Nvidia Sale
The Nvidia sale is one of Damodaran's most-cited recent investment decisions and a worked example of his framework in action. He had held Nvidia for years, profiting from the AI infrastructure build-out that drove the stock from approximately $25 in late 2022 to over $140 by mid-2025. He sold the entire position by the end of 2025. His stated reasoning was not that he had lost faith in Nvidia's business — he explicitly considers it "a company that delivers" — but that the market price had reached a level that required nearly impossible outcomes to justify. "You need too much to go right to break even," he said. The sale is a textbook application of his pricing vs. valuation distinction: the price had moved beyond the intrinsic value, even under optimistic narrative assumptions.
In the AI Story
The sale illustrates a discipline that distinguishes Damodaran's investment behavior from the dominant pattern in growth investing. Most growth investors hold winners as long as the narrative remains intact, on the theory that the multiple expansion drives most returns. Damodaran's discipline requires comparing the multiple to the intrinsic value at every point, and selling when the multiple has expanded beyond what the intrinsic value supports — even if the narrative remains favorable.
The mechanics of the decision involved a specific narrative-to-numbers translation. Nvidia's price implied that revenue would continue compounding at rates that the addressable market could not sustain, and that margins would remain at current levels despite eventual competitive entry from custom silicon (Google's TPUs, Amazon's Trainium, hyperscaler in-house chips). The narrative required by the price was not impossible — Nvidia might capture and hold dominant share — but the probability-weighted intrinsic value was below the market price. The sale followed.
The decision is consistent with Damodaran's broader Net Zero Thesis. The aggregate AI infrastructure build-out is real; some companies will capture lasting returns; but the price required to participate in the AI infrastructure trade as of late 2025 had reached a level that priced in winning outcomes too aggressively. The sale was not a bet against AI; it was a recognition that the price had moved beyond what the AI thesis could support.
The contrast with his Microsoft hold is instructive. Microsoft's value derives substantially from its cloud business — "a utility essential to modern life" whose stability does not depend on AI's speculative upside. The narrative-to-numbers translation for Microsoft did not require optimistic AI assumptions to justify the price. He held. Two stocks in the same broad AI ecosystem; two opposite decisions; both grounded in the same framework applied to specific company-level analysis.
Origin
Damodaran disclosed the Nvidia sale in his year-end 2025 commentary and elaborated in subsequent interviews in early 2026, including the January 2026 conversation that produced the "net zero" framing.
Key Ideas
Selling on price, not narrative. The sale was triggered by the divergence between price and intrinsic value, not by deteriorating fundamentals.
"Too much to go right." The price required nearly all favorable scenarios to materialize; probability-weighted intrinsic value was below market price.
Discipline against growth-stock holding orthodoxy. Most growth investors ride winners; Damodaran's framework requires selling when the multiple exceeds intrinsic value.
Coherent with the Microsoft hold. Same framework, different company analysis, opposite conclusion — illustrating the framework's company-specific application.
Appears in the Orange Pill Cycle
Further reading
- Aswath Damodaran, "Nvidia: A Valuation Update," Musings on Markets blog (multiple posts 2023-2025)
- Aswath Damodaran, January 2026 interviews on AI investment
- Various financial press coverage of Damodaran's Nvidia disclosure (December 2025 - January 2026)