Year to date, many of the market's most beloved software stocks have been slammed. 2025's market darling, Palantir Technologies (PLTR +1.43%), for example, is down about 22% already in 2026. And long-term wealth compounders Adobe, Salesforce, and ServiceNow have all seen their shares slide about 25% to 30% so far this year. These declines exemplify a beating that focused on most of the market's pure-play software companies. What's even more interesting is that this sell-off comes as these companies report strong results, with artificial intelligence (AI) looking more like a catalyst for their businesses than a headwind.
So, what's causing this widespread fear in the market toward software stocks? Based on most headlines, the explanation is this: Investors are becoming increasingly worried about AI's potential to disrupt software.
But this logic may be broken.
The deep domain knowledge and complex integrations with mission-critical enterprise elements that companies like Palantir or ServiceNow provide make it nearly impossible for someone to throw together copycat software of this caliber. Additionally, just as AI is helping new market entrants, it's also being used by incumbents to enhance their competitive advantages and expand their growth opportunities.
Image source: Getty Images.
But just because the idea that AI will disrupt these software companies is probably broken logic, it doesn't mean there isn't a good reason for these stocks' recent sell-off.
I propose that there is an entirely different reason for software stocks' recent pullback: Their stocks had run ahead of their underlying business fundamentals, creating valuation risk -- and the idea of AI disruption, even if flawed, was just an excuse for these stocks to take a breather.
Understanding valuation risk
With so many stocks rising over the last few years as the S&P 500 climbed 78% between 2023 and the end of 2025, the idea of valuation risk may now seem like a foreign concept to many investors -- especially those who started taking investing seriously during that period. But the premise of valuation risk, of course, is simple: For stocks that trade at high valuations, there's always a chance that the market decides they no longer deserve such a robust premium, and the stock can get rerated lower even as the fundamentals continue compounding nicely.
Therefore, as investors debate whether software stocks can truly be disrupted by AI, I think the bigger question is whether their recent rerating was fair.
I believe it was.
Case in point: Yes, Palantir stock is down about 22% year to date, and its price-to-sales ratio has compressed significantly. But the narrative is quite different when you zoom out a bit.
Since the beginning of 2025, Palantir stock has risen more than 80%, obliterating the S&P 500's gain during this period, and the AI data platform company's price-to-sales multiple has expanded by more than 20% to 75. When viewed this way, you could say the market has rewarded Palantir handsomely for its strong revenue growth (including 70% year-over-year growth in its most recent quarter) and its recent swing to substantial profitability.
Of course, this is an extreme example and isn't necessarily representative of most software companies. Adobe, for instance, has seen both its stock price and the valuation investors are willing to pay for the stock get absolutely demolished since the beginning of 2025. The same is true for Salesforce and ServiceNow. For these stocks, however, it could be a similar story -- just on a much longer time horizon. For instance, if you zoom out over the past 15 years, all three of these companies are handily beating the S&P 500's more than 400% gain over that period, even when you factor in their recent sell-offs.
NASDAQ: PLTR
Key Data Points
The sell-off could get worse
All of this isn't to say that the stock prices of Palantir, Adobe, Salesforce, and ServiceNow are going to worsen from here. These stocks were simply cherry-picked to highlight a broader investment theme.
With that said, there are more software companies than Palantir that still look somewhat egregiously valued even after the recent sell-off in software stocks. Consider the price-to-sales multiples of Snowflake, CrowdStrike, and Shopify of about 14, 23, and 13, respectively. Snowflake's and CrowdStrike's price-to-sales ratios are particularly striking considering they are running at trailing-12-month net losses.
While some software companies, of course, deserve to trade at high valuation multiples, this doesn't mean a sell-off of software stocks overall wasn't unwarranted. Investors may simply be becoming more discerning in gauging the durability of software companies' competitive advantages, rerating valuations lower to reflect those concerns. The disruptive potential of AI may simply be an excuse -- not the core underlying reason for the sell-off.
Sure, investors may be able to find some oversold software companies after the recent sell-off if they look closely enough. But that doesn't mean software stocks overall don't have more room to decline. While there's no guarantee of a further decline, investors should never underestimate the threat of valuation risk.
