Stock Indexes Split as Chip Stocks Fall, Oil Surges

The Nasdaq Composite fell 0.7% (to ~11:16 a.m. ET) as semiconductor losses drove broad risk-off sentiment, while the Dow fell only 0.2% and the S&P 500 was down 0.3%. In Korea, SK Hynix plunged 15% (U.S.-listed shares down 8.4%) and the KOSPI slid 5.5%, with circuit breakers triggered 35 times in 2026. At the same time, the Strait of Hormuz ceasefire collapsed/negotiations stalled, with oil up >3%, adding geopolitical pressure ahead of Q3 earnings season for major banks (JPMorgan, Bank of America, Goldman) and large corporates (Netflix, J&J, UnitedHealth).

Analysis

This is a classic factor rotation into durability and away from capex intensity: the market is effectively saying the marginal dollar of growth in semis and AI infrastructure is being repriced until management teams prove returns, not just spending plans. That makes the near-term tape more dangerous for MU and NVDA than for the broader software stack, because memory and AI hardware are the cleanest proxies for second-order capex skepticism; if banks and megacaps report acceptable demand and stable credit this week, the index-level pressure can ease quickly, but the semis need a catalyst, not just sympathy.

The geopolitical leg matters more through inflation expectations than through energy beta alone. A sustained oil shock would likely support XLE/USO and hurt the most rate-sensitive, long-duration growth names by pushing real yields and discount rates higher, while also raising the probability that corporates defer data-center builds and other discretionary infrastructure. That would favor cash-generative defensives like JNJ/UNH and, to a lesser extent, JPM if earnings show trading strength and no credit deterioration.

Contrarian view: the semiconductor selloff may already be discounting a slowdown in future AI spend before there is any hard evidence of a demand break. If earnings/guidance merely confirm that cloud and hyperscaler capex is still on schedule, the bounce could be violent because positioning has become crowded on the short side; the key falsifier is not one bad trading session, but whether July/August guidance from NVDA/MU and their customers turns capex plans lower. On the energy side, the move is only durable if shipping disruptions persist beyond a few sessions; a partial reopening of Hormuz would unwind the macro scare fast.

Key Decisions for Investors

• Trade the dispersion: short MU or NVDA against long JNJ/UNH for 2-6 weeks. The setup favors a quality/value hedge if earnings guide capex lower, with semis carrying the more fragile multiple and defensives offering lower event risk.
• Buy a short-dated put spread on MU into the next catalyst window; use a tight premium budget and target a 2:1 to 3:1 payoff if memory pricing and guidance weaken. Falsifier: stable pricing commentary or an earnings beat with raised capex visibility.
• Initiate a tactical long in XLE or USO for the next 1-3 weeks only if shipping disruptions remain unresolved. Treat this as a headline-driven trade, not a structural thesis; cut if Hormuz traffic normalizes or Brent gives back the spike.
• Stay patient on JPM/BAC/GS until bank results confirm trading revenue benefits without credit stress. If earnings show stable reserves and stronger markets revenue, these names can outperform the tape over the next month even in a risk-off backdrop.
• Avoid bottom-fishing NVDA/MU until post-earnings guidance or customer capex commentary gives hard evidence that AI spend is intact. The contrarian long only works if hyperscaler demand is reaffirmed; otherwise the multiple de-rating can extend 1-3 months.