Never Go All in on Stocks (EP. 437)

Episode Summary

Executive Summary: The episode centered on market concentration, AI enthusiasm, and how those forces are reshaping investing, risk, and business behavior. Ben and Michael debated whether mega-cap tech and AI are in a bubble, argued that index concentration is distorting breadth signals, and emphasized that diversification and cash/bonds still matter. They also covered tariffs, demographic inequality, data-center capex, AI’s impact on jobs, Bitcoin/Strategy’s financing model, housing bifurcation, and consumer strain.

Main Topics: Tech concentration and distorted market breadth (Priority: 5/5): The hosts opened with how a handful of mega-cap tech stocks are driving index returns, creating strange divergences where many stocks fall while broad indices hold up. They argued that traditional breadth signals are less useful in a market dominated by the top names. Whether AI is a bubble (Priority: 5/5): They debated the AI investment cycle from both valuation and capex perspectives. Michael leaned toward calling it bubble-like because expectations and spending are enormous, while Ben argued the term bubble is too rigid and that future outcomes could still justify today’s investment. Diversification, sequence risk, and retirement planning (Priority: 5/5): Jason Zweig’s argument for maintaining bonds/cash was discussed as a reminder that stock-market upside is not guaranteed. The hosts emphasized sequence-of-returns risk and the danger of being overexposed to equities near retirement. Data-center buildout and real-economy consequences (Priority: 4/5): They discussed the massive construction of AI infrastructure, especially Meta’s Louisiana project, as evidence that the AI boom is not just a stock-market story. They highlighted knock-on effects for construction, farmland, water access, and industrial activity. AI, jobs, and chart skepticism (Priority: 4/5): They pushed back on claims that AI is already causing a collapse in job openings, calling some charts a 'chart crime' because they conflate causation with correlation and ignore prior labor-market tightening and post-COVID normalization. Strategy/MicroStrategy, Bitcoin, and capital structure risk (Priority: 4/5): A long segment covered Strategy’s falling premium to Bitcoin, its B-minus credit rating, and Michael Saylor’s view of Bitcoin as a superior risk framework. The hosts were skeptical of the company’s narrative and highlighted the mismatch between Bitcoin assets and dollar liabilities. Housing, consumer bifurcation, and demographics (Priority: 4/5): They noted the growing K-shaped economy: million-dollar homes are taking a larger share of sales while first-time buyers are older. The discussion broadened into generational inequality, younger workers’ struggles, and the role of parental wealth transfers.

Key Arguments: Market breadth looks weak, but in a highly concentrated index it often becomes a coincident indicator rather than a useful bearish signal. A few mega-cap stocks can keep indices elevated even when most stocks are down, so equal-weight and internals matter more than headline index levels. The AI buildout is too large and too capital-intensive to dismiss as pure hype, but that does not automatically mean a classic dot-com-style collapse is imminent. Calling AI a bubble depends on definition; a bubble need not imply permanent failure, only expectations that outrun plausible fundamentals. Investors and retirees still need bonds/cash because stock returns are path-dependent and sequence risk can permanently alter retirement outcomes. The data-center boom is a real economy phenomenon with major spillovers to construction, farmland, and utilities, not just a market narrative. Claims that AI immediately caused job-opening declines are overstated; the labor market was already cooling for other reasons. Strategy’s business model relies on leverage, market confidence, and favorable capital access, but the market may no longer be willing to pay a large premium for Bitcoin exposure through the stock. Consumer weakness is increasingly bifurcated: wealthier households still spend heavily while lower-credit and younger households face more pressure. Demographic timing, delayed adulthood milestones, and parental wealth/inheritance are becoming increasingly important to economic outcomes.

