Nvidia (NVDA +0.87%) is one of the hottest stocks on the market -- especially after its 10-for-1 stock split made it easier for smaller investors to experience the artificial intelligence (AI) boom that sent its shares up 150% this year alone. But while stock splits are exciting, they don't change a company's market cap (the value of all shares combined) or its valuation relative to growth potential and earnings.
With AI technology in its early stages, Nvidia still looks like a long-term winner because of its exposure to the market. But new investors should be aware that they are late to the party. Let's explore three reasons I'm downgrading the industry-leading chipmaker from a screaming buy to an optimistic hold.
NASDAQ: NVDA
Key Data Points
1. Potentially unsustainable margins
When you crack open Nvidia's income statement, the first thing that jumps out is the company's first-quarter gross profit of $20.4 billion. This metric represents the revenue earned from selling its products minus the direct costs of creating them (before overhead expenses like office salaries, research, or advertising). And with a gross margin of 78.4%, this number is extraordinarily high for a hardware company.
Unlike software, which is essentially just computer code, chip manufacturing requires building components from scratch out of an array of raw materials like silicon, copper, cobalt, and other rare earth metals with the help of lithography machines, which are very complex. Nvidia outsources most of this work to third parties, such as Taiwan Semiconductor Manufacturing, which creates the cutting-edge chips at its Taiwan-based factories.
Nvidia's high gross margin suggests it is selling its products at a tremendous markup over the fees it pays its manufacturing partners to make them. And this situation might not be sustainable over the long term. According to Morgan Stanley analysts, TSMC is considering increasing its production fees for Nvidia, which could eat into its gross margin -- and, by extension, its profitability.
2. The consumer-facing AI industry is not very profitable
While AI is taking Wall Street by storm, it isn't very profitable on the consumer-facing software side of the equation. According to research from Sequoia Capital, the industry has already spent around $50 billion on Nvidia's cutting-edge AI chips, while generative AI start-ups have only brought in a comparatively measly $3 billion in revenue. These companies are almost certainly burning through huge amounts of cash to operate.
Image source: Getty Images.
As a hardware supplier, Nvidia's "picks-and-shovels" exposure to the industry allows it to make money even when its clients are losing. But this situation can't last forever. If the consumer side of the industry doesn't turn around, Nvidia's clients will become unable to sustain their current demand for chips, leading to a revenue slowdown.
3. Poor diversification
A slowdown in AI chip sales wouldn't be so terrifying if Nvidia had other businesses to fall back on. But after over two years of breakneck growth, its data center business has practically made its other segments irrelevant.
In the first quarter, its data center segment represented 87% of Nvidia's $26 billion revenue, driven by sales of advanced graphics processing units (GPUs) to run and train AI applications. The company's previously core gaming and PC segment generated just $2.6 billion (10%) of revenue in the period.
Over time, Nvidia's revenue streams could rediversify as use cases like self-driving cars and warehouse automation increase demand for chips developed by the company's automotive and robotics segments. But in the near term, Nvidia remains exceptionally undiversified and vulnerable to any potential reduction in demand for data center AI chips.
Nvidia is a hold
After considering Nvidia a strong buy over the last two years, I'm ready to downgrade the stock to an optimistic hold. While the chipmaker remains the best way to bet on the long-term expansion of the AI industry, the near-term risks are becoming too big to ignore. Investors may want to wait for a significant pullback in the stock price before buying shares.
