Blue Sky Stock Volatility Report

Notable Movers & Volatility

The tape was dominated by the memory and storage complex. SNDK led all names at +83.13% ($998.19 → $1827.99), with STX (+45.07%), SKHY (+42.97%) and MU (+40.42%) close behind, while WDC ran the other way at -29.75% ($580.00 → $407.48) — an unusually wide intra-sector dispersion.

Realized volatility exceeded implied essentially everywhere, so options were systematically under-pricing movement. AMZN was the biggest surprise at 7.51x, followed by MSFT (5.81x) and AAPL (5.99x) — notably the lowest-IV names in the set (27.5%, 28.0%, 23.8%). Only NVDA (1.67x) came close to fair.

Range placement diverged sharply from headline moves. TSLA closed at 94.76% of its range and MU at 76.73%, while CSCO closed at 11.69%, AVGO at 14.43% and AAPL at 21.14%. Beware conflating range with return: STX swung 45% yet finished -0.15%, INTC +31.52% range but -2.60%, and AMD +25.02% range against a -10.38% return — round-trip action rather than trend.

Broader Market Context

The pattern here — violent two-way movement concentrated in memory, storage and AI hardware, with mega-cap software and platforms quietly realizing double their implied vol — is consistent with a market re-pricing the *durability* of AI capex rather than abandoning it outright.

Plausible contributing threads:

• Cost-curve competition in memory. Reports of aggressive DRAM/NAND capacity ramps from lower-cost Asian suppliers would likely pressure pricing assumptions embedded in MU, SNDK, WDC and STX, and could explain why the same sub-sector produced both the top gainer and the top decliner.
• Efficiency-first AI models. Continued progress on cheaper inference and domestic accelerator design abroad likely raises the question of whether Western hyperscaler spend is over-provisioned — a concern that hits hardware suppliers before it hits platform owners.
• Macro and geopolitics. Renewed Middle East escalation risk tends to lift energy prices, which feeds affordability and inflation, tightening the rate path. That is a headwind for long-duration AI valuations and for data-center power economics specifically.

Net: sub-1.0x realized/implied ratios were absent here, so hedging remained cheap relative to delivered risk — a stance worth respecting until dispersion narrows.