Tired of ads? Enjoy an ad-free experience by signing up.
👩‍🍳 How we use AI at Tech in Asia, thoughtfully and responsibly.
🧔‍♂️ A friendly human may check it before it goes live. More news here

SoftBank sells $5.8b Nvidia stake, sparking AI bubble concerns

SoftBank Group has sold its entire stake of 32.1 million Nvidia shares, raising US$5.8 billion to fund its AI projects, according to its latest quarterly results.

Market analysts noted that the timing of SoftBank’s Nvidia sale comes amid broader warnings from Wall Street leaders about possible overvaluation in AI stocks.

SoftBank’s rising focus on OpenAI has also brought more attention to the company’s exposure to fluctuations in the AI sector.

The Japanese tech investor made the sale in October to support CEO Masayoshi Son’s AI initiatives, including the Stargate project to expand US data center capacity and a reported multibillion-dollar commitment to OpenAI.

Nvidia shares fell over 2% following the news, contributing to a drop in the S&P 500 index, while concerns about the sustainability of AI-related valuations increased.

SoftBank also sold roughly US$9.2 billion in T-Mobile shares, further strengthening its cash reserves for future AI investments.

🔗 Source: Reuters

🧠 Food for thought

Implications, context, and why it matters.

SoftBank’s Nvidia exit tracks broad institutional shifts in AI stocks

  • SoftBank sold its entire 32.1 million-share stake in October. Other investors took different paths, as Crestwood Advisors added 189,084 shares in Q3 and Primoris Wealth Advisors lifted holdings by 7.6% in Q2 12.
  • In the latest quarter, 2,788 funds bought Nvidia while 2,233 cut exposure, and Compound Global Advisors slashed its stake by 94.8% by selling 11,592 shares 13.
  • Trades came as Nvidia hovered near a 52-week high of $212.19 after a run from $86.62. A P/E of 56.71 and a market value above $4.8 trillion suggest some profit taking at high prices, not a shift in AI demand 2.

Tech vendors can tap Stargate, SoftBank’s planned network of large U.S. data centers

  • Five sites are slated for Shackelford County TX and Doña Ana County NM 45. Lordstown OH, Milam County TX, and one Midwest location are also planned 45. Capacity tops 5.5 gigawatts, with two sites able to reach 1.5 gigawatts within 18 months 45.
  • Vendors can move early. Lordstown is under construction and targets 2026 service 4. Milam County offers a fast-build design with powered infrastructure from SB Energy, a utility-scale renewable power developer 4. These builds need heavy fiber, as the Abilene site used 6,400 workers 6.
  • OpenAI is evaluating 16 states, including Pennsylvania, Oregon, and Wisconsin 7. Site visits are planned 7. Construction equipment makers, power providers, and data center cooling vendors should track rolling updates, since each 1-gigawatt campus costs $50-60 billion 5.

Recent SoftBank developments

Stay ahead in Asia’s tech landscape

You've reached your 2 free content limit for the month. Sign up for free to read the full story.

🏄 For casual readers / 👶 Free

Basic

US$0

Free forever

Get instant access to this article and more every month

0 premium content

Unlimited news briefs

5

5 articles

Ad-free reading experience

Just US$0 per day

⌛Sign up in 20s. No payment details needed.

📖 For learners / 👍 Starter

Lite

US$4.92/month

Billed annually at US$59/year

Get instant access to this article and more every month

4

4 premium content

Unlimited news briefs & articles

Ad-free reading experience

Just US$0.17 per day

Cancel anytime

Our subscriber community includes professionals from these companies:

Stay updated on the go with our mobile app.

Get latest insights with smoother, more personalized experience through TIA mobile app.