Salesforce's blowout quarter reversed the February "SaaS doomsday" selloff, lifting the State Street SaaS ETF 45 percent to a record as high-moat software names recovered while low-moat vendors lagged.
返回
Software stocks rally 45% as Salesforce earnings reverse AI doomsday
The AI-replacement panic that wiped hundreds of billions from software valuations in February has reversed, with the State Street SaaS ETF up 45 percent from its low to a record after Salesforce's blowout quarter.
"The 'SaaS doomsday' narrative was complete nonsense," Salesforce Chief Executive Marc Benioff told CNBC, days after the company's shares posted their best session in six years.
Salesforce's Agentforce annual recurring revenue topped $1.5 billion, up more than 240 percent year over year, and the company raised its full-year revenue forecast, sending shares up 23 percent on Aug. 27. Identity-security vendor Okta jumped 29 percent that day and is up 86 percent year to date, while CrowdStrike rose 20.5 percent and Atlassian climbed to its highest level since July 2025.
The rebound has split the sector along a single fault line: how hard a vendor's core asset is for a frontier lab to rebuild. High-moat names are leading the recovery, while low-moat products that AI tools can replicate still trade below their pre-panic levels.
What ended the panic
The selloff began in February when Anthropic said its Cowork assistant could accelerate contract review and compliance workflows inside corporate legal teams. Four sentences of product description were enough to erase hundreds of billions in software market value, as investors reasoned companies would build internal tools with AI rather than pay for packaged software.
Nvidia Chief Executive Jensen Huang called the reaction "the most illogical thing in the world," and analysts argued the selloff painted every software company with the same brush. Seven months later, the market has reached a more granular verdict.
Salesforce's quarter was the turning point. Agentforce ARR of $1.5 billion, up 240 percent year over year, and 3.2 billion Agentic Work Units in the quarter showed customers were paying for AI agents rather than replacing software with them. The company's expanded partnership with Anthropic, which layers Claude's reasoning on top of Salesforce's customer data, reinforced Benioff's argument that frontier models depend on CRM data instead of displacing it. Nine of the top 10 AI companies now run on Salesforce and Slack, with combined spending up 435 percent year over year.
The moat test separates winners from laggards
Okta's recovery shows what investors now reward. New products made up 30 percent of the company's bookings in its latest quarter, up from 5 percent the prior period, as enterprises treat identity security as a first priority before scaling AI agents, Bloomberg Intelligence analyst Mandeep Singh said.
The laggards tell the opposite story. Monday.com, whose workflow tools are easy for companies to replicate with lightweight AI-built productivity apps, is down 35 percent year to date. Tax-software giant Intuit has fallen 48 percent. In private markets, database startup Airtable was acquired by Bending Spoons for $2.25 billion, a fraction of its $11.7 billion valuation in 2021, a deal analysts read as a salvage purchase of a product that is essentially an enhanced spreadsheet.
"We all came to realize the industry is going to be more resilient than expected... We're not going away," KeyBanc analyst Jackson Ader said. "That's why Salesforce put up a decent but not spectacular quarter and got such a huge reaction."
The divergence is likely to persist as long as AI deployment favors vendors with proprietary data and specialized workflows. Salesforce's own numbers show the stakes: overall revenue grew just 11 percent in the quarter, and organic growth was 6.4 percent excluding the Informatica acquisition, yet the forward guidance, not the reported results, drove the 23 percent move. For investors, software valuations now hinge on which companies pass the moat test rather than on the sector's aggregate exposure to AI.
This article is for informational purposes only and does not constitute investment advice.
