Gareth Soloway argues that the AI trade is beginning to unwind after a sharp overnight selloff in Korea, Japan, and U.S. tech, and he warns this could turn into a larger tech correction over the next 3–6 months. He backs the call with chart-based evidence: negative RSI divergences, lower highs, crowded positioning, and a growing belief that cheaper Chinese AI models could commoditize margins.
Watch on YouTube ›Get the market thesis, key claims, assets, contradictions, and follow-up questions from any financial video — then unlock a version personalized to your portfolio, watchlist, and favorite speakers.
Gareth Soloway frames the session as the start of a potentially bigger unwind in AI and mega-cap tech after a violent overnight selloff. He says South Korea’s KOSPI dropped 10% because two AI-chip makers make up more than half the index, Japan’s Nikkei fell over 5%, and the Nasdaq was down as much as 3% premarket. His core thesis is that the AI trade has become too concentrated and too narrative-driven, and that the first signs of a larger correction are now visible on the charts and in the news flow. He repeatedly returns to technical analysis as the anchor for his view. On the S&P 500 and Nasdaq, he points to lower highs, lower lows, and a possible break in trend that would confirm downside continuation. …
Immediate setup is fragile: if tech cannot reclaim the opening weakness, the current move looks like the first leg of a crowded AI unwind. A bounce is possible, but failure near the lows would keep the downside pressure in place.
Over the next few weeks, the market likely rotates from crowded AI leaders into beaten-down large caps only if earnings and index breadth stabilize. If cheaper AI alternatives keep gaining traction and the leaders keep breaking support, this becomes a broader tech correction rather than a one-day shock.
Structurally, he is arguing that AI is moving from scarcity and pricing power toward commoditization, which would compress margins and reduce the durability of the current hype cycle. The longer-term regime implication is that the most crowded beneficiaries may underperform even if AI adoption continues.
The AI trade is more concentrated than the dot-com bubble and an unwind is inevitable — when hedge funds all pull the rip cord, retail usually gets stuck.
Speaker compares concentration in AI stocks to dot-com bubble as a historical analogy, warning that hedge fund positioning will unwind eventually, leaving retail holding the bag.
AI models from China are so much cheaper to run than US models that many companies globally are switching away from ChatGPT, Claude, and Gemini, which will commoditize AI and crush the margins of hyperscalers.
Speaker argues that cheaper Chinese AI models (DeepSeek, Quen, Kimmy) are gaining adoption outside the US, which will commoditize AI, compress margins for hyperscalers who spent billions on data centers, and trigger a crash in tech stocks.
Nvidia is forming a bear flag pattern plus a double-top head-and-shoulders pattern, and a breakdown would confirm further downside.
Speaker points to a sequence of lower highs and lower lows on Nvidia's chart forming a bear flag, alongside a double top/head and shoulders pattern with a neckline that if broken would signal more downside.
Unlock the full claims, asset map, scores, related transcripts, follow-up questions, and AI chat — shaped around your portfolio, watchlist, favorite speakers, and risks.