SaaS Stocks Enter Bear Market: AI Fears Erase $2T
Software stocks have lost nearly $2 trillion in market cap as AI agents threaten the per-seat licensing model. Is the selloff overdone?
The B2B software sector has entered a bear market. The iShares Expanded Tech-Software ETF (IGV) is down more than 21% year-to-date, and the damage to individual names is worse. Nearly $2 trillion in market capitalization has evaporated since January.
This isn't a garden-variety rotation out of growth. It's an existential reckoning with what AI means for the software business model that created trillion-dollar companies over the past two decades.
The Damage
The numbers are brutal:
| Company | YTD Decline |
|---|---|
| Intuit | -40% |
| Workday | -33% |
| Adobe | -32% |
| Salesforce | -31% |
| Microsoft | -23% |
| Shopify | -26% |
These aren't speculative names. They're the pillars of enterprise software—companies with fortress balance sheets, sticky customer relationships, and decades of growth. When the market sells them off this aggressively, something structural is happening.
The AI Disruption Thesis
The core worry: AI agents are replacing human workflows at scale. If enterprises need fewer human users, they need fewer software seats. The per-seat licensing model that made Salesforce and its peers rich looks suddenly vulnerable.
Recent surveys suggest 40% of IT budgets are being reallocated from traditional SaaS subscriptions to "agentic" platforms and large language model token usage. That's money flowing out of Salesforce and into Anthropic, OpenAI, and the AI infrastructure providers.
The question isn't whether AI disrupts software—it clearly does. The question is how fast and how completely. Bears see an existential threat to the entire category. Bulls see a transition that takes years, giving incumbents time to adapt.
The Contrarian Case
JPMorgan strategists argue the selloff has gone too far. The market is pricing "worst-case AI disruption scenarios that are unlikely to materialize" in the near term.
Their logic: enterprises don't rip and replace mission-critical software overnight. Salesforce CRM data, Workday HR records, Adobe creative workflows—these systems are deeply integrated into how companies operate. Switching costs remain high even if AI offers theoretical improvements.
And the incumbents aren't standing still. Microsoft has Copilot. Salesforce has Einstein. Adobe has Firefly. They're embedding AI into existing products, potentially strengthening rather than undermining their competitive positions.
The bears counter that this misses the point. AI doesn't need to replace Salesforce entirely to crush its valuation. It just needs to slow seat growth and compress margins. A company growing seats 15% annually trades very differently than one growing 5%.
What the Flow Shows
Options activity in software names has tilted bearish. Put-call ratios on IGV hit 1.8 last week, near the highest levels since the 2022 tech selloff. Institutional money is hedging or outright shorting the sector.
The IPO market tells a similar story. Seven of the top ten largest Q1 software IPOs traded below issue price within 30 days. Investors aren't just selling existing software stocks—they're rejecting new ones.
That kind of across-the-board pessimism sometimes marks bottoms. But it also reflects a fundamental repricing that could persist for quarters.
Where the Value Might Be
Not all software is equally exposed. Infrastructure-layer companies—those selling to developers rather than end users—may actually benefit from AI adoption. More AI applications mean more cloud compute, more databases, more observability tools.
Enterprise automation platforms could go either way. If AI agents need orchestration, companies like ServiceNow could thrive. If AI agents replace the need for orchestration entirely, they're in trouble.
The most vulnerable are horizontal application providers selling commoditized features. Why pay for a standalone expense management tool when the AI assistant handles it?
The Trade
This isn't a sector to bottom-tick. The structural questions are real, and clarity won't come for quarters. But 21% down with valuations approaching historical lows deserves attention.
A barbell approach makes sense: short the most vulnerable horizontal apps while going long the infrastructure layer and the incumbents with credible AI strategies. Microsoft at 23% off highs looks different than a speculative SaaS name at the same discount.
The market will eventually sort winners from losers. Until then, expect volatility to persist and the AI narrative to dominate every earnings call through the summer.