Is the SaaS apocalypse coming? The bullish case for software stocks

Author: Anthony J. Pompliano, Founder and CEO of Professional Capital Management; Translated by: Shaw Golden Finance

The term “SaaS doom theory” has taken off quickly on Wall Street. Since last September, the market cap of the S&P North American Technology Software Index has fallen 32%. Such a massive drop in just a few months is truly rare.

Salesforce’s stock price is down more than 26%, Adobe’s market cap has shrunk by 20%, and dozens of mid-cap SaaS companies have seen their share prices effectively cut in half. If you’ve been watching enterprise software stocks recently, you’ll see that the entire sector is undergoing a full-scale revaluation of value. And for investors holding these stocks, the performance is undeniably heartbreaking.

But what if everyone is panicking—getting ready to cut their losses and exit at the bottom of the pullback? What if the worst phase is already over?

I don’t have a crystal ball to predict the future, and I don’t intend to forecast the market bottom with precision. But contrarian investors would make this argument right now: the market is overreacting. The mainstream bearish narrative is at odds with the facts. Investors who enter now may be approaching the best layout opportunity in a decade.

Before we discuss where things may go next, let’s first clarify why the market has gotten to where it is today.

This round of declines was caused by two very different shocks. The first shock is macro-level impact. In 2022, the Federal Reserve raised rates at the fastest pace in history, dealing a heavy blow to duration assets.

Very few assets have durations that can exceed those of high-growth software companies—whose cash-flow valuations typically look out over the next ten years. Take the Bessemer Cloud Index as an example: the median revenue multiple for publicly listed SaaS companies fell from the historical peak of 18.4x forward revenue on September 18, 2021, down to about 6x at the start of 2026. Just this one metric’s adjustment has been brutal.

By itself, this valuation metric has dropped 65%.

The second shock comes from structural changes in the industry. To be frank, this is what scares investors even more. The core reason can be summarized like this: the rise of AI intelligent agents has led to what analysts are calling “seat compression”—one AI agent can handle the work of multiple employees, so the number of software licenses enterprises need declines accordingly.

After Atlassian first reported a decline in the number of enterprise seats, its stock plunged 35%. Workday announced layoffs of 8.5% and directly listed AI as one of the reasons. Even if these cases aren’t alarming enough, a CIO survey in January 2026 shows that enterprise IT budget growth is expected to be only 3.4%.

Most analysts believe that the sharp slowdown in budget growth is because money is being redirected toward AI infrastructure builds by mega-scale cloud vendors, with related planned spending exceeding $660 billion. That number is undeniably staggering. Once the market sees massive capital plus panic-driven assumptions, it quickly falls into a selloff.

Early this year, market behavior matched this pattern. A typical example is that enterprise price-to-sales multiples compressed from 9x to about 6x within just a few weeks—an all-time low since the mid-2010s.

But it’s worth noting that software stocks didn’t all fall in lockstep.

There is a clear split between AI infrastructure software and traditional application software. Palantir’s stock price surged 135% in 2025. The reason: its U.S. commercial revenue grew 121% year over year, and it also provided revenue guidance of $7.2 billion for fiscal year 2026—well above analysts’ expectations. Quality companies with real growth can almost ride out any cycle against the tide.

Similar cases include: Microsoft’s Azure, whose quarterly revenue topped $50 billion and still grew 39% year over year; Oracle’s cloud infrastructure, which grew 84% in a single quarter and announced an unfilled backlog of $553 billion. A company holding an order backlog of half a trillion dollars is hard to argue is not more valuable in the future.

Now let’s compare the other end of the market. Salesforce is a sobering example: its market cap has evaporated by more than a quarter. Adobe’s annual revenue growth is still 12%, but its forward P/E multiple has been compressed to about 10x—such a low valuation typically implies the company has entered a permanent decline.

CrowdStrike’s situation is even harder to make sense of. As a widely recognized structural winner in cybersecurity—where its market position grows more important as AI adoption spreads—its valuation is 20% lower than the five-year average price-to-sales ratio. From these examples, I draw the key conclusion: the market is treating all non-infrastructure software as impaired assets across the board. Contrarian investors believe this is the market’s mistake.

So where exactly is the problem?

The mainstream bearish view argues that AI will upend the SaaS industry. The pessimists claim that businesses (especially large groups) will stop paying for software subscriptions because AI agents will take on a large amount of work. Is that possible? Of course. But based on the facts we have today, this is absolutely not an obvious conclusion.

