The SaaSPocalypse: Why Wall Street Is Rotating From Software to Semiconductors

Wall Street is currently navigating a fundamental reassessment of where value accrues in the age of artificial intelligence. A significant investment rotation is underway, shifting capital away from the application layer of software and toward the hardware that powers it.

The performance gap between these sectors is stark. Software and Services as an industry group has declined 14% year-to-date and lost 9% over the past 12 months.

Conversely, Semiconductors and Semi Equipment have surged, rising 38% year-to-date and climbing 104% over the last year.

The Institutional Exodus from Software

According to Goldman’s U.S. Weekly Kickstart, published May 22, the shift is evident across trillions of dollars in equity positions. analysing $9 trillion in positions at the start of the second quarter of 2026, the report highlights a deliberate move by major funds.

View this post on Instagram about Weekly Kickstart, Even Microsoft

From Instagram — related to Weekly Kickstart, Even Microsoft

Hedge funds have reduced software to its lowest weight in their long portfolios since 2019. Meanwhile, mutual funds—excluding Microsoft—are carrying their widest underweight position in software since 2012.

This movement is not a result of general de-risking. Hedge fund net leverage is currently running at the 85th percentile of the last five years, suggesting a calculated consensus that software is currently the wrong allocation.

Did You Know? Hedge fund net leverage is currently at the 85th percentile of the last five years, indicating that the rotation out of software is a deliberate strategic call rather than a broad move to reduce risk.

Targeted Portfolio Shifts

During the second quarter, hedge funds increased net positions in LRCX, AMAT, and ASML. Mutual funds showed similar bullishness toward semiconductors, piling into INTC and SITM.

Even Microsoft, often viewed as AI-proof, was cut on net by both mutual funds and hedge funds last quarter.

Skepticism is also reflected in earnings projections. While Info Tech is forecast to grow earnings by 31% in 2026, Goldman’s top-down estimate of $92 in sector EPS contribution is notably lower than the bottom-up analyst consensus of $106.

The Battle for AI Value

The current volatility stems from a debate over whether AI will destroy the Software-as-a-Service (SaaS) model or simply evolve it. Some analysts, including those at JPMorgan, argue that long-term contracts and switching costs provide structural moats for incumbents.

Goldman has separately projected that the global app software market could reach $780 billion by 2030, suggesting that agentic AI might actually expand the overall software market by the end of the decade.

However, the era of 20x revenue multiples and seat-based subscriptions may be fading. The bull case for software now depends on incumbents’ ability to reprice around usage and leverage proprietary data.

Expert Insight: Samantha Carter notes that the pivot from seat-based pricing to outcome-based models represents a high-stakes transition for software vendors. The ability to monetize actual business results rather than user counts will likely determine which incumbents survive the platform shift.

Enterprise Demand and the “Wrapper” Problem

Corporate buyers are already demanding changes. CIOs and CTOs have begun renegotiating contracts and threatening to replace tools that lack AI-native capabilities.

🎙️ EP 201: Goldman Sachs Just Hired an AI. The SaaSpocalypse Is Here.

Joel Hron, CTO of Thomson Reuters, suggests the dividing line is between companies with deep, proprietary, domain-specific data and those that are essentially “interface wrappers.”

ServiceNow has leaned into this shift. At its Knowledge 2026 conference, President and COO Amit Zavery declared that “the era of sidecar AI is over,” stating that customers now want outcomes rather than cobbled-together pieces.

The company is betting on its Context Engine, a governance layer built on 7 trillion annual transactions and 100 billion workflows, to provide the necessary guardrails for AI agents.

Potential Risks and Contradictions

The narrative is complicated by conflicting research from Goldman. In March, the firm found that AI adoption showed meaningful productivity gains specifically in software development and customer support—the core workflows of SaaS platforms.

research from May suggests that “FOMO,” rather than rational capital allocation, may be driving the AI infrastructure boom. The firm concluded that hyperscalers may be prioritizing the AI arms race over their current shareholders.

If the semiconductor supercycle is partially driven by narrative, the current rotation away from software could eventually face its own market reckoning.

Frequently Asked Questions

What is the “SaaSPocalypse” investment trend?

It refers to a major portfolio rotation where hedge funds and mutual funds are reducing their exposure to software stocks—reaching weights not seen since 2019 and 2012, respectively—and shifting that capital into semiconductors.

How are enterprise customers changing their behavior toward software vendors?

CIOs and CTOs are taking a harder line by renegotiating contracts, demanding pricing based on outcomes, and seeking tools with native AI capabilities rather than add-ons.

What is the difference between “interface wrappers” and data-driven AI companies?

According to Thomson Reuters CTO Joel Hron, the distinction lies in whether a company possesses deep, proprietary, domain-specific data to train and differentiate its AI models, rather than simply providing a user interface for existing AI.

Do you believe the shift toward semiconductor hardware is a permanent structural change or a temporary bubble driven by FOMO?