Politics

AI Stocks Dip as US Considers New Chip Export Restrictions

AI Stocks Dip as US Considers New Chip Export Restrictions

Since early last year, investors have eagerly embraced the potential of artificial intelligence (AI), investing heavily in shares of companies strategically positioned to capitalize on this next-generation technology. However, as the bull market reaches its two-year milestone, many investors are pausing to reassess the market landscape, and some are seizing any opportunity to secure profits. Against this backdrop, a notable decline in AI stocks was observed on Tuesday. Chip designer Arm Holdings experienced a 6.7% drop, AI chipmaker Nvidia saw a 4.9% decrease, chipmaker Advanced Micro Devices (AMD) dropped by 4.8%, semiconductor device supplier Broadcom fell by 3.7%, and chip foundry Taiwan Semiconductor Manufacturing (TSMC) dipped by 2.6% as of 12:50 p.m. ET. The catalyst behind this decline in AI stocks stemmed from reports that the U.S. government is considering implementing new restrictions on chip exports.

Rumors are swirling that the Biden administration is contemplating curbing sales of advanced AI processors from companies like Nvidia and AMD, as initially reported by Bloomberg citing sources familiar with the matter. This move would represent the latest regulatory step to address concerns that advanced technologies such as AI could be leveraged against the United States and its allies. The government is reportedly exploring a cap on the number of export licenses for specific countries, with national security cited as the primary reason for these potential restrictions. It is important to highlight that the U.S. already imposes strict limitations on the level of AI chip technology permitted for sale to certain countries, including China and 40 other nations across Asia, the Middle East, and Africa. Currently, U.S. chipmakers must secure government licenses to sell advanced semiconductors to clients in these regions.

The ongoing discussions extend the existing restrictions, potentially applying on a country-by-country basis, with a focus on countries within the Persian Gulf region. Although these considerations are still in the nascent stages and no definitive decision has been made, the plans have reportedly been gaining traction in recent weeks. The potential ramifications of limiting the sale of advanced AI chips to certain countries are significant for AI-centric stocks. Nvidia, the dominant provider of graphics processing units (GPUs) that power AI systems, stands to lose the most due to its control over as much as 98% of the data center GPU market, according to TechInsights. AMD, which has been fiercely competing with Nvidia for supremacy in the GPU sector, recently shifted its focus to prioritize AI processors, placing its traditional gaming chips in a secondary role.

Imposing restrictions on advanced processors could undermine these ambitions. Arm Holdings, known for providing intellectual property and chip designs for some of the most advanced chips used by companies like Nvidia and AMD, might see its revenue impacted if processor sales are curtailed. Broadcom, offering a range of products that complement GPUs in data centers, such as Ethernet switching and application-specific integrated circuits (ASICs), could also experience a decline in sales if GPU sales falter. TSMC, the world's leading foundry responsible for 62% of global semiconductor production and approximately 90% of the advanced processors used for AI, would likely see its revenue affected by any limitations on processor sales. While investors are concerned about the potential impact on Nvidia's and others' sales, historical trends suggest that they may be overestimating the threat. Similar apprehensions arose on several occasions when the U.S. government contemplated or announced restrictions on chip exports to countries like China.

Despite these fears, Nvidia experienced triple-digit growth for five consecutive quarters. Moreover, recent reports indicate that the company's Blackwell chips are sold out for the next year, indicating that, despite potential export curbs, demand for AI chips remains strong. Additionally, current valuations deserve consideration. Arm Holdings, AMD, Nvidia, Broadcom, and TSMC are trading at forward earnings multiples of 96, 46, 46, 36, and 28, respectively. From a value perspective, TSMC presents a more appealing buy, yet this analysis does not account for the accelerated growth trajectory propelled by AI. Utilizing the forward price/earnings-to-growth (PEG) ratio, which incorporates growth factors, reveals that each remaining stock holds a PEG ratio of less than 1, indicating they are undervalued. As the adoption of generative AI is still in its infancy, estimates of the market's potential value range from $1.3 trillion to much higher figures. For those seeking to profit from AI, the optimal strategy involves acquiring the most promising AI stocks and maintaining a long-term investment perspective.