Software's Valuation Disconnect: A Value Investor's View on the AI Selloff

The market's mood toward software stocks is one of deep uncertainty. The group is off to its worst start in years, with a Morgan Stanley basket of application software companies down 15% so far this year, following a 11% drop in 2025. This selloff, rekindled by the rapid release of new AI tools like Anthropic's, has created a stark disconnect between market sentiment and the underlying durability of many businesses. The core question for a value investor is whether this is a rational repricing of inflated expectations or a mispricing of long-term competitive advantages.

The primary driver is clear: AI-induced uncertainty. Analysts fear that generative AI will compress growth and erode the wide moats that have long protected these companies. As one portfolio manager noted, the pace of change is about as fast as I can remember, and that makes things about as uncertain as I can remember. This anxiety has overshadowed traditional strengths like high margins and recurring revenue. The result is a market that sees no reason to own software, regardless of price, because it assumes zero catalysts for a re-rate exist right now.

Yet, even after this sharp decline, valuations remain elevated relative to the new, uncertain growth trajectory. The basket trades at about 18x forward earnings, a level that demands significant future success. This is the heart of the disconnect. Consider specific examples: Adobe, a company long seen as a fortress, has seen its shares lose more than a fifth of their value over the past 12 months as fears mount that AI lowers the cost of switching and threatens its seat-based pricing model. Similarly, Salesforce's own AI product has touted adoption but hasn't moved the needle significantly for revenue. The market is punishing these companies not for poor execution, but for the perceived existential threat to their business models.

The bottom line is that the market is pricing in a future of compressed growth and eroded moats. For a value investor, the task is to separate this fear-driven noise from the reality of a company's intrinsic value. The current price may reflect a justified discount, but it also creates a potential opportunity if the fears are overstated and the core business proves more durable than the narrative suggests.

Assessing the Moat: Which Companies Are Resilient?

The market's fear of AI disruption has created a valuation chasm, but it has also highlighted a select group of vendors with the most durable advantages. The widening gap tells a story of overreaction. The valuation discount of application SaaS relative to broader software has now reached about 37%, nearly double the level a year earlier. This is the kind of disconnect that disciplined investors watch for. It suggests the market is punishing the entire category for perceived risks, even as some companies demonstrate a clearer path to resilience.

Morgan Stanley's analysis points to a few clear leaders within its "AI Success Framework." In front-office software, the bank favors HubSpot, Klaviyo, Wix and Amplitude, upgrading Amplitude to Overweight. In the back-office, BlackLine has emerged as a new top pick, while Descartes Systems Group was upgraded to Overweight. These are the vendors the bank sees as well-positioned to navigate the AI transition. The consensus view, however, is not a broad embrace. It remains selectively opportunistic, waiting for visible AI-driven revenue upside across estimates before becoming broadly bullish.

The key insight is that the moat is not defined by a company's past success, but by its ability to adapt and expand its role in the new AI layer. Morgan Stanley notes that incumbent vendors' traditional arguments-proprietary data, large installed bases, defensibility-have not been enough to re-engage investors. The burden has now shifted to proving a new narrative: that application software will act as the primary delivery layer for AI, expanding total addressable markets and lifting growth. The favored picks are those where this transition appears most plausible.

For a value investor, the setup is one of selective opportunity. The broad category is being punished for a future that remains uncertain. The evidence points to a few specific companies where the competitive advantages appear to be widening, not narrowing, in an AI-driven world. The market's patience is thin, but the valuation gap offers a potential margin of safety for those who can identify the true winners in this reshuffling.

Catalysts and Risks: What to Watch for a Reversal

The path back to a re-rating for software stocks hinges on a single, tangible shift: the visibility of AI-driven revenue growth. The market's current pessimism is rooted in a lack of catalysts, with many investors assuming zero catalysts for a re-rate exist right now. For a value investor, the key is to identify the specific metrics and scenarios that would signal this fear is misplaced.

The primary catalyst is a clear acceleration in growth that stems from AI, not just internal product updates. Goldman Sachs sees potential here, noting AI could be a tailwind for expanding the total addressable market. This would require a visible transition where application software vendors become the essential delivery layer for AI, expanding their addressable market and lifting growth. The evidence shows this narrative is not yet materializing; as of now, most software makers haven't shown much traction with their own AI offerings. The market will need to see this shift move from promise to concrete earnings acceleration.

On the flip side, the major risk is the continued compression of growth and pricing power. Adobe's experience is a stark example, with its shares losing more than a fifth of their value over the past 12 months as fears mounted. The core threat is that AI lowers the cost of switching and could erode seat-based pricing models. If this trend persists, it would validate the market's worst fears and likely keep valuations depressed. Investors should monitor for a stabilization in customer spending and a shift from narrative to tangible earnings growth.

The broader earnings outlook provides a sobering backdrop. Earnings expansion for software and services companies in the S&P 500 is projected to slow to 14% in 2026, down from estimated growth of about 19% in 2025. This deceleration, even as chipmakers see explosive profit growth, underscores the uncertainty clouding the sector. For a reversion to occur, software companies will need to demonstrate that their AI strategies can not only protect but also expand their growth trajectory, countering this projected slowdown. The market's patience is thin, but the valuation gap offers a potential margin of safety for those who can identify the true winners in this reshuffling.