A Contrarian Take on AI: Is It Time to Buy Software Stocks?

What It Covers

Andrew examines whether SaaS stocks like Adobe, Workday, and Salesforce — down 30–40% year-to-date — represent buying opportunities, while challenging the assumption that AI disruption mirrors the 1990s internet boom, using MIT cognitive studies and OpenAI's own mental health data to reframe the AI narrative.

Key Questions Answered

• •SaaS Sector Drawdown: Six major software-as-a-service stocks have declined 30–40% since January 2025 — Workday down 40%, Intuit 38%, FICO 36%, ServiceNow 34%, Adobe 31%, Salesforce 30%. Before buying these dips, investors should distinguish between temporary sentiment-driven selloffs and genuine structural disruption to recurring subscription revenue models.
• •Subscription-to-Usage Risk: The core threat to SaaS valuations is a potential shift from per-seat subscription pricing to usage-based billing, similar to cloud computing. If companies currently charging thousands of seats for underused software must reprice by consumption, revenues could contract substantially — making this a company-by-company analytical question, not a sector-wide assumption.
• •Recency Bias in AI Investing: Markets are pricing AI as the next internet, but the two differ structurally. The internet was designed as an open, decentralized network; AI is built and controlled by a handful of for-profit corporations. Investors should audit whether their AI thesis is driven by evidence or by pattern-matching to the 1990s dot-com expansion.
• •Adobe's Fundamentals vs. Narrative Gap: Adobe's Q1 2026 results show annualized recurring revenue up 11%, deferred revenue up 15%, and gross margins holding steady despite data center costs rising 14%. The CEO departure after 18 years and segment consolidation add uncertainty, but the numbers do not yet reflect the disruption the stock price implies.
• •Buy-the-Dip Realism: Comparing Franklin Resources (bought 2016, returned 7% over five years despite strong historical metrics) against Alphabet (bought May 2024, up 80%) illustrates that strong historical numbers alone do not validate dip-buying. Moat analysis — specifically whether existing customers face sufficient switching costs — must precede any decision to add to a declining position.