November 2025 Random Ramblings

Episode Summary

Executive Summary: Andrew Walker’s monthly ramble reflects on investing psychology and process: how winning streaks can create overconfidence, how sold ideas that later fall can become dangerous re-entries, why shorting high-valuation AI names is risky, and why management-provided targets and NAVs often mislead. He frames these as questions about edge, tailwinds, and humility in valuation.

Main Topics: Wins That Lure Investors Into False Security (Priority: 5/5): Walker argues that early success in a strategy or sector can create overconfidence, using craps as a metaphor for hot streaks and his own history in legal special situations and Malone-style levered buybacks. When Passed-on Stocks Drop and Become Re-buys (Priority: 5/5): He describes a recurring pattern where he passes on a stock due to a perceived risk, the stock falls after that risk materializes, then he buys it later—often with poor results—prompting him to question his mental model and process. Shorting the AI Bubble Is Not Straightforward (Priority: 4/5): Walker discusses the growing narrative around an AI bubble and cautions against easy comparisons to dot-com-era shorts, arguing that companies like Palantir and CoreWeave may still have room to run if they become long-duration winners like Amazon, Netflix, or Tesla. Trusting Management Numbers, Targets, and NAVs (Priority: 5/5): He critiques investor-day projections and management NAV presentations, saying that even when the numbers look reasonable, the investments often underperform because the market may already know the story or because hidden overhead and structural issues are ignored. John Malone, Leverage, and Tailwinds (Priority: 4/5): Walker revisits John Malone’s levered buyback model and suggests its historical success may have depended heavily on declining interest rates and an unusually favorable media environment rather than the strategy alone. Humility, Edge, and the Limits of Diligence (Priority: 5/5): Across all topics, he emphasizes that investors need to understand where they truly have edge, avoid confusing a hot streak with skill, and remain skeptical when a thesis relies on simple valuation math or management narratives.

Key Arguments: Early success can be a function of variance, not skill; investors may mistake a hot streak for a durable edge. His own history suggests legal special situations were profitable in one era and far less so recently, raising the question of whether the original edge was real or just cyclical. Levered buybacks may have worked extremely well from 1980 to 2010 because of falling rates and strong industry tailwinds, but those conditions are much less supportive today. Buying stocks after a predicted risk materializes can feel rational, but in his experience it has often been one of his worst-performing patterns. Shorting AI names based only on extreme valuation can be dangerous because great businesses often look absurdly expensive before they become even larger winners. Past short sellers of Tesla, Netflix, Amazon, and Salesforce looked rational on valuation/accounting grounds and were still badly hurt by powerful business performance. Management-issued NAVs and long-term targets can be useful reference points, but they often fail to translate into returns because overhead, incentives, and market awareness are ignored. A stock trading below management’s stated NAV is not automatically a thesis; the discount may exist for a structural reason that the market already understands.

Data Points: Gambling budget: $300 - Walker’s self-imposed craps budget during Planet MicroCap in Las Vegas. One-off bet size: $20 - He says he sometimes bets $20 on a basketball game for entertainment. Hot streak example: 5x - His friend reportedly made about 5x his money on the first night at the craps table. Legal investing experience: First 7 years vs. past 3 years - Walker says legal special situations were mostly winners in his first seven professional investing years but losers over the last three. John Malone era: 1980 to 2010 - He cites this period as the prime era for Malone’s levered buyback model. Management target horizon: 3 years - He notes companies often publish three-year targets and frequently miss them. Spac bubble reference: 2021 - He uses 2021 SPAC projections as an example of overly aggressive long-term guidance. AI capex claim: $1 trillion - He references Sam Altman being questioned about a trillion dollars of capex over 10 years. AI company funding reference: $10 billion company - He describes OpenAI as a large company making very large capex promises. Liberty Sirius XM example: $4, $25, $40 - He says Sirius XM traded around $4/share, Liberty Sirius around $25/share, and implied NAV was about $40/share.

Pivotal Quotes: "wins that lure you into security" — Andrew Walker: The core theme of the episode and a repeated framework for discussing strategy overconfidence. "I am shorting the top of the dot-com bubble" — Andrew Walker: He contrasts current AI-bubble short arguments with historical shorting of bubble-era tech names. "management's got NAV at 10, the stock's at five, that's probably not a thesis" — Andrew Walker: His summary of why simple NAV discount stories often fail without a deeper catalyst or edge.

Implications: Listeners should treat recent success, management guidance, and valuation-based shorts with skepticism. Durable returns likely require a real edge, awareness of regime changes, and a catalyst—not just a cheap-looking spreadsheet or a popular narrative.