The AI Trade Is Rotating: Money Is Moving Out of Chip Stocks and Into Software and Applications

The Quick Take

• After being the primary AI trade beneficiary for 18 months, chip stocks (Nvidia, AMD, TSMC) are seeing money rotate into AI software and application layer companies
• The rotation thesis: chip valuations have run to levels that fully price in near-term demand, while AI software stocks still have room to re-rate on revenue momentum
• Application layer beneficiaries include enterprise AI software, AI-native SaaS companies, and companies converting AI infrastructure investment into productivity gains
• The rotation is not a 'sell AI' signal — it's a maturation signal: the infrastructure build is recognized as complete enough that market attention is moving to monetization

The AI investment trade has historically moved in waves: first, infrastructure picks-and-shovels (semiconductors, data center equipment) lead as the market prices build-out; then, software and application companies follow as the monetization layer becomes visible. The current rotation — money flowing out of chip stocks that have already delivered exceptional gains and into AI software and enterprise application names — is consistent with this pattern. Nvidia is not losing its fundamental AI demand thesis; it is simply transitioning from 'early and obvious' to 'widely held and fully priced' in the institutional portfolio allocation sense.

The software and application layer beneficiaries of this rotation span a wide range: enterprise AI platform companies (Palantir, C3.ai, ServiceNow's AI extensions), cloud providers whose AI services are generating visible revenue (Microsoft Copilot, Google Gemini enterprise, AWS Bedrock), and vertical AI solutions in healthcare, finance, and legal. What these companies have in common is that their AI revenue is now measurable in quarterly filings — investors can put a growth rate on the adoption curve rather than relying on speculative addressable market analysis.

For portfolio positioning, the rotation signal suggests a tilt toward AI companies with demonstrated software revenue growth over pure hardware exposure. This doesn't mean abandoning semiconductor names — demand for Nvidia H100/H200 remains strong and HBM cycle is still building — but it means the next period of AI-related alpha may come from companies one or two steps up the value chain. Watch quarterly earnings from enterprise software names for AI revenue line-item disclosures, as these will be the data points that sustain or reverse the rotation thesis.

Synthesized from 2 sources.