Market data reflects trailing 7-day stock price performance through February 4, 2026 for the 25 publicly traded software companies with the largest declines over that period. Public market analysis conducted on F2. Image created using Gemini 3.
F2 AI, Gemini 3
On Tuesday, ~$300 billion of market value evaporated across SaaS, data, and software-heavy investment firms. This was not an earnings miss or a macro shock. It was an AI product release.
The damage had been building for months. By the time markets reacted, the IGV Software Index was already down roughly 30 percent from its late-September peak. What changed last week was not the direction, but the speed.
Several of the most entrenched enterprise software companies fell sharply in a single day. Salesforce, ServiceNow, Adobe, and Workday each dropped around 7 percent. Intuit fell nearly 11 percent. At the same time, valuation multiples across the sector compressed violently. The average forward earnings multiple for software companies collapsed from roughly 39x to about 21x in just a few months. Short sellers have already made over $20 billion in 2026 betting against legacy SaaS and are doubling down.
Markets do not erase that much value unless a core assumption breaks.
The assumption that just broke was the durability of legacy SaaS growth.
For most of the past two decades, enterprise software benefited from a remarkably stable economic story. Software was expensive to build. Switching costs were high. Data lived in proprietary systems.
Once a platform became the system of record, it stayed there. That belief underpinned everything from public market multiples to private equity buyouts to private credit underwriting. Recurring revenue was treated as a proxy for predictability. Contracts were assumed to be sticky. Cash flows were assumed to be resilient.
AI is now testing every part of that logic at once.
What spooked investors last week was not that AI can generate better features. Software companies have survived feature competition for years. What changed is that modern AI systems can replace large portions of human workflow outright. Research, analysis, drafting, reconciliation, and coordination no longer need to live inside a single application. They can be executed autonomously across systems.
The mood driving the selloff was captured bluntly by Chamath, who posted on X:
“The Great SaaS Meltdown has started and there’s no going back… A new AI-oriented workflow is coming… the great SaaS meltdown has started.”
The $300 billion repricing was not random. It reflected investors accelerating their expectations around workflow substitution risk.
Feature competition compresses margins.
Workflow replacement redirects spend.
When workflows move, value moves with them.
Customers do not need to rip out legacy systems overnight for this to matter. They consolidate. They renegotiate. They reduce usage. Cash flow weakens before logos disappear. Markets understand this dynamic instinctively. That is why companies long considered “sticky” sold off together, regardless of near-term fundamentals.
But the story investors are telling themselves is incomplete.
What is happening is not the collapse of software. It is a reallocation of where value accrues inside a much larger market.
Recent research from Goldman Sachs projects that AI agents will materially expand the overall software market by the end of the decade, while capturing a disproportionate share of the profit pool. In their framework, agents do not merely enhance applications. They become the interface to work itself. By 2030, more than 60 percent of software economics could flow through agentic systems rather than legacy SaaS seats.
The profit pool in software is expected to shift toward AI agents
Goldman Sachs, Gartner
This is the key distinction. The market is growing, not shrinking. But legacy software economics are being diluted as intelligence, memory, and execution move outside static applications and into adaptive systems that operate across tools.
In other words, companies are not paying less for software. They are paying less for licenses and more for outcomes.
That shift explains both the selloff and the opportunity. When profit pools move faster than revenue disappears, public markets react immediately. Private markets follow later.
The implications are especially significant in private equity and private credit.
Over the last decade, enormous amounts of capital flowed into software businesses based on a shared set of assumptions: predictable revenue, low churn, and high recovery value. These assumptions justified leverage and covenant structures that treated software cash flows as among the safest in the economy.
AI does not break these portfolios overnight. It creates a lag. Spend compression appears before churn. Margin erosion shows up before covenant breaches. Economic reality diverges from reported metrics.
For debt investors, this is the most uncomfortable scenario. Risk increases before the numbers make it obvious. By the time the data confirms the problem, pricing power is already gone. It is no coincidence that asset managers with heavy software exposure traded down alongside the applications themselves.
This forces a harder question: what does “recurring” really mean in an AI-native world?
Recurring revenue used to imply predictability because replacing software required human labor, long implementations, and organizational pain. AI reduces that friction dramatically. Intelligence becomes portable. Data becomes interpretable. Workflows become replicable.
Software increasingly looks like a container. Intelligence becomes the product.
That inversion breaks many of the heuristics investors relied on for years.
Some argue this risk is overstated. Enterprises are unlikely to rebuild mission-critical systems from scratch with unstable, homegrown tools. They will continue to “rent” software rather than build it themselves. To some degree, that is true.
But the rent model is changing.
Rather than sticking with legacy SaaS incumbents, forward-leaning organizations are shifting toward purpose-built, LLM-agnostic AI platforms that deliver portable intelligence and workflow execution without the baggage of outdated architectures. These are not feature add-ons. They redefine what the software layer actually does.
This distinction is clearest in private markets, where the real constraint has never been software availability. It has been human throughput.
Private equity and private credit teams are not over-softwared. They are under-leveraged. Highly paid professionals still spend extraordinary amounts of time on manual data extraction, reconciliation, memo drafting, and diligence coordination. Legacy tools record information. They do not execute the work.
Vertical AI agents win here not by displacing core systems of record, but by replacing expensive, offline human workflows that sit between them.
At F2, we are not competing with spreadsheets, document management systems, or data rooms in the traditional sense. We are eliminating the labor-intensive processes wrapped around them. The result is not lower software spend. It is higher return on decision-making.
I have seen this shift from multiple angles. Earlier in my career, I was the private equity associate buried in Excel at 2 a.m., underwriting businesses that depended on exactly these assumptions. Later, building fintech infrastructure and now AI tooling for private markets, I have watched how quickly those assumptions unravel once intelligence moves closer to the work itself.
The lesson is not that software disappears. It is that legacy software economics are being rewritten.
As AI agents expand the market, value shifts toward systems that own execution rather than interfaces. Revenue becomes less about access and more about impact. Recurrence matters less than replaceability.
For investors, the implication is uncomfortable but clear. Traditional signals arrive too late. By the time churn shows up, workflow substitution has already occurred.
In an agent-driven market, the edge belongs to those who understand where intelligence is moving before the income statement does.
That is the real disruption investors are pricing in. And it is only just beginning.