Tech • AI • Robotics • Game
The Market Cries Out for Help, Neocloud FUD, Strauss Zelnick Fires Back at Take-Two Slander
7/10TL;DR
A post-election market rout wiped out roughly $2.2 trillion as investors slashed exposure to AI infrastructure, chipmakers and data-center plays while shifting into Treasuries.
KEY POINTS
Tech selloff deepens
The Nasdaq fell about 5%, pushing it into correction territory after a drop of roughly 10% from recent highs. The S&P 500 slid 4% and the Dow Jones Industrial Average lost around 1,600 points, erasing more than the prior session’s post-election gains and marking one of the sharpest risk-off moves since 2025.
AI trade takes the brunt
Selling was concentrated in semiconductors, server suppliers and AI-linked data-center names as investors reassessed how a changed political backdrop could slow the pace of US buildouts. Market participants increasingly focused on the possibility of tighter permitting, higher costs and weaker returns for new capacity tied to the AI boom.
Money moves into safety
The selloff triggered a classic flight to quality. US Treasury prices rose, sending the 10-year yield to its lowest level since early September, as investors rotated out of volatile growth stocks and into government debt.
Neo-cloud and infrastructure names hammered
Several high-beta AI infrastructure stocks were hit especially hard. Cerebras was cited as down about 54% over the past month, while CoreWeave was down roughly 50%. Other names under pressure included Nebius, Core Scientific, Bloom Energy, ASML, Applied Materials, Seagate, Teradyne and Corning.
Hedge fund stress and forced selling fears
Traders speculated that liquidations were amplifying the move, with attention on funds exposed to semiconductors and AI infrastructure. Engineers Gate, Lone Pine and Walleye were among firms mentioned in market chatter, though concrete details remained limited. Walleye was said to be winding down one long-short strategy while maintaining other products.
Hyperscalers hold up better
Even as the Nasdaq sank, the largest cloud and platform companies appeared comparatively resilient, with several mega-cap tech names down closer to 2%. One explanation gaining traction is that established hyperscalers already control significant data-center capacity and power access, making them less vulnerable than newer AI infrastructure entrants that still depend on permits, financing and construction.
Core question shifts from capacity to revenue
The market debate is no longer only about whether AI demand is real, but where the next wave of monetization will come from beyond coding tools. Investors remain unconvinced that areas such as AI for science, drug discovery or chip design can turn breakthroughs into revenue quickly, especially when real-world validation, manufacturing and human trials slow adoption.
Enterprise AI spending still looks narrow
Coding remains one of the few functions where companies can justify very large AI budgets. By contrast, legal, finance and marketing use cases have not yet produced the same willingness to spend tens of thousands of dollars per employee each year, limiting how fast AI labs can expand revenue beyond developers and technical teams.
Anthropic IPO timing comes into question
The market shakeout also cast doubt on whether Anthropic would press ahead with a public offering on the timetable many investors had expected. A weaker valuation, a smaller float or a delay all became more plausible as sentiment deteriorated. Even so, awareness of the company among retail and institutional buyers remains unusually high for a software listing.
Data-center debt hits the market
In another sign of stress, a roughly $500 million portfolio of data-center loans was reportedly offered for sale through a competitive bidding process. The package included debt tied to projects connected to operators such as Vantage Data Centers and CyrusOne, as well as exposure to a site linked to the OpenAI, SoftBank and Oracle Stargate project in New Mexico.
Politics and local resistance matter more
Investors are increasingly pricing in the risk that communities and state governments demand better terms for new data centers, especially around power prices, water use, noise and local economic benefits. That could slow new AI infrastructure in some regions while potentially raising the value of already-permitted, already-powered facilities.
CONCLUSION
The selloff underscored that enthusiasm for AI remains intact but is no longer enough on its own to support every valuation. The next phase will hinge on whether infrastructure buildouts can proceed politically and whether AI companies can prove a broader path to revenue beyond coding.
Ask a question