Software Stocks: Are They a Bargain or a Trap?

The software sector has been getting hammered, and the fear is palpable. Just last Monday, the iShares Expanded Tech-Software Sector ETF, known as IGV, tumbled 4.8% in a single day. That drop is part of a brutal year so far, with the fund down 24% and on pace for its worst quarter since 2008. The culprit is a wave of pessimism about artificial intelligence. Investors are worried that new AI tools could let anyone write their own code, making traditional software products obsolete. As one strategist put it, these companies are "guilty until proven innocent."

This panic has driven valuations to levels that scream extreme fear. The sector's price-to-earnings ratio has fallen to less than 21 times earnings, its lowest point in over three years and far below its historical average. For context, that multiple is rarely seen since the 2008 crisis. The market is pricing in a catastrophe.

Yet here's the disconnect: the business fundamentals haven't collapsed. In fact, the results are often strong. Of the 15 software companies in the S&P 500 that have reported earnings this season, 87% have beaten profit expectations and 67% have topped revenue forecasts. That divergence between solid results and a collapsing stock price is what makes this setup interesting. As one analyst noted, software is becoming "the new value sector."

This is where the core investment thesis comes in. The market is overselling based on AI fears, but the real test is whether these companies have useful products and loyal customers. Deep value investor Lee Roach's insight is relevant here: when the market expects a company to "immediately die," but it's actually only "slowly declining," the potential returns can be astonishing. AI may have weakened the moat of software companies, but it doesn't mean their terminal value is zero. The setup forces you to ask a simple question: is the product still needed, and do people still pay for it? That's the common-sense check that the panicked market is ignoring.

Kicking the Tires: Product Utility and Brand Loyalty

The market's fear is that AI will make software products obsolete. The common-sense test is simpler: are these tools still essential for running a business? The recent partnership news with AI startup Anthropic suggests the answer is still yes, for now.

On Tuesday, shares of several major software firms climbed after announcing collaborations with Anthropic to build new AI "plug-ins" for tasks like investment banking and HR. Companies like LSEG, FactSet, and DocuSign saw their stocks tick higher, helping lead a sector-wide rebound. This isn't just hype; it's a practical move. The new tools aim to integrate AI directly into workflows, suggesting that established software platforms are still seen as the necessary foundation for critical business operations. People are still paying for these products to solve real problems.

That said, the AI threat is real and it's eroding the traditional "moat" of software companies. The fear is that if anyone can "vibe code" with tools like Anthropic's, the demand for expensive enterprise suites could dry up. This is the core tension. The market is pricing in a terminal decline, but the business reality shows customers are still buying. As one strategist noted, the disruption may be "years out yet," and the market is already pricing in the worst-case scenario.

The upcoming earnings reports from giants like Salesforce and Workday will be the next real-world test. Can management prove their products remain indispensable? For Salesforce, the stock is trading at a record low multiple, making any proof of enduring utility potentially very rewarding. The setup is clear: the product must still work, the brand must still command loyalty, and the utility must still be undeniable. If it is, the current panic may be the last major hurdle. If not, the erosion could accelerate. For now, the partnership deals show the market is still willing to pay for integration, not just disruption.

The Simple Math: What to Watch for a Turn

The market is pricing software as if it's already dead. The real test is watching for signs it's still breathing. For the sector to stop being "guilty until proven innocent," investors need to see concrete evidence that the slow decline is slowing-or reversing.

The first thing to kick the tires on is customer behavior. Watch for signs that people are still paying for the product. That means looking past the headline revenue numbers to see if revenue per user is holding steady or growing. If that metric starts to fall, it signals customers are either using less or finding cheaper alternatives. Similarly, watch for customer churn rates. A sudden spike would confirm the AI threat is hitting the bottom line now, not years from now. These are the real-world utility metrics that matter more than any earnings call promise.

Management guidance will also be a key indicator. The partnership news with Anthropic is a start, but it's about integration, not just hype. The next earnings reports from giants like Salesforce and Workday will show whether leadership is talking about practical AI tools that customers can actually use, or just grand visions. If management focuses on how AI plug-ins are boosting sales or reducing support costs, that's a positive signal. If they keep talking about "transformative" AI as a distant future risk, it suggests they're not yet confident in the near-term fix.

The ultimate catalyst for a turnaround is a shift in valuation. The sector's price-to-earnings ratio has fallen to levels rarely seen since the 2008 crisis. A sustained rebound in that multiple back toward historical averages would be the clearest sign that the market's extreme fear is unwinding. It would mean the sentiment is shifting from "guilty until proven innocent" to "innocent until proven guilty." That's the setup where the common-sense investor can step in.

For now, the math is simple. Watch the customer metrics, listen to the guidance, and wait for the P/E to tick higher. The panic may be overdone, but the proof will be in the numbers that show people still need these tools.