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

Software sell-off fears look overstated as Microsoft posted +18% y/y revenue with Azure momentum and Microsoft 365 Copilot seats now exceeding 30M, ServiceNow grew subscription revenue +23% (constant currency) and signed 123 deals over $1M, while Salesforce reported +14% y/y Q1 revenue and nearly $34B in remaining performance obligations with a record 98 deals. The main near-term risk is heavy AI capex—Microsoft spent $41B last quarter—which could pressure earnings, plus a potential seat-licensing drag from more capable AI agents. Overall, the article frames the pullback (MSFT roughly flat YTD, CRM down 24%, NOW down 24%) as a potential buying opportunity supported by strong fundamentals and continued long-term earnings estimates.

Analysis

The market is treating AI as if it only accrues to infrastructure suppliers, but the larger economic moat is likely in the application layer where MSFT, NOW, and CRM sit. Their advantage is not just distribution; it is that they control the workflow, identity, and data permissions that make AI agents enterprise-grade, which should preserve pricing power and reduce churn even if seat counts grow more slowly. That creates a second-order winner set: systems integrators, boutique point solutions, and smaller workflow vendors are more exposed to displacement than the incumbents being sold off.

Near term, Microsoft’s capex intensity is the key denominator risk, not a business-model break. If AI spending keeps rising faster than monetization, the stock can stay rangebound despite strong top-line metrics because the market will keep discounting peak margin pressure. The catalyst path is the next 1-2 earnings cycles: watch Copilot attach, Azure growth, and billings/RPO quality rather than headline AI mentions. A miss there would falsify the thesis quickly; otherwise, the multiple can re-rate over 6-18 months as free cash flow normalizes.

The contrarian point the consensus may be missing is that AI agents are a double-edged sword: they can both deepen software lock-in and reduce the number of paid seats. That makes this less of a clean multiple expansion story and more of a stock-selection story within software. The best risk/reward is to own the names with the strongest enterprise switching costs and the clearest ability to monetize workflows, while avoiding vendors whose value proposition is easier to substitute with AI-native tools.

Key Decisions for Investors

• Build a 3-6 month long basket in MSFT/NOW/CRM on additional weakness, but size modestly until the market gets another earnings read-through; thesis breaks if next-quarter Azure/Copilot or subscription growth decelerates materially.
• Favor NOW over lower-quality software peers on a relative basis: workflow control should convert AI adoption into seat expansion and pricing power faster than generic SaaS. Use a pair trade long NOW / short a broad software ETF if available, or keep NOW as the core long in the sector.
• Avoid chasing the move in NVDA as the implied beneficiary of all AI spending; the better risk/reward over the next 1-3 months is in lagging software names with intact fundamentals, where revisions can improve as capex converts into monetization.
• For event risk, consider call spreads into the next earnings cycles rather than outright stock if valuation sensitivity is a concern; upside comes from multiple re-rating, while downside is capped by continued capex drag. Reassess if MSFT capex keeps rising without corresponding Azure acceleration.