the standard in AI-accelerated enterprise data centers. Nvidia's hardware is effectively the "brains" behind the split-second decision-making and computational power needed to train large language models and run generative AI solutions.
Recently, the semiconductor analysts at TechInsights released data showing that 3.85 million GPUs were shipped in 2023. Nvidia was responsible for 3.76 million (98%) of these shipments. This makes it easier to understand why the company's Data Center segment more than quintupled sales in the fiscal first quarter (ended April 28), compared to the prior-year period.
Moreover, demand has completely overwhelmed the available supply of AI-GPUs. When the demand for a good or service swamps supply, it's normal for the price of that good or service to meaningfully increase. It's not uncommon to see H100 GPUs selling for around $30,000, which has lifted Nvidia's adjusted gross margin to a scorching-hot 78.4%!
Nvidia's first-mover advantages are helping it on the innovation front as well. With rivals attempting to play catch-up with the H100, Nvidia has been busy developing its next-generation AI-GPU architecture. It unveiled Blackwell in March, which it'll begin rolling out in the second half of the current calendar year, as well as Rubin, which was revealed in June and is expected to be released in 2026. On a compute basis, catching Nvidia could prove challenging for its external competition.
With Nvidia blowing the doors off of Wall Street's sales and growth expectations for more than a year, it's understandable why the fear of missing out (FOMO) has taken hold among investors. Unfortunately, the combination of FOMO and next-big-thing investment trends has historically been a train wreck waiting to happen.
Image source: Getty Images.
Nvidia might struggle to remain a trillion-dollar company by 2026
The single biggest enemy for Nvidia and its shareholders is history. While history shows that the major stock indexes rise over long periods, it's also quite clear that next-big-thing investment trends undergo a maturation process that involves a bubble-bursting event.
Since the mid-1990s, every touted game-changing technology, innovation, or trend has resulted in an early-stage bubble. While not a comprehensive list, this includes the internet, businesses-to-business commerce, genome decoding, nanotechnology, housing, China stocks, 3D printing, cannabis, blockchain technology, augmented reality, and the metaverse. Without fail, investors always overestimate the uptake of these game-changing innovations/trends by consumers and businesses, which leads to lofty expectations not being met. It would be foolish (with a small "f") to expect AI to change this trend.
To add fuel to the fire, most businesses lack a blueprint of how they're going to utilize artificial intelligence to grow their sales. While many of the most influential businesses in America are investing in AI solutions because that's the hot thing to do right now, it's not really moving the needle for a majority of these businesses (outside of hardware players like Nvidia). Every technology needs time to mature, and AI isn't anywhere close to being a mature innovation at this stage of the game.
Competition is another clear problem. Even if Nvidia retains its GPU compute advantages over its peers, the company appears destined to lose share. Advanced Micro Devices and Intel are both rolling out their AI-GPUs designed to directly compete with the H100 in AI-accelerated data centers. With demand grossly overwhelming supply for Nvidia's chips, AMD and Intel should have no trouble earning share from impatient enterprise customers.
As I've pointed out on countless occasions, Nvidia's competition is also internal. Microsoft, Meta Platforms, Amazon, and Alphabet comprise roughly 40% of Nvidia's net sales.
While it's fantastic that Nvidia can call the world's most influential businesses its top customers, it's equally concerning that Microsoft, Meta, Amazon, and Alphabet are internally developing AI-GPUs for their respective data centers. Once again, Nvidia can maintain its competitive compute advantage and still lose if these four companies choose to lean on their own chips and lessen their reliance on the largest publicly traded company by market cap.
As the sheer number of AI-GPUs deployed increases, the scarcity that's powered the selling price of the H100 into the stratosphere is going to ebb. In other words, it's a scenario for Nvidia's adjusted gross margin to retrace back to historic norms.
Having witnessed similar scenarios play out numerous times with next-big-thing innovations over the last three decades, it's only logical to expect Nvidia's FOMO to wear off as well. When it does, which I expect to occur by or before 2026, Nvidia could struggle to remain a trillion-dollar company.
