Nvidia Corporation is risky at current levels.
ASML Holding is much better positioned to benefit from AI with a limited downside.
I present my outlook for both companies.
Money transfer between smart phones
There's no doubt that artificial intelligence ("AI") has been the hottest topic of 2023, and many investors are anxious to jump on the AI train in anticipation of life-changing gains. I get it, it's exciting, but it's exactly times like these when it's important to stay grounded, fight the FOMO (fear of missing out), and rely on proven investing principles.
Today I want to illustrate my point by first looking at $NVIDIA Corp(NVDA)$ , which is a great example of what I consider a risky stock, which trades at a very high P/E multiple due to the assumption that future revenues will skyrocket. I then want to present an alternative - $ASML Holding NV(ASML)$ which allows exposure to the AI trend, but is much more suitable for conservative investors that are looking to first and foremost preserve their capital and aim for total annual returns around 10% with limited downside.
Nvidia
The last time I covered the stock was after Q1 2023 earnings, when the company shocked everyone by guiding towards 52% revenue growth over a single quarter. As a result, the stock doubled in price from about $200 per share to $400 per share.
Nonetheless, I issued a strong sell rating for Nvidia, because I couldn't justify buying it at a forward P/E of 50x, especially with extremely aggressive growth assumptions already priced in. Frankly, I liked the business a lot, but not the valuation.
Since then, the stock did nearly touch $500 per share but has sold off soon after and now trades just 6% higher compared to when I last wrote about the stock. This is despite the fact that Nvidia managed to deliver on their $11 Billion revenue guidance and actually beat it with revenues of $13.5 Billion in Q2 2023.
data
This illustrates, nicely, why it's rarely a good idea to jump into an expensive stock with high growth expectations. When everyone expects miracles, even if they do materialize, there simply isn't much upside left.
Going forward, Nvidia has guided towards revenues of $16 Billion in Q3 and seems to be on track to deliver EPS of about $10 per share this year. Beyond, analysts' consensus calls for 50% growth in 2024 and about 20% during 2025-2028. That means that EPS could reach $31 per share by 2028.
But here's the thing. By that point, Nvidia will have grown by a factor of 3x, and its market cap will have surpassed $3 Trillion. At some point, future growth will inevitably start to slow and the stock will re-rate to a lower multiple. I say inevitably, because in order to achieve these growth rates, Nvidia will have to increase its market share aggressively and that simply cannot go on forever.
For the sake of my calculation, I'll assume a P/E of 25-30x five years from now, which is in line with what established tech companies with about 10% earnings growth tend to trade at, think $Microsoft(MSFT)$ $Apple(AAPL)$ .
For the bull case, which assumes that everything goes right, and the company delivers on analyst's ambitious growth estimates, we can expect a total annual return of 15%. Anything above this is highly unlikely in my opinion.
data
But let's not forget that inflation remains sticky and missed expectations in August and the Fed delivered a clear hawkish message just last week increasing the probability of higher interest rates for longer. Moreover, the deeply inverted yield curve along with the nearly depleted excess savings of consumers suggest that a recession is certainly on the cards during the next 12 months.
And like it or not, the chip industry is highly cyclical. The macroeconomic situation is likely to present some short to medium-term headwinds to Nvidia, which could easily result in lower-than-expected revenue growth and/or worsening margins.
I think that growth could easily slow to 20% next year and 15% beyond, which would still be well above the 9% CAGR which ASML expects for the semiconductor market over the rest of the decade. In this case modestly bearish case, the total expected return would be just 6.6%, likely below what investors could get elsewhere.
data
And frankly, things could get a lot worse, if the company fails to increase its market share and only grows with the sector as a whole at 9% per year. In that case, we're looking at zero or negative returns over the next five years.
ASML
At the beginning, I promised an alternative suitable for conservative investors who still want at least some exposure to AI, so here it goes.
ASML is a Netherlands-based manufacturer of extremely sophisticated extreme ultraviolet (EUV) and deep ultraviolet (DUV) lithography machines that are necessary for production of some of the smallest and most advanced semiconductors.
Notably, ASML is the only company in the world that is capable of producing EUV machines and therefore effectively runs a monopoly, which gives it significant pricing power.
Currently, the company ships most of their products to Asia, in particular Taiwan (34%), South Korea (27%), and China (24%). This is a direct consequence of the fact that the vast majority of semiconductors and chips are produced in those countries these days. Of course, this presents a potential risk if, for example, the China-Taiwan situation escalates. But here's the thing. The world will need chips no matter what, and the West is already trying to move some of the production capacity back. So, while geopolitical tensions could provide near-term headwinds for ASML, the long-term outlook remains intact.
data
Management has stated on the earnings call that they continue to see strong demand for their products, and despite an anticipated slowdown in 2024 due to macroeconomic headwinds, they remain on track for their long-term targets.
In particular, they expect to grow revenues from about EUR 28 Billion this year to EUR 30-40 Billion by 2025 and EUR 44-60 Billion by 2030. Moreover, they also expect their gross margin to improve from about 51% today to 54-56% by 2025 and 56-60% by 2030.
Expected revenue growth stands at 6.6% per year at the low end of guidance and at 11.3% per year at the high end. These growth rates aren't overly aggressive and are roughly in line with the expected growth of the sector of about 9%.
At the low end of guidance, EPS is likely to grow from about $20 this year to $33 by 2030. Now, for the multiple, ASML is one of the very few companies where I'm fine with paying 30x earnings, because it is a virtual monopoly in a very important sector set to grow for decades as technology evolves and the company is years, if not decades, ahead of the competition.
At 30x earnings, the low-end price target stands at $1,000 per share by 2030 which corresponds to an annual price appreciation of 7.7% on top of a roughly 1% dividend, for a total expected return of 8.7%.
data
ASML has always been fairly transparent about their expectations for the business, and they tend to hit their targets. That's why I'm optimistic that they will meet their guidance and likely end up around the mid-point, which would correspond to a price target of $1,250 per share by 2030 and translate into 11.5% of upside per year on top of the dividend for a total return of 12.5%.
I don't want to be overly aggressive and assume that ASML will hit the top-end of their guidance. My expectation is for a 10-12% total annual returns for the rest of the decade. There could be some near-term, downside caused by the cyclicality of the business or geopolitical tensions, but given the modest growth assumptions and a lower valuation I think the potential downside is much smaller for ASML than for Nvidia.
Bottom Line
Nvidia is priced for perfection, and even if all goes well, the upside is likely capped at around 15%. On the other hand, if the economy slows and the company fails to aggressively gain market share, it could easily result in sub-par, even negative returns. I rate Nvidia as a sell (though not a short!) and suggest an alternative in the form of ASML.
ASML holds a virtual monopoly in a sector that our world cannot function without. It is arguably one of the best-positioned players to benefit from the AI trend, and given its lower valuation and reasonable growth assumptions, it provides a very attractive risk-reward opportunity. I expect to earn 10-12% total annual returns for the rest of the decade with a significantly lower downside compared to Nvidia and rate ASML as a buy.
Comments