The Zacks Analyst Blog Highlights SKYY, WCLD, CLOU and FCLD

Cloud computing demand is cited as accelerating—global cloud infrastructure service spending rose 35% YoY to $129B in Q1 2026, with cloud infrastructure revenues projected to top $500B in 2026. AI-optimized IaaS spending is forecast to jump 96% in 2026 to $42.3B (and 56.5% to $66.1B in 2027), supporting “cloud resilience” during pullbacks. The article spotlights cloud-focused ETFs SKYY ($3.44B AUM, 60 bps), WCLD ($331.2M, 45 bps), CLOU ($350.6M, 68 bps), and FCLD ($132.8M, 39 bps), noting many are near 52-week highs.

Analysis

This is more a momentum/flow validation note than a new fundamental catalyst. The important market implication is that the incremental dollars from AI-cloud spend accrue first to the picks-and-shovels stack: networking, accelerators, and owned infrastructure. That favors ANET and NVDA more than the software-heavy cloud ETFs, because their revenue recognition is tighter to capex cycles and less dependent on downstream monetization that can lag by quarters.

The less obvious risk is that broad cloud enthusiasm can actually pressure margins for the application layer. If enterprises are re-platforming to AI workloads, compute costs and depreciation intensity rise faster than top-line growth, which can compress free-cash-flow conversion for names like SNOW, DDOG and NTNX even if usage trends remain healthy. In other words, “cloud spend up” does not automatically mean software multiple expansion; the market may be overpaying for duration while underpricing the cost of delivering AI workloads.

The biggest second-order winner is the ecosystem around AI clusters: networking, optics, storage, and power-efficient data-center buildouts. The biggest contrarian risk is that hyperscaler capex inflects slower than consensus after the next earnings round, which would hit the highest-multiple cloud proxies first and leave the ETF baskets vulnerable to a sharp factor unwind. If AI demand is real but supply of GPU capacity keeps expanding through neoclouds, pricing power can normalize faster than the current growth narrative implies.

Key Decisions for Investors

• Prefer a relative-value long ANET / short WCLD pair over the next 4-8 weeks: ANET has cleaner exposure to the capex wave and less duration risk than the ETF’s software-heavy composition. Falsifier: if cloud software names start outperforming on improving billings, the pair is wrong.
• Use pullbacks to add to NVDA rather than chase SKYY/CLOU at 52-week highs. Best entry is a 3-5% dip ahead of hyperscaler earnings/capex commentary; stop if next-round capex guidance flattens or order growth slows.
• Avoid initiating fresh longs in SNOW, DDOG and NTNX on this headline alone; if you want exposure, hedge them against NVDA/ANET. The thesis breaks if those software names show accelerating net retention or margin expansion despite higher AI spend.
• Treat the cloud ETF strength as a tactical flow trade, not a long-duration signal. If SKYY/CLOU cannot hold recent highs after the next macro/rates repricing, fade the move with a short-term sell signal rather than a structural short.
• Watch MSFT, AMZN and GOOGL earnings for capex-to-revenue conversion. If capex keeps rising but operating margin guidance compresses, that is a cue to rotate out of the broad cloud basket and into the infrastructure winners.