Contrarian investors would point out that this logic has a fatal flaw: industry leaders are not sitting still—they’re building an AI layer on their own. Using their advantages accumulated over 20 years of exclusive enterprise data, customer relationships, and distribution channels, they can’t be replicated by startups in a short time.

In short, traditional giants are only vulnerable to being disrupted when they don’t proactively reinvent themselves.

Take Salesforce as an example. Its Agentforce AI agent platform generated $800 million in annual recurring revenue in fiscal year 2026, up 169% year over year, and is now shifting to a pricing model based on usage and outcomes. If this model can scale successfully, it will be difficult for startups to compete with Salesforce’s channel advantages.

Another area worth watching is ServiceNow’s generative AI suite, Now Assist. Its annual contract value has already surpassed $600 million, and by year-end it is expected to challenge $1 billion. A $1 billion contract value is not something to ignore.

And that’s not even to mention software giant Microsoft. It launched a new Microsoft 365 Enterprise subscription priced at $99 per user per month—65% higher than its prior highest tier—with the goal of directly monetizing AI value through its existing base of users.

Therefore, despite the loud chorus of bearish voices, these cases are not a threat to the SaaS business model. Instead, they indicate that SaaS is evolving into a stronger form—one that directly benefits from the AI technology wave.

Another key point is that many enterprises are shifting from charging by user seat to charging by usage. They no longer charge by headcount; they charge based on tasks completed or outcomes delivered. This business-model iteration, combined with expectations that enterprise software spending will grow 15% this year, makes the “software stock downturn” narrative even harder to sustain.

The ones best positioned to capture this incremental spending are the industry leaders that already have enterprise customer relationships, compliance frameworks, and workflow integration capabilities. Bain’s research also confirms that most customers would rather buy solutions from existing vendors that include AI features. That makes sense: as long as the technology and cost are competitive, enterprises will prioritize familiar partners and trusted brands.

So, what conditions are needed to make the bearish camp miss the mark and help software investors regain confidence?

This situation is somewhat complex. There’s no one-step solution. The market needs a few key turning points. First, AI monetization must break through a credibility threshold: if Agentforce’s annual recurring revenue exceeds $1 billion, the market’s confidence in Salesforce will recover; Now Assist must achieve $1 billion in annual contract value; Microsoft needs to prove that Copilot can continuously and measurably improve average revenue per user. The earliest these data might come out is in the second half of this year, at which point the market narrative will shift from “AI will upend SaaS” to “SaaS monetizes with the help of AI.”

Second, multiple enterprises’ IT budget data must confirm it: even with seat compression occurring, software spending still delivers net growth. This is the most concrete evidence refuting the bearish logic.

Finally, a stable macro environment is crucial. If rate hikes persist, the sector will continue to face pressure; but if interest rates remain unchanged or even begin to move toward rate cuts, the pressure from valuation compression will gradually dissipate, and a true move toward higher valuations will begin.

If the bearish camp is wrong in its assessment, how much upside can investors realistically capture?

Good news: valuations for high-quality software stocks are already at the most attractive levels in years. Microsoft’s forward P/E is about 24x, with an expected annualized earnings growth rate of 14%. Wall Street’s average target price is $600, implying roughly 50% upside from the current stock price.

Cloudflare is similar. Analysts’ average target price is $245, about 40% above the current price. Snowflake’s revenue growth year over year is still 29%, with a price-to-sales ratio of 13x. Consensus expectations point to 43% upside. These upside opportunities are quite meaningful—and none of them have yet been priced in the most optimistic analysts’ target prices.

For those seeking extreme asymmetric returns, contrarian investors would argue that Adobe is the best candidate. Its current forward P/E is about 10x, and both revenue and profit growth remain in the double digits. As mentioned earlier, the market has priced it as a company trapped in structural decline—yet it’s still growing. If Adobe’s valuation reverts to a normal 25x P/E, that alone implies roughly 150% upside from valuation repair, not even counting earnings growth.

In the U.S. stock market, there are plenty of cases like this. Multiple institutional analysts predict that high-quality software stocks that successfully monetize AI can achieve 40%-50% gains through valuation repair alone—before adding earnings growth driven by technological innovation.

Remember: the market’s reaction to “SaaS doom” has been excessive. Some even think the entire industry will be disrupted into irrelevance. But the story told by the data is completely different: perhaps SaaS companies aren’t being upended by external forces—they’re evolving internally into more valuable forms—with higher revenue per customer, lower marginal delivery costs, and a total potential market size that could double over the next four years.

If that turns out to be true, then investors’ recent panic has already delivered these companies an “on-the-clock discount” on their entry prices. The only remaining question is whether, in hindsight, this panic was truly justified.