U.S. Software Sector Down 21% From a Year Ago as AI Disruption Drives Ongoing Valuation Reassessment
Miles Bennett
The S&P 500 Software & Services Index has fallen roughly 21% over the past year while the broader market posted clear gains — Wall Street is pricing in one core bet: AI may eliminate the need to buy software at all.
How far have software stocks fallen — and how wide is the gap with the market?
The S&P 500 Software & Services Index is down about 21% from a year ago, while the broader US market has risen notably over the same period.
This means → software is not pulling back with the market — it is sinking alone against a rising tide, a structural decoupling.
Several major names have fallen far more than the sector average: Adobe, Intuit, Workday, ServiceNow, Salesforce, and Oracle have all seen deep valuation compression.
What triggered the sell-off?
The fuse was lit in February: research firm Citrini Research published a report outlining how AI could upend the existing software industry.
Panic selling followed, wiping out more than $1 trillion in software-stock market capitalisation at one point.
In plain terms = one report turned a vague worry — "could AI kill commercial software?" — into a story with a concrete pathway, and capital fled immediately.
Why did Anthropic's coding tool pour fuel on the fire?
Anthropic then launched its Claude coding tool; users reported it dramatically improved programming efficiency.
This means → "AI replaces commercial software" was no longer just theory — it now had a product users could feel, and the narrative gained real credibility.
The market's core logic: if companies and individuals can use AI to generate their own software, paid demand for legacy vendors will shrink sharply.
Why are AI chip stocks rising while AI software stocks fall — where is the money going?
Software remains weak while AI infrastructure — chips and compute — trades strongly, creating a stark contrast.
In plain terms = the market is channelling all its AI optimism into the "pick-and-shovel" upstream, while treating application-layer software companies as the losers.
This reflects a structural divergence: the same AI wave feeds the upstream and hits the downstream — whether it lasts depends on how much AI coding tools actually change enterprise software purchasing behaviour.
Content is for reference only, not financial advice.