AI-driven momentum trading faces challenges as Goldman Sachs reports hedge funds have recorded record sales of U.S. tech stocks over the past two months.

Goldman Sachs noted that hedge funds have been net sellers of the technology sector in six out of the past eight weeks. Cumulatively, the market value of their holdings in this sector has declined by approximately 10%, marking the largest reduction in over a decade of data tracking. Goldman Sachs analysis indicates that panic selling has emerged in the technology sector; although fundamentals related to AI infrastructure remain solid, the sector will continue to face near-term challenges.

wallstreetcn28minutes ago

Moonshot AI Ignites a Shift in AI-Driven Trading: Assessing Winners and Losers in the New Landscape

China's semiconductor ecosystem, memory chip manufacturers, AI agents, and software developers are all winners, while AI model developers and high-end chipmakers are viewed as losers.

wallstreetcn04:33

Ranked ninth globally and first among open-source models, Kimi K3 immediately faced a 'circuit breaker' in computing power: its rise in rankings has instead confirmed the underlying logic of hardware demand.

Forty-eight hours after the launch of Kimi K3, Moonshot AI announced a shortage of computing capacity and suspended new C-end user subscriptions. Citi and Bank of America separately noted that the release of the K3 would not weaken the investment rationale for hardware; on the contrary, it could drive higher computing power consumption. China Merchants Securities pointed out that the performance of individual chips alone can no longer determine system capability—factors such as efficient inter-chip communication and unified memory resources are increasingly becoming critical determinants of computational efficiency.

wallstreetcn04:10

Goldman Sachs Trading Desk: Momentum trading may take several more weeks to bottom out; the AI capex narrative is faltering.

Goldman Sachs noted in its report that AI models are achieving high performance with less computational power than expected, challenging the prevailing market narrative that sustained increases in capital expenditure are necessary to win. Momentum-driven trading adjustments have not yet concluded, but U.S. equities show no signs of systemic risk, with capital rotating across sectors. The efficiency of these models has raised questions about the investment logic on the training side, though demand on the inference side remains robust. The upcoming earnings season will serve as a critical window to test the underlying assumptions of AI-related capital spending.

wallstreetcn00:13

Fund Flows | Southbound capital sold HK$6 billion worth of Hong Kong-listed stocks and has increased its position in GigaDevice for 11 consecutive days.

Track the latest developments of southbound capital flows.

Gelonghui FinanceJul 20 20:05

Is the recent pullback in chip stocks merely temporary? JPMorgan: Q2 earnings season will sound the starting gun for the next rally!

① JPMorgan believes the recent decline in semiconductor stocks does not mark the beginning of a long-term downturn, but rather a consolidation phase ahead of the next upward move; ② The bank notes that semiconductor stock prices have increasingly diverged from fundamentals, and technical indicators also show an 'oversold' condition, recommending accumulation over the summer; ③ It forecasts that AI-driven DRAM supply-demand tightness will persist through 2028, with the Q2 earnings season serving as the catalyst for the next rally.

cls.cnJul 20 18:34