Sell Nvidia, Buy Intel Stock?

As the face of the AI boom, Nvidia’s stock has risen more than 180 percent this year, bringing its valuation close to $3 trillion. The increasing demand for Nvidia’s GPUs, which are now the de facto chips for AI applications, is expected to drive the company’s revenues to more than double this fiscal year. On the other hand, Intel’s stock has had a difficult year. The market value of the stock is only $100 billion, and it has lost roughly 50% of its value so far this year. It is anticipated that Intel’s revenue will decline this year. But here’s the twist: It may be time to reconsider the AI bellwether. Also, observe a cryptocurrency mover that has increased by 300 percent in a month; XRP is only warming up. Why is that?

The markets frequently exhibit myopia and have a propensity to extrapolate short-term trends to the long term. According to Nvidia, the market for AI accelerators will continue to grow, and the company’s profit margins and growth rates will continue to be robust. Investors, however, are not optimistic about Intel’s future due to its foundry business issues and its declines in market share in the CPU market. That being said, practically everything in life is cyclical, and the semiconductor markets are no different. At this point, it might make sense to reduce holdings in Nvidia and take Intel stock into consideration. We explain below – the ‘why’. Consider the High Quality portfolio, which has outperformed the SandP and generated returns of over 91 percent since its inception, if you want upside with a more comfortable ride than an individual stock.

Nvidia may be leading the AI boom.

Over the past two or so years, businesses have invested enormous resources in creating AI models. These days, training these enormous models is more of a one-time event that calls for a significant amount of processing power. Nvidia has benefited greatly from this since its GPUs are thought to be the fastest and most effective for these kinds of tasks. This is clear from the recent increase in Nvidia’s revenue. From a meager $27 billion in FY’23 to nearly $130 billion in FY’25, sales are expected to grow. However, the field of AI might be changing. Incremental performance gains should decrease as models get bigger in terms of multiple parameters. Separately, there may be a bottleneck in the supply of high-quality data for model training. Since large language models already process a large portion of the high-quality data on the Internet, there may be a move away from large-scale, general-purpose AI models and toward more specialized, smaller models, which would lower demand for Nvidia’s powerful GPUs. Nvidia may have experienced front-loaded demand in the last few years, and future growth is probably going to slow.

Inference, the stage at which trained models produce outputs, may now become the demand for AI-related chips instead of training. Since inference requires less computing power, it may pave the way for the development of alternative AI processors. Nvidia will probably continue to dominate the inferencing market as well (it claims that inferencing makes up around 40% of its demand for data center chips), but competitors like AMD and possibly even Intel have a chance to capture a small portion of the market.

Enterprises and large tech companies rushed to purchase GPUs during the early wave of generative AI out of “fear of missing out,” without considering costs or returns on investment. With its net margins hitting more than 50% in recent quarters, Nvidia’s pricing power increased as a result. Nevertheless, businesses and their investors will eventually seek returns on their capital, which could lead them to become more cautious about AI expenses in the future, which would likely hurt margins. Additionally, Nvidia’s largest clients, including Google and Amazon, are stepping up their efforts to develop their own AI chips in addition to competitors like AMD and Intel. Using its exclusive Trainium chipsets, Amazon revealed plans on Tuesday to construct an AI ultracluster, which is essentially a huge AI supercomputer. This might endanger Nvidia’s operations as well.

Intel’s Foundry Business Is Ready for a Makeover.

The AI boom has dominated the narrative surrounding Nvidia, while Intel’s foundry business has been the source of pessimism. Compared to industry leader TSMC, the company has experienced significant losses, including an operating loss of $7 billion in 2023. However, the division is poised for a potential comeback with its newest 18A process node. Significant performance and efficiency gains are anticipated with this technology, which uses RibbonFET transistors and PowerVia backside power delivery. Major corporations like Amazon, Microsoft, and the United States have already signed contracts with Intel. S. . Department of Defense for 18A-process-based custom chip designs. With this approach, Intel has reached several significant technical milestones, and the company anticipates that its first 18A designs will be put into production by external clients in 2025. The story around Intel’s foundry business may change if this transition is carried out successfully. See why 2025 Could Be Intel Stock’s Comeback Year for an in depth look at how Intel stock could be re-rated higher.

Furthermore, Trump is expected to return to the White House in 2025, and Intel’s vast U. S. The manufacturing footprint will probably become a far more valuable asset as well. The policies that Trump is focusing on increasing domestic manufacturing and decreasing dependency on foreign supply chains may benefit Intel. Potential incentives for domestic production or tariffs on chips made abroad could give Intel a competitive edge, especially in its foundry division. Additionally, Intel is the only U.S. A. -based semiconductor manufacturer that creates and produces cutting-edge chips puts it in a strong position to secure additional contracts from the federal government.

A better risk-adjusted return might be provided by Intel.

With only 23 times the consensus 2025 earnings, Intel’s stock is trading at a fair price. Due to Intel’s present difficulties, the 2025 earnings estimate is actually lower than historical levels, at only roughly $1 per share. For comparison, Intel reported earnings of more than $5 per share in 2020 and 2021, and nearly $2 per share in 2022. This implies that the stock may do the same if Intel’s earnings return to historical levels in the upcoming years. There are several tailwinds in the chip and foundry industries, and it is anticipated that the company will resume revenue growth in 2024, with consensus estimates indicating a 6% increase. The Lunar Lake and Arrow Lake chips, which power Intel’s expanding CPU lineup, put the company in a strong position for a PC and server market comeback. Intel’s Gaudi 2 and upcoming Gaudi 3 AI accelerators may also provide incremental benefits in the AI processor market.

Nvidia has yielded substantial profits, but the Trefis High Quality (HQ) Portfolio, which consists of 30 stocks, is less erratic. Additionally, it has consistently outperformed the S&P 500 during that time frame. Why? According to HQ Portfolio performance metrics, the stocks in the portfolio as a whole offered higher returns with lower risk than the benchmark index; as a result, the ride was less turbulent.

Nvidia, however, is trading at a high 48x of its anticipated FY’25 earnings. Even though Nvidia has recently experienced remarkable growth, it is unclear if the positive trend will continue. We see minimal margin for error at the current valuation. These risks could jeopardize Nvidia’s future growth and margins, which would negatively impact the company’s profits. Investors may see better risk-adjusted returns by switching from Nvidia to more undervalued semiconductor companies like Intel as the AI market begins to change. Intel may only have one option, and that is most likely up, given the aforementioned considerations. But for Nvidia, things might become a little trickier.

Invest with Market Beating Portfolios from Trefis.