Geo Chen
· Fidenza Macro
· July 30, 2026 at 11:29
· ⏱ 12 min read
| Read on Substack ↗
Summary
The author argues the AI-driven bull market has peaked due to extreme Korean retail leverage unwinding, commoditization of AI models from Chinese open-source labs, a looming compute glut, stagflation from the Iran war, and a Fed losing credibility. This convergence of factors points to a prolonged bear market in equities, with AI semi and infrastructure names particularly vulnerable.
•Korean retail margin calls wiped out $38.7B (Citibank estimate) and hit 1.2 million accounts, one in every 30 Korean adults.
•Chinese AI lab Moonshot released Kimi K3, which outperforms Claude Fable in Code Arena; Alibaba followed with Qwen 3.8, a 2.4T open-weight model.
•OpenAI delayed its IPO due to valuation concerns; SpaceX's secondary selloff at 27% below IPO price is denting IPO prospects.
•Google announced a $205B AI capex plan for 2026, sending its stock down 7%—the market no longer rewards hyperscaler capex increases.
•The war in Iran has cost US taxpayers $37.5B (likely underestimate) and risks becoming a protracted quagmire with stagflationary effects via energy supply and government borrowing costs.
•30-year Treasury yields broke above 5.2%, signaling a new phase in the Treasury bear market and a headwind for equities.
Read time12 min
Length12,705 chars
Categoryfinance
Ideas
Geo ChenGlobal macro trader; ex-head of FX trading, Credit Suisse
Author notes Google's 7% stock drop after announcing higher AI capex ($195B to $205B), and argues the market is not rewarding hyperscalers for capex increases—indicating deteriorating sentiment and po
Author notes Google's 7% stock drop after announcing higher AI capex ($195B to $205B), and argues the market is not rewarding hyperscalers for capex increases—indicating deteriorating sentiment and potential overinvestment.
Risk: Further capex increases could lead to margin compression and debt market stress; compute glut scenario would hit Google's cloud margins.