What's happening

Across the 11-name analog/auto/specialty bucket (ADI, NXPI, MCHP, ON, MPWR, SWKS, QRVO, LSCC, INDI, CEVA, VPG), the past 30 days have produced red across the board — from roughly -2% (QRVO) to -35% (VPG). But the trend-band data tells a more specific story than a rout: nine of eleven names simply stepped down one notch, from strongly bullish to mildly bullish, in a tight late-June-to-late-July window (MPWR on June 25, MCHP July 1, ADI/NXPI/ON July 6, QRVO July 7, LSCC July 20, VPG/CEVA July 28). Only SWKS actually crossed a full band boundary into mildly bearish, on July 22 — and that coincides with its pending acquisition of Qorvo, a corporate-action overhang rather than a demand signal, especially since Qorvo has suspended standalone guidance calls pending the merger's close.

The one-year gain is broader than it looks

The hypothesis that the year's +35% run rests on one or two AI-levered names doesn't hold up: over the trailing year, VPG (+233%), LSCC (+126%), MPWR (+77%), ADI (+59%) and ON (+45%) all posted big gains, with NXPI, QRVO and MCHP up more modestly and only SWKS (-10%) and INDI (-22%) negative. The analog core — ADI, NXPI, ON, MCHP — has participated, not just ridden along.

Fundamentals still support mid-cycle, not late-cycle

The underlying prints back this up. Analog Devices posted Q2 FY26 revenue up roughly 37% year-over-year with record bookings and channel inventory back to a normal 6-7 week range across all sub-markets. NXP's Q2 revenue rose 19% year-over-year with automotive up 12-17% and a raised 2024-27 auto revenue CAGR target of 8-12%. Monolithic Power guided Q2 revenue up roughly 35% year-over-year, with AI-server power revenue up nearly 98%.

Pricing dynamics also cut against the "Chinese substitution eroding margins" fear embedded in the hypothesis: Texas Instruments has pushed through three rounds of price hikes in six months, up 15-85% on some parts, with ADI and Infineon following suit — a sign of pricing power into strong demand, not erosion. The same reporting notes China's analog self-sufficiency remains below 10% even at 40-45% of global consumption, meaning domestic substitution is a real but early-stage risk concentrated in new design wins rather than existing programs.

The specialty-tail names have idiosyncratic stories: CEVA's July slide tracked a COO transition and analyst downgrades, even as TD Cowen raised its price target following a June AI licensing deal, while Skyworks and Qorvo's own results — a Q2 beat for SWKS and expanding margins for Qorvo — read as an improving RF cycle obscured by merger mechanics rather than deteriorating demand.

Why it matters

This reads as a broad, sector-wide profit-taking pause after an outsized run rather than the front edge of a rollover: fundamentals (bookings, inventory, pricing, auto content) are still confirming the recovery thesis across the analog core, even as smaller/specialty names took the sharpest near-term hits.