Software stocks are back from the dead as AI fuels growth: can the rally continue?

Anthropic’s February launch of new Claude Cowork plugins triggered a global selloff in software firms, wiping out roughly $300B of value across US markets in a single day. The move amplified investor fears that AI-driven tools could replace traditional software economics, leaving the sector effectively “written off” at the time.

Analysis

The first-order move is usually a positioning event, but the second-order risk is real for application-layer software: if AI reduces the labor embedded in a workflow, buyers will renegotiate seat-based pricing and concentrate spend into fewer platforms. That is more damaging to high-multiple point solutions and “nice-to-have” workflow tools than to systems-of-record with distribution, data gravity, and embedded switching costs. In other words, the market should be discriminating between interface risk and franchise risk; it often does not in the first leg of the selloff.

The likely winners are the compute stack and the incumbents best placed to bundle AI into existing budgets: semis, hyperscalers, cloud/security, and broad platforms with captive user bases. A less obvious beneficiary is services/automation that can help customers implement AI faster, because the budget may shift from recurring software licenses toward integration and consumption. The losers are the software names with the weakest proof of pricing power and the clearest exposure to commoditized tasks; they will face multiple compression before any actual revenue impairment shows up.

Over the next 1-3 months, the key catalyst is whether management teams can show stable retention and attach rates rather than just product demos. If enterprise buyers delay new seats or push for outcome-based pricing, the downside can persist for another quarter; if not, the move should mean-revert sharply. The 6-18 month risk is genuine workflow substitution, but consensus is likely overstating how quickly that translates into P&L damage because procurement, integration, and security constraints slow adoption.

The contrarian read is that the market may be over-discounting disruption at the app layer while underpricing the ability of large platforms to absorb AI spend and defend share. The thesis is falsified if software NRR, renewal cohorts, and forward billings hold up through the next earnings cycle, or if managements can show AI monetization is expanding wallet share rather than cannibalizing it.

Key Decisions for Investors

• Favor a 1-3 month pair trade: long SMH or NVDA vs short IGV on rallies. Risk/reward is better on the infra side if the market keeps repricing AI as capex rather than software substitution; cover if hyperscaler capex guidance rolls over.
• Short a basket of the most duration-sensitive application software on strength into earnings — CRM, NOW, SNOW, ADBE — with a 4-8 week horizon. The setup only works if guidance or billings crack; stop out if NRR/revenue retention stays resilient.
• Relative-value long MSFT vs short a weaker standalone software basket. MSFT has the best distribution to bundle AI into existing workflows, so it is more likely to monetize AI without losing the customer relationship.
• Use the next earnings season as the decision point, not the headlines. If forward billings, cRPO, or net retention do not deteriorate, the current fear trade is likely overdone and should be faded.
• Watch for a reversal trigger: any management commentary showing AI is being sold as an upsell rather than a substitute. That would be the signal to unwind shorts and rotate back into software beta.