Dan Ives Says Software Stocks Just Saw Their Most Disconnected Sell-Off Since the Late 1990s. Here's Why He's Still Bullish.

Software stocks (Microsoft, Salesforce, ServiceNow) are down sharply YTD (MSFT flat vs Nasdaq +11%, CRM down 24%, NOW down ~29% as of the article), but the bull case argues the sell-off is disconnected from fundamentals. Microsoft reported 18% YoY revenue growth last quarter with Azure momentum and 30M+ Microsoft 365 Copilot seats; ServiceNow’s subscription revenue rose 23% in constant currency and closed 123 deals over $1M; Salesforce delivered 14% YoY revenue growth and nearly $34B in remaining performance obligations, plus a record 98 deals over $1M in new annual contract value. Key near-term risk flagged is heavy AI capex—Microsoft spent $41B in capital expenditures last quarter—which could pressure earnings, though the view is it supports long-term advantage as AI agents may eventually reduce incremental seat/license demand.

Analysis

This is less a “software versus AI” call than a quality-sorting event inside SaaS. Incumbent platforms with embedded data, identity, and workflow hooks should capture a larger share of AI budgets than point solutions, because the buyer wants governance and integration more than novelty. That makes MSFT and NOW the cleanest beneficiaries, while weaker seat-based vendors face a slower but more structural loss of pricing power as AI features get bundled into existing contracts.

The near-term battleground is free cash flow, not revenue. Heavy AI capex can leave MSFT looking optically expensive for 1-2 quarters even if demand is healthy, because buybacks and margin optics will lag the revenue narrative. The second-order winner is NVDA: enterprise AI adoption is still translating into compute demand before it translates into software monetization, so hardware captures the first dollar of spend while software absorbs the implementation burden.

The market may be overestimating how quickly AI agents cannibalize software seats and underestimating how sticky incumbent workflow systems are. The real risk is not churn tomorrow, but a 6-18 month slowdown in net new licenses and renewal uplift if customers substitute usage-based AI for headcount-based software growth. Falsifiers: a sequential drop in MSFT FCF margin from capex, a deceleration in NOW subscription growth/large-deal counts, or CRM billings conversion failing to keep pace with its RPO backlog.

Key Decisions for Investors

• Buy MSFT on 3-5% weakness over the next 2-4 weeks; target a re-rating into the next earnings cycle. Risk/reward is favorable if revenue stays >10% y/y, but cut if AI capex keeps outrunning operating income and FCF margin compresses again.
• Buy NOW as the cleaner operating-leverage beneficiary versus the broader SaaS complex over a 1-3 month horizon. Upside is strongest if $1M+ deal momentum holds; invalidate the thesis if subscription growth or large-deal counts decelerate for two consecutive quarters.
• Pair trade: long MSFT / short a basket of lower-quality SaaS names with weaker switching costs (e.g., WDAY, DOCU, SNOW) over 6-12 months. The thesis is incumbents monetize AI through bundling while weaker vendors face seat cannibalization; stop if AI-driven net retention improves broadly across the group.
• Treat CRM as a watchlist, not a fresh buy, until backlog converts into billings at a better rate. If RPO remains large but current revenue growth slows, the stock can stay range-bound despite positive AI rhetoric.
• Maintain a smaller tactical long in NVDA as the second-order compute beneficiary of enterprise AI spend, but size it below the software longs. The risk is a near-term capex digestion phase at hyperscalers; the thesis breaks if data-center spending guidance rolls over.