Data Points: Heirs firing parents' advisor: 4 out of 5 - Used in sponsor copy on wealth transfer and advisor retention Wealth transfer expected over next 20 years: $84 trillion - Sponsor mention about the coming great wealth transfer Stocks advancing over a two-day period: 227 stocks advancing; roughly 730+ declining - Illustrated market concentration despite broad weakness Average SP return on similar breadth-weak days: -2% - Historical comparison when decliners vastly outnumber advancers NVIDIA weight in S&P 500: Almost 10% - Example of index concentration Russell 2000 year-to-date return: About 12%-13% - Compared with other equity segments during the year Equal-weight S&P 500 year-to-date return: About 9% - Showed divergence from cap-weighted index Mid-cap return year to date: About 5% - Illustrated smaller-cap underperformance relative to small caps Emerging markets return year to date: 34% - Highlighted strength of international diversification MSCI EAFE / international developed return year to date: About 28% - Used as evidence that non-U.S. diversification is helping NASDAQ 100 return over past 10 years from 2009 bottom: About 22% annually - Compared with historical melt-up periods S&P 500 return in the 1950s: 491% for the decade - Used as a historical non-crash melt-up example Average decline examples during 1950s bull market: 14%, 28%, 22% - Corrections in 1959, 1961-62, and 1966 Top tech valuation context: Current big-tech PE cited at 22.8x - Compared with long-term average and margin-adjusted view Average margins used in valuation comparison: 14.5% currently vs 12.3% 10-year average - Duality Research’s margin-adjusted valuation argument OpenAI revenue vs valuation context: $13 billion revenue and $1 trillion market cap - Used as a provocative example of valuation expectations Private-sector job churn in Q4 2024: 7.5 million jobs destroyed; 7.7 million jobs gained - Showed that headline layoffs can obscure large labor-market churn Unemployment rate: 4.3% - Used to temper AI-job-loss alarmism Strategy market cap relative to Bitcoin holdings: Peak premium around 3.5x; now about 1.2x - Showed collapsing premium in MicroStrategy/Strategy stock Strategy credit rating: B-minus issuer credit rating - Discussed on earnings call and in relation to access to capital Meta Louisiana data center: 4 million square feet; over 2 gigawatts of compute - Example of AI infrastructure scale McKinsey estimate for data-center spending: Almost $7 trillion - Projected capital needed to keep pace with compute demand Apple market cap: About $4 trillion - Used to contextualize the scale of Apple and its businesses Apple iPhone revenue vs peers: More revenue than Bank of America over 12 months - Highlighted scale of iPhone business Apple Services revenue: $109 billion - Used to compare against Target and underline profitability Apple Wearables revenue: $36 billion - Compared with Starbucks revenue Apple Mac revenue: $34 billion - Compared with Schwab revenue Apple iPad revenue: More revenue than AMD - Used to illustrate magnitude of Apple’s segments Tariff charges on invoices at RAMP: Nearly 3% in 2025 vs 1.4% in 2024 - Showed tariff costs are rising slowly and remain limited in scope Top 10% of earners’ consumption share (viral claim): 50% claimed; economist argued more like 35%-40% disposable consumption - Debate over wealthiest households’ share of spending Households with subprime credit scores: 25% of U.S. population - Used to underscore financial fragility and consumer bifurcation First-time homebuyer age: 38 years old - Illustrated delayed homeownership Million-dollar home sales share: 20% of total - NAR data showing higher-end housing strength Sub-$250k home sales share: 8% of total - Showed weakness at the lower end of housing Ridiculousness host pay: $32.5 million per year - Mentioned as an entertainment-industry anecdote Number of Ridiculousness episodes per year: 336 - Illustrated scale of MTV content production

Pivotal Quotes: "This unprecedented concentration that we're seeing is going to keep producing these very weird mismatches between returns and internals." — Duality Research (quoted by Michael): Used to frame why breadth indicators are less meaningful in today’s cap-weighted market "I’d be way more comfortable saying the returns are going to be far lower in the future than they are going to crash." — Michael Batnick: His 'Grand Rapids hedge' on the likelihood of a crash versus a lower-return future "The second innovation is where we're replacing traditional risk with digital risk." — Michael Saylor: Part of his explanation for Strategy’s Bitcoin-centered risk framework

Implications: Investors should expect continued index distortion, be skeptical of simplistic bubble narratives, and keep risk management front and center. The AI boom is real but uneven, and wealth, housing, and labor outcomes are becoming more polarized.