Palantir · one lens

Richard Thaler on Palantir

Retail poured in at 450x trailing earnings while the founder sold — the record holds both halves of the same behavioral loop.

PLTR cautious
An editorial application of a published framework — not the thinker's view, affiliation or advice, and not a recommendation. See methodology.

Thaler’s behavioral-finance work doesn’t treat “irrational” as an insult; it treats it as a predictable pattern a good dataset can actually catch. Two of his mechanisms — narrative-driven demand and loss-averse anchoring — both show up in Palantir’s record at the same time, pointing in different directions.

The first runs through demand. By December 2025, CNBC reported retail traders had poured billions of dollars into Palantir, a stock up roughly 3,000% over three years and trading near 450 times trailing earnings, with retail investors describing their position in narrative terms rather than valuation ones (texxr.com/1154173). Thaler’s work on mental accounting explains why a number that large doesn’t function as a stop sign for a buyer who has already filed the position as “the AI story” rather than “a 450x multiple” (Misbehaving). The account a gain gets filed under changes how much risk feels acceptable inside it — his own finding that gamblers take bigger risks with “the house’s money” than with their original stake applies just as well to an investor already sitting on a multi-year run-up (Misbehaving).

The second mechanism runs the other way, inside the same window. Alex Karp sold $1.9 billion of Palantir shares between January 2024 and March 2025, even as the stock funding those sales kept climbing on the same retail demand (texxr.com/883414). An insider realizing gains while public enthusiasm accelerates isn’t proof of anything by itself — but it is the asymmetry Thaler’s loss-aversion research flags: the party with the most information about the business is locking in a reference point, while the party with the most narrative exposure is extending one. Revenue did accelerate alongside all of this — up 85% year over year by Q1 2026 — which is exactly why the split doesn’t resolve cleanly. The fundamentals and the behavior are both real, running through the same coverage record at once.

An editorial application of Richard Thaler's published framework — not their actual view, affiliation, or advice.

Where this lens comes from

Richard Thaler's framework is set out at the lens page, drawn from Misbehaving. This page applies it to Palantir and nothing else — the company's full record is in the dossier.

The other lenses on Palantir

← Palantir dossier