Cramer explains the divergence in tech stocks – and why software may continue to lag

IGV software ETF dropped more than 4% as Salesforce and Adobe fell about 3% and 4%, respectively, and CrowdStrike plunged 7.5%; hardware names Marvell and Intel gained nearly 5% and Corning rose 2.85%. Jim Cramer framed the move as a persistent rotation into hardware and AI infrastructure at the expense of enterprise software and expects the 'buy hardware, sell software' trade to continue in the near term.

Analysis

Market reallocation toward AI infrastructure is functioning as a multi-stage valuation rotation rather than a one-off trade: capital is shifting from annuity-like software multiples to near-term, contractable hardware revenues tied to large data-center builds. That mechanics favors vendors with modular BOMs and visible backlog (fabs, substrate, specialty glass) and penalizes companies whose growth is priced on long-duration ARR assumptions, amplifying volatility in ETF/quant flows that bucket exposure by sector rather than fundamental nuance.

Second-order supply-chain effects are underappreciated. If hyperscalers accelerate chassis and accelerator orders this year, expect upstream bill-of-material beneficiaries (substrate/packaging, thermal solutions, front-end test services) to see margin expansion within 2-4 quarters, while there is a 3-6 month risk window for cyclical inventory builds at ASIC suppliers that could create a mid-cycle pullback once OEMs digest initial shipments.

Key reversal catalysts: a meaningful fall in real yields or signs of software ARR resiliency (reaccelerating net-new bookings, fewer delayed renewals) would re-rate software quickly; conversely, a geopolitical shock to Asia supply chains or an abrupt hyperscaler pause would vaporize near-term hardware earnings visibility. Position sizing should reflect this asymmetric information environment — lean into hardware where backlog visibility is high, hedge with short-dated protection for software exposure, and keep time horizons explicit (tactical 1-3 months vs structural 6-24 months).

Key Decisions for Investors

• Pair trade (6-12 months): Long MRVL (target +30% / stop -18%) financed by short CRM (target -25% / cover if CRM > +15% from entry). Rationale: MRVL exposure to networking/storage silicon benefits from capex pull-forward; CRM is the high-duration software victim of multiple compression. Size as 1:0.6 notional to reflect different vol profiles.
• Tactical option spread (3-9 months): Buy IGV 3-month put spread (buy 1x 10% OTM put, sell 1x 25% OTM put) to capture continued software ETF weakness with defined max loss = premium; target 2.5x premium if software momentum deteriorates further. Use this as hedging overlay against long hardware exposure.
• Capital-efficient hardware upside (9-18 months): Buy INTC calendar/vertical structure — buy near-term ATM calls, sell further-dated 25–35% OTM calls to fund — or alternatively a 12-month call spread (buy ATM, sell 30% OTM). Expect asymmetric upside if share gains come from AI server market-share recovery; cap max loss to premium paid.
• Material/infra long (12-24 months): Buy GLW with a 12–24 month horizon (target +25% / stop -15%). Rationale: structural growth in data-center glass and fiber demand with less execution/leverage risk than some semiconductor names; keep exposure as a core cyclical overweight but size conservatively given macro uncertainty.