TIP823: From Railroads to AI: The Timeless Patterns Behind Market Bubbles w/ Kyle Grieve

— Bubble formation mechanics: Bubbles are driven by human psychology (greed, fear, FOMO) rather than technology itself. They follow predictable patterns across centuries—from plank roads in the 1850s to Beanie Babies to the tech bubble—making pattern recognition a practical risk-management tool.

— Kindleberger's five-stage framework: Displacement (triggering event), overtrading (volume spike), monetary expansion (easy credit), revulsion (initial pop), and discredit (sentiment reversal). Insana refines this with five bubble ingredients: Eureka moment, easy money, government largesse, favorable economic conditions, and external stimulant.

— Why investors rationalize bubbles: New generations believe they're smarter than predecessors; career incentives push fund managers to own hot stocks; investors invent useless KPIs (page views, user growth) when fundamentals don't exist; herd behavior drowns out skeptics.

— Historical case studies: Yahoo required 18 billion customers to justify its peak valuation; RCA stock multiplied 120x (1923–1929) while trailing P/E expanded from 15x to 285x despite solid 35% annual EPS growth; plank road companies promised 10–40% dividends but delivered 0.7% and capital destruction.

— AI bubble assessment using Insana's framework: AI has Eureka moment, easy money (though rates higher than 2020), and government support (CHIPS Act). But economic conditions are weaker (2.1% GDP growth, 4% unemployment), and mass public participation hasn't yet peaked in IPO markets. Private deals (Thinking Machines at $50B without product) show speculative froth.

— Protective strategies: Cap exposure to narratives; verify winners compound intrinsic value, not just multiple expansion; perform terminal value reality checks (does the business's growth path make mathematical sense?); compare current multiples to historical averages; avoid pure AI build-out investments; focus on profitable businesses already leveraging AI today.

— Separate three layers of risk when evaluating AI exposure: (1) Technology will transform society—true—but technological success ≠ investor returns. Avoid speculative pre-revenue buildout plays. (2) Business layer is winner-take-most; 99% of early-1900s US car makers failed. Demand proof of competitive moat or durable profitability, not just narrative. (3) Value layer is where bubbles inflate. Compare current multiples to 10–20 year historical averages; when trading 2x historical average, treat as warning.

— Reality-check terminal value math before buying: If you believe a business compounds intrinsic value at 15% annually, map what it must accomplish (e.g., three to five acquisitions per year) and honestly assess if that's plausible. Thinking Machines at $50B valuation needs $5B ARR within years with zero product—mathematically implausible unless you have exceptional conviction; if you don't, pass.

— Monitor objective deal signals in your portfolio's industry: Are M&A prices, debt coupons, and private fundraises reasonable or narrative-driven? Example: Meta, Alphabet, Oracle issuing 30-year bonds at 5.7% coupon (only 1% above Treasuries) despite uncertain ROI on $5T AI capex suggests speculative investor appetite but not yet panic-level euphoria. Track these alongside media mentions and dinner-party tips; when neighbors brag about holdings before fundamentals improve, a top is near.