Global Chip and AI Stocks Suffer Sharp Three-Session Selloff

Semiconductor and artificial intelligence stocks suffered a sharp, broad-based selloff across global markets over three consecutive sessions through 15 October 2026, with the Philadelphia Semiconductor Index (INDEX:SOX) and the Nasdaq Composite (INDEX:IXIC) bearing the brunt of a wave of selling that spread from Asia to Europe and the United States. The rout was triggered by a confluence of competitive threats, sobering executive warnings and a reassessment of whether the AI infrastructure boom can sustain its pace of investment.

The selloff matters because semiconductors have been the backbone of the AI trade that powered equity markets for the better part of two years. A sustained retreat in chip stocks risks unwinding a key pillar of the bull market, with the AI hyperscalers — Microsoft, Alphabet, Amazon, Meta Platforms and Oracle — expected to deploy around $795 billion in capital expenditure this year alone, according to industry estimates.

What happened

The three-session decline accelerated on 13 October 2026, when Asian and European markets were roiled by a combination of competitive and macro shocks. South Korea's benchmark index crashed more than 7% and Japan's Nikkei plunged 4%, with both moves driven by a deepened tech sell-off. Memory-linked stocks fell sharply after China's ChangXin Memory Technologies (CXMT) reported meaningful progress on a new DRAM architecture designed to boost chip memory density, raising the prospect of intensified competition for established players including Samsung.

The competitive alarm from CXMT compounded existing anxiety in the sector. Intel and Broadcom's chief executives had already warned that access to capital could emerge as a major headwind to the artificial intelligence boom. Broadcom's chief executive went further, likening AI investing to a "casino" — a remark that rattled sentiment across chip equities on 13 October 2026.

• South Korea's equity benchmark fell more than 7% on 13 October 2026, its steepest single-session drop in the current selloff.
• Japan's Nikkei plunged 4% on 13 October 2026 as the tech sell-off deepened across Asia.
• Memory-linked stocks fell for a second consecutive day on 13 October 2026 following CXMT's chip architecture announcement.
• Broadcom's CEO compared AI investing to a "casino," amplifying the risk-off mood in semiconductor equities.
• Intel and Broadcom CEOs both flagged capital access as a potential major headwind to the AI investment cycle.

What drove the selling?

Three distinct forces converged to produce the three-session decline. First, the competitive threat from China's CXMT unsettled memory-chip investors who had assumed that Western and South Korean manufacturers would maintain a durable technology lead. Second, executive commentary from the chief executives of two of the sector's largest companies cast doubt on the sustainability of AI capital spending. Third, investors have grown increasingly nervous about whether the pace of AI infrastructure buildout can be maintained without a slowdown in demand.

The concern over AI spending is not new, but it has sharpened. Infrastructure and energy companies invested in the AI buildout face a potential slowdown, and any significant delay in data-centre construction timelines carries material financial consequences for the supply chain. Shares of industrial companies tied to the AI buildout, including GE Vernova and Caterpillar, had already dropped in mid-September 2026, alongside neocloud CoreWeave and AI infrastructure provider Oracle, as investors began pricing in the risk of a deceleration.

Calls to slow the development of powerful AI models have also fuelled doubts over whether the sector's promised profits can justify its staggering costs. Anthropic's chief executive Dario Amodei warned in mid-September 2026 that a swarm of rogue AI agents could be capable of taking over the internet within six to twelve months — a statement that added a regulatory and reputational dimension to the investment calculus.

How did markets react?

The Nasdaq Composite (INDEX:IXIC) tracked the Philadelphia Semiconductor Index (INDEX:SOX) lower across all three sessions, reflecting the outsized weight of chip and AI-adjacent names in the index. Indian equity markets were also bracing for a weak opening on 15 October 2026, with analysts citing the Korean and Japanese crashes as the primary external driver.

Nvidia (NVDA), the bellwether of the AI chip trade, entered the selloff with a trailing price-to-earnings ratio of 28.74 and a forward price-to-earnings ratio of 24.41, based on its latest reported quarter in which earnings per share came in at $2.22 against an estimate of $2.09, and revenue reached $96.22 billion against an estimate of $92.27 billion.

Despite those beats, the valuation reset underway across the sector has not spared Nvidia. Advanced Micro Devices (AMD) also fell, with the AMD logo appearing prominently in market coverage of the chip rout as a proxy for the sector's broader pain.

ASML Holding N.V. (ASML), the Dutch maker of extreme-ultraviolet lithography machines whose equipment is essential to advanced chip production, has previously rattled the sector with a weak demand outlook for non-AI chip segments — a concern that resurfaced during the current selloff as investors questioned whether the AI-driven portion of demand is large enough to offset weakness elsewhere in the semiconductor cycle.

Why it matters

Chip stocks had been widely regarded as a relatively safe expression of the AI theme, offering exposure to the infrastructure layer of the buildout rather than the more speculative application layer. The current selloff challenges that thesis. As one market analysis noted, chip stocks "were a safe AI play" that have now "turned into the market's pain trade" — a shift that reflects both the valuation reset and the emergence of credible competitive threats from Chinese manufacturers.

The macroeconomic stakes are also significant. Global AI spending was projected to reach $375 billion in 2025 and $500 billion by 2026, according to industry estimates, with economists warning that without AI-driven investment, the United States economy could face a sharper slowdown. The AI boom has been described as masking underlying economic weakness, meaning a sustained deceleration in chip and AI capital expenditure would have consequences well beyond equity markets.

The Broadcom CEO's "casino" remark crystallised a fear that has been building since mid-September 2026: that the capital flowing into AI infrastructure may be running ahead of demonstrable returns, and that a correction in sentiment could arrive faster than the technology's commercial payoff.

What to watch next

Several catalysts in the coming weeks will determine whether the three-session selloff marks a temporary correction or the beginning of a more sustained repricing of the semiconductor and AI sector.

1. Earnings from major chip companies: Results from bellwether names including Nvidia (NVDA) and Advanced Micro Devices (AMD) will be scrutinised for any revision to demand guidance, particularly for AI accelerators and memory chips.
2. ASML Holding N.V. (ASML) order data: ASML's booking figures for advanced lithography equipment will serve as a leading indicator of whether AI-driven chip demand is holding up or softening.
3. Hyperscaler capital expenditure guidance: Quarterly updates from Microsoft, Alphabet, Amazon, Meta Platforms and Oracle on their data-centre spending plans — collectively expected to total around $795 billion this year — will be the most direct test of whether the AI infrastructure cycle is intact.
4. CXMT competitive developments: Further disclosures from China's ChangXin Memory Technologies on its new DRAM architecture will be watched for signs of how quickly Chinese manufacturers can close the technology gap with Samsung and Micron.
5. Regulatory and policy signals: Any government action in the United States or Europe on AI development constraints or export controls on chip technology could amplify or dampen the current risk-off mood.