Bill Gurley & Michael Mauboussin - Putting Theory into Practice - [Invest Like the Best, EP.370]

Episode Summary

Executive Summary: Patrick O’Shaughnessy hosts Bill Gurley and Michael Mauboussin for a wide-ranging talk on AI, increasing returns, regulation, capital allocation, venture dynamics, and genius. The conversation frames how technology waves create value, why incumbents are often advantaged, and where investors should look for durable edge.

Main Topics: AI: hype, limits, and incumbent advantage (Priority: 5/5): They separate LLMs from broader AI, warning against inflated expectations while noting incumbents moved fast. Increasing returns and network effects (Priority: 5/5): They explain why rare feedback loops can compound value for customers and firms. Intangibles, recombination, and innovation (Priority: 4/5): They argue modern growth is driven by scalable digital assets and recombining existing ideas. Regulatory capture and blocked progress (Priority: 5/5): They say entrenched regulation helps incumbents and slows innovation in healthcare, energy, finance, and telecom. Venture capital structure and cyclicality (Priority: 4/5): They criticize VC’s low barriers to entry, high barriers to exit, and too much capital chasing too few great deals. Capital allocation under low rates (Priority: 4/5): They show companies did not behave as theory predicted during ZIRP, with buybacks, cash, and leverage diverging from expectations. Physical-world technology and hard-tech (Priority: 4/5): They see opportunity in energy, nuclear, robotics, and healthcare, but note regulation and capital intensity raise the bar.

Key Arguments: AI is profound, but LLMs are narrower than AI and won’t quickly deliver fantasy outcomes. Incumbents have an edge because AI adoption is highly choreographed and API-based. Foundational model companies face startup-like valuations without normal startup-market discipline. True increasing returns are rare but can create exponential value by raising willingness to pay. Intangibles scale fast, but they are easy to copy and can become obsolete quickly. Regulation often becomes the friend of the incumbent and reduces competition after major laws pass. Venture has low barriers to entry and high barriers to exit, so too much capital destroys returns. Public companies shrank 46%, yet private-market expansion is not a simple fix. Companies used a ~15% hurdle rate even when the cost of capital changed materially. The biggest market winners matter most: 2% of companies created $50T of $55T of wealth.

Data Points: Date of recording: April 15th in 2024 - Patrick timestamps the discussion for AI-era context Magnificent Seven economic profit share: about 45% - Michael says the group generated this share of U.S. stock market economic profit Market cap share of Magnificent Seven: 25 or 30% - Michael contrasts valuation share with profit contribution Return on capital trend: starting around the year 2000, that flipped - Michael says large-company returns overtook smaller-company returns around then Tech CapEx vs energy CapEx: 2x - Michael says top five tech companies spend twice top five energy companies’ CapEx U.S. public-company count decline: down 46% - Michael cites the shrinkage in public listings Venture market size: a little over a trillion, maybe trillion and a half - Michael estimates U.S. venture AUM Benchmark fund size: 450 or 500 million dollars - Bill references his firm’s funds in contrast to mega-funds Cost of capital hurdle rate used by CFOs: roughly 15% - Michael cites John Graham’s CFO survey Treasury-bill underperformance share: just under 60% - Michael summarizes Bessembinder’s public-market study Aggregate wealth destroyed by underperformers: $9 trillion - Michael cites Bessembinder’s estimate Aggregate wealth created by winners: $64 trillion - Michael cites Bessembinder’s estimate Net U.S. market wealth creation: $55 trillion - Michael gives the net result from 1926 through 2022 Share of companies creating most wealth: 2% - Michael says this tiny slice created $50T of the $55T Value created by the top 2%: $50 trillion - Michael cites Bessembinder's concentration result Probability of platform success: less than one in seven - Michael references an academic study on firms seeking to become platforms Cost decline from Wright’s Law: 20% - Michael says cost per unit falls for each doubling of cumulative output Historical auto diffusion example: within a decade - Bill says automatic transmissions spread across manufacturers in about 10 years Boeing/bridge comparison: 12 days - Bill cites the I-95 bridge being rebuilt in 12 days as an exception to bureaucratic delay

Pivotal Quotes: "What could go right?" — Bill Gurley: He describes the best mindset for partner meetings and evaluating new ventures "The patterns have a half-life and they decay." — Bill Gurley: He warns against overusing pattern recognition in venture investing "If you want to be an optimist, that's certainly one way you could argue for that." — Michael Mauboussin: He frames recombination and digital tools as a source of future progress

Implications: Investors should separate real compounding mechanisms from hype, and watch where AI, energy, and healthcare can genuinely rewire workflows despite regulation.