AI spending fears overshadow record Anthropic IPO as Asian stocks plunge

  • Nearly 5% was wiped off Japan’s Nikkei 225 and South Korea’s KOSPI. Major chipmakers and technology shares led the sell-off.
  • With Japanese inflation accelerating, expectations of a Bank of Japan rate hike strengthened, sending the Nikkei to its lowest since April.
  • Disappointing technology earnings, concerns over US exports and limited geopolitical summit outcomes weighed on Hong Kong and mainland Chinese indices.

Nearly 5% declines hit Japan’s Nikkei 225 and South Korea’s KOSPI on Friday. Across Asia, equities fell sharply following Wall Street’s overnight sell-off. Persistent questions about the profitability and sustainability of vast artificial intelligence investments overshadowed excitement over Anthropic’s record-breaking stock market debut. Technology heavyweights suffered the largest losses, pulling the region’s main benchmarks down.

TMGM Analysis: Financial Market News, Economic Calendar & Market Insights

At around 56,630, Japan’s Nikkei 225 was down nearly 5%, taking its weekly decline beyond 7% and marking its lowest level since April. Kioxia Holdings, Advantest and SoftBank Group were among the technology and AI-linked shares recording steep falls. Inflation data added pressure: Japan’s annual rate rose from 2.1% in September to 2.3% in October, its highest since November 2025. The acceleration reinforced expectations of further Bank of Japan interest rate increases.

A nearly 5% slide took South Korea’s benchmark KOSPI to roughly 4,890, wiping out the previous session’s advance. Samsung Electronics and SK Hynix, the country’s leading chipmakers, drove the decline. Despite Anthropic’s strong market debut, investors remained concerned about AI-related capital expenditure.

Substantial closing losses also extended to mainland Chinese and Hong Kong shares. Broad selling in technology stocks pushed Hong Kong’s Hang Seng Index down 1.7% to around 22,450, its lowest since May 2025. Disappointing third-quarter results from Tencent, NetEase and Kuaishou weighed on sentiment. Tencent also warned that computing constraints and tighter US semiconductor export restrictions could hold up training for next-generation AI models.

Mainland benchmarks retreated, with the Shanghai Composite losing 1.57% to 3,520 and the Shenzhen Component declining 2.05% to 10,370. A trilateral meeting of Presidents Xi Jinping, Donald Trump and Vladimir Putin offered investors few catalysts. Beijing and Moscow reported broad agreement across four areas: United Nations reform, refined petroleum supply, International Criminal Court overreach and nuclear risk reduction. However, according to a Kremlin aide, separate bilateral talks between Trump and Putin on the war in Ukraine delivered no breakthroughs.

APEC receptions signal changing diplomatic hierarchy

Contrasting receptions at APEC illustrated shifting geopolitical alignments, according to Rabobank analysts. They noted that Donald Trump was “greeted by Xi Jinping with a hug and pom-poms” whereas Vladimir Putin “was met by a lower-level, taller Chinese official, which looked a deliberate diplomatic snub.” The analysts viewed the US delegation’s warm welcome and the more restrained reception for Russia’s president as reinforcing the summit’s function as a stage for changes in the global hierarchy, with possible implications for markets.

AI stocks: key questions answered

As an academic discipline, artificial intelligence aims to reproduce human cognitive functions, logical understanding, perception and pattern recognition in machines. Commonly shortened to AI, it encompasses sub-fields such as artificial neural networks, machine learning or predictive analytics, symbolic reasoning, deep learning, natural language processing, speech recognition, image recognition and expert systems. Artificial general intelligence, or AGI, is the field’s ultimate objective: a machine capable of solving arbitrary problems without having been trained to solve them.

Artificial intelligence has a broad range of applications. Among the best known are generative AI platforms, including ChatGPT and Google’s Bard platform, which draw on training with large language models (LLMs) to respond to text-based questions. User-written text also provides the basis for Midjourney to create original images. Elsewhere, probabilistic techniques help AI assess an entity’s qualities or how it is perceived. Upstart’s lending platform, for example, applies an AI-enhanced credit rating system to assess applicants’ creditworthiness, searching the internet for information on their careers, wealth profiles and relationships. Scientific research databases support other AI systems in proposing potential new pharmaceuticals for laboratory testing. By collecting and organising information about viewing habits, YouTube, Spotify, Facebook and other content aggregators also use AI applications to recommend personalised content.

Semiconductor company Nvidia (NVDA) produces AI-focused computer chips alongside some of the platforms engineers use to develop AI applications. Because it supplies the tools for further AI development, many supporters regard Nvidia as a picks-and-shovels investment in the AI revolution. Another participant is Palantir Technologies (PLTR), a “big data” analytics business with substantial US intelligence community contracts. Its Gotham platform enables intelligence users to sift through data, identify leads and support pattern recognition. Major corporations use its Foundry product to monitor employee and customer data, supporting predictive analytics and anomaly detection. Microsoft (MSFT), meanwhile, holds a substantial stake in OpenAI, the privately held creator of ChatGPT, and has incorporated OpenAI’s technology into its Bing search engine.

Well over 200% was added to Nvidia’s share price in the six months after ChatGPT’s public release in late 2022. It was one of many AI-related stocks to rally, prompting immediate questions among Wall Street commentators about another technology bubble. Prominent investor Stanley Druckenmiller, who has held substantial investments in both Palantir and Nvidia, argued that bubbles never last just six months. Were AI enthusiasm to develop into a bubble, he said, extreme valuations would persist for at least two and a half years or longer, as they did during the DotCom bubble of the late 1990s. At the midpoint of 2023, the best assessment was that equities were not in a bubble, at least for the time being. Although Nvidia was then valued at 27 times forward sales, analysts projected exceptionally strong revenue growth for years ahead. For comparison, the NASDAQ 100 traded at 60 times earnings at the DotCom bubble’s peak, against 25 times earnings in mid-2023.