The July 13 selloff began with oil and ended in the market’s most crowded positions.
In China, the Shanghai Composite fell 2.1%, Shenzhen Component 3.5%, ChiNext 3.1%, and STAR 50 3.4%. Shanghai and Shenzhen turnover totaled about RMB 2.82 trillion, with roughly 4,400 stocks declining across the two exchanges (source: Eastmoney market data, 3:05 p.m. CST on July 13). Semiconductors lost 5.6% and defense 7.6%. Elsewhere, Nikkei 225 fell 1.9%. Korea’s KOSPI closed down 9.0% after triggering a circuit breaker, while SK hynix dropped 15.4% and Samsung Electronics 10.7% (sources: AP, KRX, and Yonhap, July 13 close).
Oil Became a Risk Variable Again
U.S. and Iranian forces resumed missile and drone attacks after the weekend, while Iran said the Strait of Hormuz had been closed again. WTI crude moved above $74 a barrel during Asian trading, at one point rising more than 3% (sources: Reuters and AP, July 13 Asian session).
This was not a simple haven rotation. Higher oil raised both inflation and corporate cost concerns, while the dollar and Treasury yields also firmed. Risk assets therefore faced pressure on discount rates and margins at the same time. Energy-importing markets such as Japan and Korea were natural early casualties.
Crowding Determined the Depth of the Fall
Geopolitics explains the opening direction, but not the distribution of losses. If the session were only a war-risk trade, defense and rare-earth shares should have held up better. Instead, Chinese defense underperformed the broad market and rare earths weakened sharply in the morning, while banks, energy, and consumer staples were relatively resilient.
The cleaner explanation is that an external shock triggered profit-taking already waiting to happen. AI hardware, memory, semiconductor equipment, and thematic small caps had carried some of the market’s largest prior gains and the heaviest positioning. Through early July, global chip stocks repeatedly fell even after strong operating updates as the market began questioning how quickly AI capital expenditure and memory pricing could keep growing. On July 13, Korea’s two memory leaders fell far more than KOSPI, while Chinese semiconductors fell much more than CSI 300. That pattern looks more like concentrated exposure being repriced than new evidence of a sudden deterioration in industry fundamentals.
How This Reading Could Be Wrong
One session cannot prove a complete causal chain. If oil quickly reverses but semiconductors and high-beta small caps continue to underperform, valuation and crowding are probably doing more of the work. If oil stays high, energy and banks retain their relative advantage, and broad risk premia rise together, the geopolitical and inflation shock deserves more weight.
The most informative detail was not the index decline itself. Even supposed war beneficiaries failed to resist the selling. Oil lit the match; positioning supplied the fuel.
Sources: Eastmoney market data, KRX, Yonhap, Reuters, and AP. The data window is the July 13, 2026 close for Asian equity markets; WTI is an Asian-session snapshot. Coverage is limited to major Chinese, Japanese, and Korean indexes, selected sectors, and representative semiconductor companies, not a full-market scan. The causal interpretation is a research judgment based on price distribution and contemporaneous reporting, not a proven single cause. This piece is a personal observation and does not constitute investment advice.