Some Experts Argue Software Stock Sell-Off Was ‘Too Harsh’ Despite AI Fears

All the talk of AI “eating” software could have some investors eating their words, according to a recent Jefferies report.

A rough year for software stocks morphed into “Saaspocalypse” last week when the release of new agentic tools from Anthropic amplified fears that AI is set to disrupt the software industry. The iShares Expanded Tech-Software Sector ETF (IGV) fell about 8% last week, putting the fund down about 22% since the start of the year. But Jefferies analysts see opportunities for investors who can handle uncertainty. (Software stocks were surging Monday afternoon, helping push major indexes sharply higher.)

“The negative sentiment is reaching extreme levels,” wrote Jefferies analysts in a note on Sunday. By their measure, sentiment among software investors is nearly as negative as during the 2008 Global Financial Crisis and the Dotcom Crash of the early 2000s.

After last week’s rout, more than 40% of the software stocks in Jefferies’ coverage are trading near historically low valuations. But analysts argue proclamations of software’s death are “premature,” and that “incumbents that embrace the transition should emerge as [long-term] winners.” 

Software stocks have been dogged by fear that AI threatens the industry on multiple fronts. AI agents and what’s becoming known as “vibe coding” could erase software companies’ competitive moat. Increased competition from AI-native startups like OpenAI and Anthropic could pressure industry margins. And AI-driven efficiencies at software buyers could shrink headcount, undermining the industry’s seat-based pricing model.

Momentum in other corners of the technology sector isn’t helping sentiment in the software space, according to Jefferies. Tech giants, racing to add computing capacity to train and run AI, are expected to spend about 60% more on infrastructure this year than in 2025.

A huge share of that capital is flowing to semiconductor companies and other suppliers of data center hardware. Advanced Micro Devices (AMD), one of the slower growing AI hardware suppliers, is expected to increase revenue by more than 30% this year, compared with growth in the mid-teens for major software providers like Salesforce (CRM) and Intuit (INTU). 

Many software companies are also struggling to demonstrate that AI is adding to the top or bottom lines. Jefferies estimates that AI accounted for no more than 3% of revenue at any of the application software companies in its coverage last year. Meanwhile, chip makers’ sales have skyrocketed amid booming demand from AI data centers. 

Jefferies believes that over the long term, today’s AI concerns will prove overblown and that benefits will accrue to software providers that are “nimble and adapt to the realities of the AI transformation.” They say the companies that are best positioned are those with superior access to data, established distribution networks, and entrenchment in enterprise workflows.

As for the sector as a whole, a few catalysts could spark a rebound in software stocks, according to Jefferies. The industry would be aided by clarity on the software ambitions of AI model providers like OpenAI and Anthropic. Do they want to replace software or augment it? An answer to that question could clear a big dark cloud hanging over the industry. 

Evidence that software companies are benefitting from AI would also help to shift the narrative. A meaningful increase in enterprise deployments could boost AI-driven revenue, while implementing AI internally could expand margins. Finally, moderating data center spending could buoy the industry by weakening AI infrastructure’s tailwinds, making software’s modest growth marginally more attractive to investors.