Money & Markets

SaaSpocalypse Wiped $285 Billion in Two Days. Markets Got It Wrong

Claude Cowork's January 2026 launch triggered a $285 billion two-day wipeout in software stocks. Oxx's Richard Anton says the panic broke live private deals — and exposed public markets' pricing problem.

By Nathan Brooks

4 min read

Public Markets Are Frequently Wrong. The SaaSpocalypse Shows Us Why
Public Markets Are Frequently Wrong. The SaaSpocalypse Shows Us Whysergeitokmakov / pixabay

What's News

  • An estimated $285 billion in market cap was wiped from software, data analytics and fintech companies over two days in January 2026.
  • The BVP Nasdaq Emerging Cloud Index fell 31% by April 2026, then rose 83% from its low by end of October.
  • CrowdStrike dropped 27% in early 2026, then gained 114% from the start of the year to end of October.
  • Oxx's Richard Anton says multiple private transactions in process early 2026 fell away after false public-market signals.
  • The number of U.S. public companies has halved since the late 1990s, a trend accelerated by the 2002 Sarbanes-Oxley Act.

Traders erased an estimated $285 billion in market capitalization from software, data analytics and fintech companies in just two days after Anthropic's Claude Cowork launch triggered the January 2026 panic now known as the SaaSpocalypse. By April, the BVP Nasdaq Emerging Cloud Index had fallen 31% since the start of the year. By the end of October, it was up 26% year-to-date and up 83% from its April low.

That round trip is the story. Richard Anton, co-founder and general partner of the software-focused venture firm Oxx, argues in a new analysis that the episode exposed the limits of public markets in judging software value — not the end of the SaaS business model itself.

"The SaaSpocalypse didn't reveal the end of software. It revealed the limits of public markets in judging its value," Anton wrote.

How the markets got it wrong

The panic had a specific trigger. Claude Cowork launched in January 2026, and additional plug-ins announced later that month targeted enterprise support workflows such as HR and legal. From there, valuations went into a tailspin, with traders treating software companies indiscriminately — punishing the ones that would thrive in the AI era alongside the ones genuinely at risk.

This wasn't the market's first AI misjudgment, either. When OpenAI launched ChatGPT to the public in November 2022, Alphabet stock was punished even though Google later emerged as a clear winner in generative AI. Then the market swung the other way: it was slow to respond to generative and agentic AI's implications for SaaS through the end of 2025, and enterprise software market caps kept rising until the January break.

Anton, who says he has invested in software "over many decades," frames this as a familiar pattern of underreaction followed by overreaction. In his view, AI has not ended software companies. Instead, it has driven a shift in how software is built, in where user growth will come from — primarily agents rather than humans — and in the sources of defensible competitive advantage.

That shift was always going to be uneven, he argues. Companies deeply embedded in their customers' core business processes stand to benefit. Others face genuine threat. The public market's error was refusing to make that distinction during the sell-off.

What does the recovery look like?

The bounce-back has been bifurcated. At Oxx, Anton says the firm "never believed in the SaaSpocalypse." Sub-sectors like cybersecurity and data and analytics have fared well, while others, like workflow management, have fallen back.

CrowdStrike is the standout example. Its share price dropped 27% in early 2026, then rose 114% between the start of the year and the end of October — a swing that illustrates just how disconnected the January pricing was from the company's underlying position.

Anton describes the arc as the standard hype cycle that plays out whenever new technology emerges: "a surge of excitement, then a correction, then productivity." In the short term, he says, markets simply overestimated the harm AI would inflict on the software sector as a whole.

A bigger problem than sentiment

The knee-jerk reactions are not just temporary market mood, in Anton's telling. He ties them to a structural shift that began in the late 1990s and accelerated in 2002, when the Sarbanes-Oxley Act introduced tougher investor protections after the dot-com crash.

The unintended consequence: company ownership migrated from public to private markets. The number of public companies in the U.S. has halved, while private markets have grown dramatically. That is where much of the smart money now sits, along with "a more accurate, and patient, assessment of software companies' value," Anton writes.

What does this change for founders and dealmakers?

The most concrete damage may be in private transactions. Anton says the public-market panic sent false signals that broke live deals: "multiple transactions which had been in process early this year fell away in the wake of the false signals which came from public markets."

His prescription starts with recognition. "Few people recognize there's a problem with the pricing function of public markets, and the first step to solving this issue is to recognize there is actually a problem," he writes.

For founders weighing an exit, the implication is direct. If public markets can swing a cloud index down 31% and back up 83% within a single year, an IPO may no longer look like the ultimate goal — and being publicly owned may no longer be appealing. The SaaSpocalypse, on this reading, was less a verdict on software than a stress test the public markets failed.

Oxx is a specialist software investor backing European founders aiming to transform their industries with AI as they ready to scale.

Original: crunchbase.com

Share this article:

More from Nathan Brooks

Nathan Brooks

Show full bio

News editor covering regulation, policy and their impact on entrepreneurship, startups and venture funding at Business Bearings.

2,460 articles

Related articles

« Previous articleNext article »