$OIL - Wu is long oil based on Iran war dynamics, Strait of Hormuz closure, and belief that Trump's negotiating position is weak. He sees most scenarios leading to higher oil prices regardless of specific outcome.
Bearish:
$MSFT - Microsoft pulled back CapEx in Q1 due to losing exclusive OpenAI deal and disappointing corporate adoption of Copilot. Wu sees Microsoft as more conservative and facing structural issues with its AI strategy.
$MU - Wu argues memory chip pricing power is short-lived due to commoditized nature of the market, Chinese competition entering DRAM/NAND, and typical boom-bust cycles. New capacity coming online 2027-2028 will crash prices.
$QQQ - Wu is short Nasdaq based on CapEx slowdown, AI commoditization, regulatory risks, and bearish technical reversal. He sees the AI infrastructure build as predicated on continued capability improvements that are plateauing.
$GOOGL - Google's earnings are 60% dependent on advertising revenue which is highly cyclical. Wu sees limited room for market share expansion and vulnerability in a recession. Company is also considering raising $80-85B in equity to finance CapEx.
$META - Facebook pulled back CapEx in Q1 and is considering raising equity to finance future CapEx. Like Google, highly dependent on cyclical advertising revenue with limited room for market share expansion.
$AIMODELS - Wu argues large language models are commoditizing as competitors quickly catch up to breakthrough models. He sees regulatory constraints limiting distribution of frontier models like Claude Mythos, which has only reached 150 users in two months.
$DATACENTER - Wu sees AI data center CapEx as creating an optical illusion of earnings growth through income transfer to component makers. The CapEx-to-operating-income ratio has reached unsustainable 135% for hyperscalers, forcing equity raises.
$GLD - Wu states he is bearish gold as part of his overall positioning alongside being long oil and short equities.