Semiconductor ETFs Hit Record Highs as Software ETFs Plunge in 2026

Semiconductor ETFs have surged to new record highs since late March, while the iShares Expanded Tech-Software ETF has fallen to levels last seen in late 2023. Snowflake and HubSpot posted their worst weekly performance in years, and other software names including Cloudflare, Intuit, Atlassian, Workday, Zscaler, Datadog, DocuSign, and RingCentral also sold off sharply. TrendLabs said fresh software-stock lows could signal broader market weakness, though the U.S. Dollar Index trigger has not yet been met because the dollar remains weak.

Analysis

The tape is telling us this is not just a factor rotation; it is a dispersion regime where profitable, cash-generative software is being re-rated like duration while semis are being treated as the only credible AI monetization rail. That means the pain is likely to stay concentrated in the middle of software beta first—workflow, observability, security, and collaboration names with elevated multiples and slower near-term billings reacceleration—before it reaches the mega-cap software franchises. The second-order effect is margin pressure on adjacent enterprise IT spend: if CIOs perceive semiconductor-led capex as the only “must-own” AI budget, software procurement cycles can elongate by 1-2 quarters.

The setup is also a positioning squeeze risk in reverse. The recent low-to-low move in software suggests systematic de-risking is still in the early innings, which often produces another 5-10% downside once momentum funds and vol-control sellers finish resetting. But semis at fresh highs are vulnerable to an air pocket if the AI spend narrative narrows too much; if leaders keep outperforming while breadth deteriorates, any earnings miss or guidance haircut can trigger a fast 8-12% factor unwind across the whole tech complex.

The most interesting contrarian is that the market may be over-discounting software resilience rather than under-discounting semiconductor fragility. Several of the pressured software names have tangible AI attach opportunities and operating leverage that can reassert quickly if bookings stabilize; the better long is likely not broad software, but specific quality compounders with net retention and free-cash-flow durability. The macro trigger to watch is the dollar: a sustained rebound would be the cleaner signal that the rotation has room to extend, while continued dollar weakness argues this is still mostly a relative-growth trade, not a broad risk-off break.

Key Decisions for Investors

• Short IGM/IYW basket vs long SOXX for 2-4 weeks: express the divergence while breadth is weak; target 6-8% relative outperformance for semis, but cut if SOXX loses its prior breakout level on volume.
• Buy puts on NET or DDOG 30-45 days out after any 2-3 day relief bounce: these are the highest convexity short-duration expressions if the software de-rating continues; aim for 2:1 payoff into another leg lower.
• Pair long SNOW vs short a software index basket only if SNOW holds recent lows and guides to accelerating consumption; this is a contrarian quality-vs-beta trade with asymmetric upside if AI data workloads reaccelerate.
• For hedging, add small semis downside via SOXX put spreads rather than outright shorts: semis remain the trend leader, but the crowding makes them the cleanest hedge against a broader tech unwind.
• Defer adding to WDAY/TEAM/INTU until post-earnings and only on confirmation of stabilization; these can be tactical longs if they reclaim 20-day moving averages, but they are not yet worth catching on the way down.