Dan Ives Names 5 Software Stocks to Buy Amid AI Selloff
Get the Finance newsletter
Daily finance — markets, central banks, M&A, the prints that move money. Free.
- Anthropic's Claude Cowork release, including plug-ins for legal, sales, finance, data, marketing, and customer support, spooked investors into dumping legacy software and SaaS stocks, with the S&P North American Technology Software Index falling more than 30% from its early September peak.
- Wedbush analyst Dan Ives argues the selling is overdone because enterprises are unlikely to abandon software and infrastructure they've invested decades in for unproven technology, framing the pullback as a buying opportunity for top-tier software names.
- Nvidia CEO Jensen Huang countered the software-is-doomed narrative at an event last week, calling the idea that "the software industry is in decline and will be replaced by AI" "the most illogical thing in the world."
- Ives' top five picks and their implied upside to his price targets: Microsoft ($575, 42%), CrowdStrike ($600, 44%), Snowflake ($270, 51%), Salesforce ($375, 103%), and Palantir ($230, 70%), with Palantir carrying a longer-term path to a "trillion-dollar market cap" (209% upside).
- The damage so far: Microsoft is down 25% from its peak, CrowdStrike 25%, Snowflake 35%, Salesforce 44%, and Palantir 36% (with Palantir trading at a steep 210 times earnings), according to the article.
- Palantir's Artificial Intelligence Platform (AIP) has seen "unprecedented demand" delivering proven ROI on mission-critical use cases where the company "has virtually no competition," per Ives' bull case for focusing beyond current valuation.
Why it matters: If Ives is right that enterprises won't rip out decades of software infrastructure for AI agents, then a 30%+ sector drop triggered by a single Anthropic product release is a buying window — and his 103% Salesforce and 209% long-term Palantir calls show how wide the upside could be. The bear case, by contrast, is that even Nvidia's CEO felt compelled to publicly dismiss the disruption thesis, suggesting the software model's vulnerability is real even if the panic is overdone.
