The iShares Expanded Tech-Software Sector ETF has shed roughly 14% year-to-date while the S&P 500 gains about 6%, despite software delivering its strongest earnings beats in years. The paradox is not a market error, it is a deliberate repricing of business models that have not broken yet but may be about to.
There is a strange logic operating in software markets right now. The sector is producing some of its best fundamentals in years, and Wall Street is selling it anyway. The iShares Expanded Tech-Software Sector ETF (IGV) surged nearly 16% over three trading sessions earlier this year on the back of strong fourth-quarter earnings, then reversed sharply into an eight-session losing streak that erased the entire gain. As MarketWatch recently reported, software has remained a major source of pressure inside technology shares, with IGV falling again this week even as semiconductor stocks rebounded. Today IGV sits about 14% below where it started January, making software one of the weakest major technology groups even as the broader index holds gains near 6%. During that same stretch, software companies posted double-digit earnings growth, with 93% of S&P 500 software names beating profit estimates in Q4, well above the 74% beat rate for the rest of the index.
The wedge between fundamentals and price action has a name on Wall Street: the SaaSpocalypse. The term sounds dramatic, but the underlying logic is coherent. AI-native coding agents and workflow automation tools are advancing fast enough that investors are already discounting the future of per-seat subscription licensing before any actual revenue impairment shows up on an income statement. If an enterprise deploys AI agents that accomplish what 500 licensed seats used to handle, the renewal math changes. Goldman Sachs estimates software stocks have lost roughly $2 trillion in collective market capitalization this year, a number that reflects anticipated structural change in the revenue model, not current results.
The disruption is beginning to show up selectively in company data. Atlassian reported its first decline in total enterprise seat counts, and shares fell roughly 35%. Google has disclosed that AI writes approximately 25% of its new code, a statistic that shifted from an internal milestone to a sector-wide valuation signal almost overnight. ServiceNow beat Wall Street's Q1 2026 expectations, growing subscription revenue 22% year-over-year to $3.67 billion, and still lost about 17% of its market value after reporting, because investors focused on something else in the transcript: more of the company's growth is moving toward non-seat-based pricing structures, including tokens and consumption models. ServiceNow is actively changing the pricing architecture that justified its valuation multiple, and doing so by choice, which tells you something about where management thinks the pressure is heading.
What makes the current moment genuinely difficult to read is that the bear case and the bull case are drawing on the same data set. ServiceNow's Now Assist AI suite is tracking toward $1.5 billion in annual contract value, up from a prior target of $1 billion. Gartner forecasts 14.7% global software spending growth for 2026. EPS revisions for the sector are among the strongest in the S&P 500, and the sector's earnings growth rate is projected to climb from 17% last year to 21% this year. Analysts surveying enterprise SaaS earnings calls this cycle found AI described overwhelmingly as a revenue accelerant, not a headwind. The numbers support optimism. The price action does not.
Goldman Sachs has been explicit about why the selloff has been too broad. The firm draws a line between software companies whose value sits in data gravity and workflow lock-in versus those whose value sits almost entirely in license volume. The former category may actually expand as AI increases the economic value of underlying data platforms. Goldman identifies Snowflake, MongoDB, Shopify, and CrowdStrike as companies with architectural moats that extend beyond the application layer, and rates them accordingly. The implication is that the SaaSpocalypse is real but unevenly distributed, and that markets pricing the whole sector down equally are creating a selection opportunity for investors willing to do the work.
Adobe reports second-quarter fiscal 2026 earnings after the close today, and the result carries unusual weight for the sector. The stock is down roughly 30% year-to-date as investors debate whether generative AI tools hollow out its creative software revenue or expand the total addressable market for content production. Adobe is expected to report about $6.45 billion in revenue and adjusted EPS near $5.81, with traders pricing in a sharp move in either direction. Investors will be watching less for the headline numbers than for the tone of forward commentary on AI monetization, CEO succession, and pricing power. A strong print with clear AI revenue attribution could offer the sector a test case for what a credible recovery narrative looks like.
For institutional investors, the relevant question is not whether AI disrupts software. It clearly does at the margin, and likely more at the center over time. The question is whether the disruption timeline is fast enough to justify current price action in 2026 or whether stocks are discounting 2028 damage today. If AI augments software spend and the earnings trajectory holds, IGV sitting well below its January level against a rising market is one of the sharper disconnects in technology. If enterprise seat counts begin falling at scale across Salesforce, Workday, and ServiceNow through the back half of this year, the selloff will look prescient in hindsight. Adobe's number tonight is the next inflection point in that argument, and the software sector will read it accordingly.
Also read: Gold sinks to its lowest level since November as the Fed's tightening bias overrides every safe haven argument • Japan's NTT Is Raising $1 Billion to Build AI Data Centers in America and Institutional Money Is Quietly Writing the Check • American traders moved billions through Polymarket's offshore exchange while the company quietly lobbied Washington to change the rules
