Nvidia, AMD, Micron fall as AI stock sell-off deepens

Published Wed, Oct 14 2026 8:42 PM EDT

In premarket trading on Thursday, shares of Nvidia, Micron, and AMD experienced declines, reflecting a broader sell-off in AI stocks that has led to a staggering reduction of over $1 trillion in market capitalizations within the week.

Nvidia saw a decrease of 1.8%, Micron dropped 4.4%, and AMD fell by 3.5%. Other notable losses included CoreWeave, which declined 5.3%, and Nebius, down by 4%. Oracle's stock was also down by 2%.

To date, 20 of the most valuable global AI and semiconductor stocks have lost over $1 trillion, driven by factors such as increasing interest rates, challenges in AI financing, and stiff competition, all of which have adversely impacted market sentiment.

The sell-off extended to technology hardware stocks in both Asia and Europe on Thursday.

In South Korea, chip manufacturer SK Hynix suffered a decline exceeding 10%, while competitor Samsung Electronics fell by 8.73%. Other South Korean semiconductor stocks also faced downturns, including LG Innotek, which dropped 8.95%, Seoul Semiconductor at a loss of 9.28%, and Samsung SDI, down by 6.26%.

Meanwhile, in Japan, SoftBank Group, a significant player in technology investment, saw a plunge of 7.9%. Semiconductor testing equipment provider Advantest experienced a dramatic drop of nearly 14%, with chipmaking equipment supplier Tokyo Electron losing 6.45%. Renesas Electronics also recorded a decline of 8%. Additionally, Taiwan's TSMC, recognized as the world’s largest contract chipmaker, fell by 1.81%.

In Europe, ASML, a leading semiconductor company in the Netherlands, saw a decline of 6.6%, while Infineon Technologies and ASM International reported losses of 9.4% and 9.6%, respectively. Nokia’s stock fell by 5%.

Why are stocks selling off?

The downturn in stock prices is attributed to growing investor concerns that increased borrowing costs may impede the extensive capital expenditures necessary for developing AI infrastructure, thereby affecting technology firms that have benefitted from the surge in AI investment.

Additionally, worries regarding a potential oversupply of global memory chips by 2029, as manufacturers increase production to meet the substantial demand driven by artificial intelligence, are dampening market sentiment.

Chinese firm ChangXin Memory Technologies (CXMT) announced a breakthrough in developing a new architecture that could significantly enhance memory density by enabling more data storage in smaller areas, indicating an intensifying competition within the memory chip market. Elon Musk also noted in a post on X that China is poised to produce "a LOT of RAM in 2028."

Furthermore, CEOs from both Broadcom and Intel cautioned on Monday that financing problems could emerge as a substantial challenge to the growth of the AI sector. Specifically, Broadcom's Hock Tan emphasized that this might prove to be a more significant obstacle to AI advancement than both chip shortages and electricity availability.

The ongoing sell-off was preceded by a tough trading day on Wall Street Wednesday, during which technology shares led market losses following unexpectedly high U.S. inflation data, raising fears about the Federal Reserve potentially increasing interest rates.

The tech-heavy Nasdaq Composite index fell by 1.88%, closing at 26,692.87, while the S&P 500 decreased by 1.11%, finishing at 7,718.94. The Dow Jones Industrial Average experienced a slight dip of 0.09%.

Dan Ives, a technology analyst at Yorkville Ives, commented that the current sell-off in AI stocks reflects apprehension regarding the financing costs of large-scale infrastructure projects amidst rising U.S. Treasury yields exceeding 5%. However, he noted that demand for AI computing power continues to be robust.

He stated, "Higher yields raise the cost of financing these builds, and investors are right to focus on balance sheets and contract structures, but rates are a financing question, not a demand question."

Ives characterized the recent pullback as a favorable buying opportunity, asserting that the infrastructure boom for AI is still in its nascent phases despite concerns regarding funding costs.

According to him, "What would change our view is a demand signal, not a rate signal," highlighting that warning signs would include reductions in hyperscaler capital expenditure plans, contract cancellations, or declines in GPU rental prices.