Macro factors like higher energy prices and Treasury yields are dampening risk appetite and could curb high-capital expenditure by enterprises and cloud providers.

China will tighten export oversight on 40 key technologies spanning artificial intelligence, photovoltaics, and biomedicine, with immediate registration and future bans or licensing for subsets of these technologies.

In AI hardware, global shipments show cautious momentum: H1 accelerator shipments totaled about 2.008 million units, with GPU shipments down about 26% to 991,000 units while non-GPU accelerators rose nearly 80% to 1.016 million.

Beijing’s push for chip self-sufficiency is contributing to market shifts and appears to be eroding Nvidia’s dominance in China amid policy and domestic ecosystem momentum.

Analysts note investors care more about earnings per share than free cash flow, with high depreciation from data-center investments potentially pressuring EPS unless profits rise.

Some analysts see value in the sell-off, with Goldman Sachs and BNP Paribas flagging selective buying opportunities while others stay cautious awaiting clearer valuations.

Despite volatility, there is a view that the pullback may be a prolonged adjustment rather than a fundamental collapse, with long-term reforms and returns remaining a focus.

Strategic investors and research notes suggest selective buying amid reforms aimed at improving capital market stability, even as wider markets pull back.

Company results show strong profitability momentum despite heavy capital expenditure in AI and AWS infrastructure, indicating AI investments may eventually lift earnings power.

AWS backlog stands at $578 billion, while trailing 12-month free cash flow is negative $42.4 billion, raising questions about the speed at which spend translates into cash generation and shareholder returns.

Anthropic’s multi-trillion market valuation signals a structural shift in capital allocation toward physical computing infrastructure, with ongoing top-line growth but rising long-term cash needs for AI development.

Barclays cautions that AI financing may be entering a pre-crisis-like phase as data-center builders borrow heavily while cash-rich customers like Alphabet and Amazon continue to spend.