Dan Loeb: The Lost Art of Short Selling, and Why Stock Picking is Back

What It Covers

Dan Loeb, CEO of Third Point managing nearly $30 billion AUM, covers his evolution from anonymous internet troll and early short seller to multi-strategy investor, explaining why short selling has returned as a viable strategy, how AI forces technological literacy on all investors, and his criminal justice reform work including the Ross Ulbricht pardon.

Key Questions Answered

• •Short Selling Framework: Avoid purely valuation-based short positions — stocks with irrational valuations can get squeezed by Reddit-driven momentum indefinitely. Loeb's team targets structural impairments instead, such as their home builder short thesis built around hidden land-pool liabilities disguised as options and unsustainable post-COVID construction cost inflation that buyers can no longer absorb at current mortgage rates.
• •Investor Skill Requirements: Pre-2008, fund managers could generate returns while being technologically or economically illiterate. Today, neither is viable. Technology now runs through every asset class and capital pool, meaning even non-tech-focused allocators like private credit firms must understand the tech through-line in every business they evaluate or risk systematic blind spots.
• •Management Quality Assessment: After 30 years, Loeb still evaluates management teams through qualitative pattern recognition rather than any quantifiable rubric. Given AI disruption compresses competitive moats faster, the key filter shifts from product durability to whether a leadership team demonstrates a proven, repeatable ability to adapt ahead of disruption cycles.
• •Multi-Strategy Platform Construction: Third Point expanded from event-driven equity into structured credit, high-yield, CLOs, private credit, direct lending, and a wholly-owned insurance company. The insurance vehicle specifically captures investment-grade returns from private structured credit and whole loans, while surplus capital gets deployed opportunistically — demonstrating how hedge funds can build permanent, lower-cost capital bases.
• •Nvidia Valuation Thesis: Loeb argues Nvidia is undervalued on a two-to-three year earnings basis, comparing current skepticism to historically misguided "ceiling" thinking applied to Google and Amazon. Long-short pod structures force managers to short something liquid and large-cap, making Nvidia a default target — creating a technically-driven discount unrelated to fundamental earnings trajectory.