TLDR: Goldman says hedge funds and mutual funds rotate out of software and into semiconductors, betting AI value lies elsewhere.
Key Takeaways:
- Goldman tracks $9 trillion in equity positions and warns the software selloff is not panic, even as leverage stays near the five year high.
- Software and Services is down 14 percent year to date and down 9 percent over 12 months, while Semiconductors and Semi Equipment is up 38 percent and 104 percent.
- Money is concentrating in chip names like LRCX, AMAT, and ASML, while mutual funds add INTC and SITM; even Microsoft got cut.
The market is treating the SaaSPocalypse like a portfolio trade, not a religion. The quiet tell is leverage staying high while software gets treated like the wrong place to harvest AI value.
The market is treating the SaaSPocalypse like a portfolio trade, not a religion. The quiet tell is leverage staying high while software gets treated like the wrong place to harvest AI value.
Q&A
If hedge funds are not de risking, what does their software underweight signal about future deal terms and pricing power?
It implies investors expect weaker valuation support for seat based SaaS, and higher pressure for outcome based pricing that ties revenue to measurable AI native workflow results.
Why would AI shift value away from the application layer even if incumbents keep their customer relationships?
Because the most valuable differentiation may concentrate in proprietary domain data, governance, and workflow context that helps agents act reliably inside enterprises, not just present features.
What could cause the semis trade to cool even while software remains out of favor?
A narrative reset around AI infrastructure returns, faster than expected capex optimization, or evidence that demand is broadening beyond the picks and shovels setup for chips.
Goldman previously found limited productivity links from AI adoption, except customer support and software development. How does that square with the software rotation?
It suggests productivity gains are real but likely accrue to firms that control the workflows end to end, including systems that operationalize agent behavior, not merely to providers perceived as interface wrappers.
What should CIOs and CTOs watch for next when they push vendors toward outcome based pricing?
They should look for which vendors can operationalize AI governance and context, prove performance and compliance guarantees, and demonstrate switching costs that are earned through reliability, not contracts.
No comments yet. Be the first to share your thoughts!