Wedbush says the recent sell-off in AI leaders such as Alphabet reflects impatience, not a breakdown, estimating $700 billion in capex this year and forecasting compute and memory costs will ease. Separately, Oracle secured $67 billion in AI infrastructure deals last quarter, lifting its backlog to $638 billion despite a 21% stock decline.
Wedbush technology analysts describe the recent pullback in major AI stocks, including Alphabet, as short-term impatience rather than a structural shift, framing it as an early buying opportunity in a decade-long AI buildout. They highlight that core AI leaders have been treated like bear-market stocks despite strong long‐term fundamentals.
The team estimates roughly $700 billion in capital expenditure across hyperscalers and memory chipmakers this year, placing the sector in an “air pocket” stage between heavy data-center buildouts and revenue recognition. Rising compute and memory costs have sparked concern, but analysts project those expenses will begin to decline over the next year.
Oracle reported $67 billion in AI infrastructure contracts in its latest quarter, swelling its remaining performance obligations to $638 billion. Cloud Infrastructure revenue jumped 93%, yet the stock has slid 21% over the past month as investors weigh execution risk against a premium 27.9 P/E ratio and high price-to-sales multiple.
