Woofun AI reports that Goldman Sachs strategist indicates Wall Street's AI trading has entered a complex phase where the market distinguishes between valuation frameworks for different AI companies. Investors continue to support direct beneficiaries of AI infrastructure, such as Nvidia, TSMC, and semiconductor suppliers, driven by sustained capital expenditures from Amazon, Alphabet, Meta, and Microsoft.

However, these hyperscalers have not seen equivalent stock price performance, leading to caution regarding their ability to convert billions in AI investment into profit and shareholder returns.

Goldman Sachs describes this dynamic as a "stretched rubber band," with hardware suppliers seeing raised profit expectations while tech platforms face capital expenditure pressure. The firm asserts the AI trade has moved from thematic investing to a return validation stage, focusing on who can generate real cash flow. Risks for hardware include demand growth failing to exceed expectations, while hyperscalers face short-term pressure from high spending. A key variable remains the AI cost curve, where improved model efficiency or lower-cost development by regions like China or Japan could challenge the high capital expenditure logic of U.S. tech giants.