Nvidia stock down 27%, and an eventual 75% pullback could be in the cards

For most of the past three decades, investors have had a technology, innovation or trend that captivated their attention.

Since the Internet completely changed the trajectory of business growth in the mid-1990s, we have witnessed massive addressable markets in genome decoding, business-to-business trading, Chinese stocks, US housing, nanotechnology, 3D printing, cryptocurrency, blockchain. technology, cannabis, augmented/virtual reality and the metaverse. However, it is artificial intelligence (AI) that could be the cream of the future trends.

Conformable Size of the awardwhich is an in-depth report released by analysts at PwC, AI has the potential to add $15.7 trillion to the global economy by the end of the decade. These gains are expected to come in the form of increased productivity and various consumer benefits.

A $15.7 trillion opportunity leaves plenty of room for several companies to emerge as winners — perhaps none more so than the semiconductor behemoth Nvidia (NVDA 8.15%).

Image source: Getty Images.

The AI ​​revolution has driven Nvidia’s historic rise

When 2022 ended, Nvidia was a $360 billion company that had marginal importance in the technology sector. Less than 18 months later, shortly after completing a historic 10-for-1 stock split, its valuation would tip the scales at a peak of $3.46 trillion, making it (briefly) the most valuable publicly listed company in the world.

At no time in history have we seen a market-leading business add more than $3 trillion in value in less than 18 months. The catalyst behind this unprecedented move has undoubtedly been the AI ​​revolution.

It didn’t take long for Nvidia’s graphics processing units (GPUs) to become the top choice for companies running generative AI solutions and looking to build/train large language models (LLMs). Orders for the company’s H100 GPU and next-generation Blackwell chip, which is scheduled to debut in the first quarter of 2025, are pending.

To add to this point, Nvidia’s CUDA platform plays a key role in keeping enterprise customers loyal to its ecosystem. CUDA is the toolset that developers use to build LLMs and get the most out of their Nvidia GPUs.

Controlling roughly 98 percent of the AI-accelerated data center GPU market in each of the past two years has given Nvidia an astonishing degree of pricing power. While businesses pay between $10,000 and $15,000 for the MI300X AI-GPU from Advanced microdevicesthey’re a steal at $30,000 to $40,000 for Nvidia’s H100. A 100% to 300% premium to competing tokens has translated into better than 10 percentage points growth in the company’s gross margin over the past six quarters.

Despite this seemingly ideal execution, Nvidia’s share price is down 27% from its June high and could be poised for an even steeper decline.

Nvidia may eventually lose three-quarters of its value

Given the high growth and profit projections that Wall Street expects Nvidia to achieve in the coming years, a possible 75% drop from its all-time high of $140.76 probably sounds absurd. However, history has not yet been beaten when it comes to future innovations.

Although quite a few game-changing technologies and trends have become quite successful for patient investors (for example, the Internet and business-to-business commerce), one thing that all new technologies, innovations, and trends share is early-stage bubbles.

While it’s impossible to pinpoint when the euphoria surrounding a potentially game-changing innovation will wear off, one consequence of Wall Street’s “next big thing” is that investors always overestimate how quickly a technology or trend will become mainstream.

Based on the amount of money companies are spending on their AI data centers, you’d probably think artificial intelligence is on the path to rapid adoption. But dig deeper and you’ll find that most companies don’t have a clear game plan for how they’ll implement AI or monetize the technology to their advantage. It’s a telltale sign that we’re witnessing the next bubble taking shape. The market leaders of subsequent bubbles have a history of losing 75% to 99% of their value from peak to trough.

The only saving grace for Nvidia is that it had other core operating segments before the AI ​​euphoria took hold. Even if the technology takes several years to mature and sales slow dramatically, Nvidia will be able to rely on sales of GPUs for gaming and cryptocurrency mining, as well as revenue from its virtualization software segment, to dampen the top and bottom results somewhat. .

Image source: Getty Images.

Nvidia is also poised to face fundamental pressures

Beyond historical precedent, there are also fundamental reasons to believe that Nvidia’s luster will fade as external and internal competitive pressures mount.

As we alluded to earlier, AMD’s MI300X is priced a notable 50% to 75% below the retail price of the H100. Between Nvidia’s extensive inventory and the lure of considerably less expensive AI-GPUs, AMD and other Nvidia rivals should have no trouble locating demand for data center hardware.

What seems to fly under the radar for most investors is the potential for internal competition.

Four members of the “Magnificent Seven” account for nearly 40% of Nvidia’s net sales. While on the one hand it’s fantastic that Nvidia is the core hardware provider for America’s most influential businesses, investors should also understand that all four of these companies are internally developing AI-GPUs to implement in their data centers. Even though Nvidia’s H100 and Blackwell chips retain their computing advantage, these Magnificent Seven companies are motivated to use these cheaper and more available internally developed chips in a complementary way in their data centers.

The real problem with an increase in domestic competition is that it reduces the AI-GPU shortage that has driven Nvidia’s exceptional pricing power. With less pricing power over time, the expectation would be for the company’s gross margin to decline.

Moreover, domestic competition will reduce opportunities for Nvidia to secure valuable data center real estate from America’s top companies. This suggests that we may have witnessed a peak in orders from the Magnificent Seven.

Nvidia is also historically expensive. There have only been a few cases in the past 30 years where a business at the forefront of a future innovation has had a trailing 12-month price-to-sales ratio (P/S) close to 40. Previously cases have resulted in these stocks losing around of 90% of their value, from peak to trough.

Whether it’s historical precedent or the fundamental roadmap. Nvidia looks likely to finally give up much of its early 2023 gains.