Tech Selloff Explained: Why Investors Are FLEEING | Sam Rahman

In this interview, Sam Rahman from Hedgei Asset Management analyzes the July 2nd market selloff where the NASDAQ declined 1.2% amid a rotation out of technology stocks into defensive sectors like healthcare and consumer staples. He explains that the primary driver is an unwind of overcrowded semiconductor and memory stock positions that benefited from the AI buildout narrative. Interestingly, Rahman notes that the "Magnificent 7" tech stocks have actually underperformed year-to-date, while semiconductor companies like Micron and Applied Materials have been the true outperformers, creating a dynamic where the beneficiaries of capex spending have become stretched relative to the spenders themselves.

Regarding Meta's announcement of cloud computing ambitions, Rahman characterizes it as "more of an excuse than a real reason" for the selloff, arguing that Meta faces existential risk from AI as engagement time shifts toward AI agents, threatening advertising revenue. He compares Meta to Oracle, suggesting the company is searching for the next transformative strategy after the metaverse proved to be a money-losing diversion. Rahman articulates that Meta lacks control over compute (relative to hyperscalers), the app layer/UI platform, and distribution channels—three critical pillars that historically define successful tech companies.

On the investment framework Rahman uses, he categorizes opportunities into three buckets: (1) companies with durable competitive advantages that must adapt to avoid disruption, (2) S-curve investment cycles like AI spanning multiple industries, and (3) special situations involving business transformations. He argues that machine learning and AI trading will expand algorithmic trading to retail participants, but that human investors retain an advantage in multi-year, long-term analysis where they can exploit corrections that machines' positive feedback loops overshoot.

Rahman expresses cautious optimism on Apple, emphasizing the iOS ecosystem's lock-in effects and high margins from services, while predicting a substantial innovation cycle from Apple Intelligence features driving iPhone upgrades and eventual dominance in home robotics. On Alphabet, he addresses search cannibalization concerns from AI overviews, arguing that the top search results become more valuable and that YouTube's human-generated content provides durable advantages against AI-created alternatives.

On the economy, Rahman downplays the significance of weak jobs data (57,000 vs. 115,000 expected), noting that employment statistics have become less valuable predictors while themes like reshoring and multi-year AI capex spending operate independently of labor market cycles. He expects bond yields to peak and decline as inflation cools, benefiting rate-sensitive sectors like housing.

Final key insight: Rahman argues that while large language models (LLMs) from Anthropic and OpenAI will become commoditized as they leapfrog each other, the real future capex opportunity lies in specialized language models (SLMs) built internally by Fortune 500 companies using proprietary data, which will drive sustained AI investment beyond the current hyperscaler focus. On the NASDAQ prediction market, he predicts the index finishes positive for the year despite near-term digestion pressure, supported by continued hyperscaler capex reiterations expected in Q2 earnings.