Chip Stocks Haven’t Done This Since 2008
Summary
Brian explains that the AI build-out is operating on two different timelines: the immediate expansion of massive data centers and the long-term rise of edge computing. Brian characterizes data centers as the 'brain' where learning occurs, while edge devices like smartphones and autonomous vehicles act as the 'reflexes' that require local processing for speed. Brian believes the recent $2 trillion sell-off in AI stocks was driven by debt concerns that do not align with the continued high spending from major players like Meta, Amazon, and Google.
Mentioned Stocks
Reasoning: Brian labels Broadcom as 'slightly undervalued.' While Brian's system notes uncertainty regarding $71 billion in financing for a customer (likely Anthropic), Brian believes Broadcom's position in custom AI chips for Google and Meta remains strong.
Reasoning: Brian keeps Micron off Brian's valuation ladder due to extreme historical volatility in earnings. Brian notes that while memory may be entering a 10-year cycle instead of a traditional cyclical one, Brian currently lacks a valuation Brian trusts.
Reasoning: Brian finds Amazon 'slightly undervalued' ahead of its earnings report. Brian views Amazon as a key winner due to its massive AWS backlog of nearly $500 billion and its ability to monetize AI through Prime subscriptions.
Reasoning: Brian states that Nvidia is 'severely undervalued' on Brian's valuation ladder. Brian notes that the stock trades at 23 times next year's earnings, which Brian considers cheap for its growth rate. Brian argues the recent sell-off created a phenomenal buying opportunity.
Reasoning: Brian views Qualcomm as 'slightly overvalued.' Brian notes that while Qualcomm pushes edge computing heavily, phone makers are building fewer devices due to expensive memory costs.
Reasoning: Brian states that ARM flips between 'slightly overvalued' and 'fairly valued.' Brian points out that investors are paying 100 times next year's earnings, which Brian suggests means the market has already priced in the potential of edge computing.
Reasoning: Brian considers Tesla 'fairly valued' but is skeptical of Elon Musk's timelines for robotics and robo-taxis. Brian notes that 'whales' have been exiting the stock for three quarters, though Brian likes the recurring revenue from FSD subscriptions.