Your Budget Sucks (EP.397)

Episode Summary

Executive Summary: The episode centers on the market shock from DeepSeek and what it could mean for AI winners, especially Nvidia and the Mag 7, while arguing the selloff may be an overreaction and that cheaper inference could ultimately benefit big tech and consumers. The hosts also discuss rotation into non-tech stocks, crypto regulation, hedge fund fees, inflation/rent data, consumer stress, ETF/leveraged-product mania, and lifestyle topics like remote work, private chefs, and movies.

Main Topics: DeepSeek and the AI market selloff (Priority: 5/5): The hosts analyze the abrupt drawdown in Nvidia, semiconductors, and the broader AI trade after DeepSeek's emergence. They debate whether the news is bearish for chip demand or simply a catalyst for market rotation, and lean toward uncertainty rather than a definitive verdict. AI winners, losers, and model commoditization (Priority: 5/5): They discuss Ben Thompson's argument that cheaper inference and open-source model commoditization could favor big tech distribution platforms like Microsoft, Amazon, Apple, and Meta more than pure model or hardware leaders, while potentially pressuring Google and possibly OpenAI. Market breadth and rotation away from tech (Priority: 4/5): The episode highlights unusually strong breadth: the S&P 493 outperforming, value and defensive sectors holding up, and a possible shift from concentrated Mag 7 leadership into small caps, international stocks, dividends, and staples. Bubble behavior vs. valuation (Priority: 4/5): The hosts distinguish between a true bubble and pockets of mania, citing Howard Marks' view that current markets may look rich but lack the broad behavioral excess of prior bubbles. They note examples like meme coins, quantum stocks, and leveraged ETFs. Crypto, regulation, and token proliferation (Priority: 4/5): They cover Brian Armstrong's claim that a million new crypto tokens are created weekly and discuss the flood of speculative crypto ETF filings, including levered products tied to meme coins and political tokens. Macro, consumer data, and household finances (Priority: 3/5): The episode reviews falling rent inflation, lower GDP volatility, rising credit-card minimum payments, and a debated Brooklyn household budget, using these to question how much economic stress is actually broad-based. Personal/lifestyle and culture recommendations (Priority: 2/5): The hosts end with lighter banter on Naples, Future Proof, remote work preferences, private chefs, dishwashing, parking etiquette, and movie reviews including Gladiator 2, Goonies, A Real Pain, and Meet the Parents.

Key Arguments: DeepSeek's lower-cost AI approach may be bearish for Nvidia in the short run, but it could be positive for consumers and platform companies over the long run. The market reaction may be an overreaction; the hosts expect uncertainty, not a clear one-way outcome, and reject confident predictions about AI winners. A rotation out of concentrated tech leadership and into the rest of the market would be healthy and potentially the most beloved bear market setup ever for active managers. A true bubble requires a behavioral mania, not just high valuations; current markets show pockets of excess but not full-blown dot-com-style euphoria. Cheaper AI inference should reduce data-center and GPU spending for some firms while increasing usage and strengthening distribution-driven businesses. The current crypto environment is structurally different because supply is exploding, but demand may still concentrate in a few dominant assets like Bitcoin. Economic volatility has been dampened by policy and structural changes, and falling rent data argues against an immediate inflation resurgence. The share of consumers making only minimum credit-card payments is concerning, but still within a historical range and may not be enough to drive macro deterioration alone. Leveraged ETFs and speculative products are trading tools, not buy-and-hold investments; long-term holding can be devastating due to path dependency and drawdowns. Remote and hybrid work remain highly valued, reflecting a durable post-pandemic shift in how people think about time, commuting, and productivity.

Data Points: China retail web sales (2023): $2.1 trillion - Used in sponsor read contrasting China's online commerce scale with the U.S. U.S. retail web sales (2023): $1.1 trillion - Sponsor read comparison point for U.S. e-commerce scale. China total retail sales (2023): $6.5 trillion - Sponsor read discussing China's consumer market size. China internet population (2023): over 1 billion - Sponsor read noting internet adoption in China. China internet penetration: 77% of population - Sponsor read comparing China's penetration to the U.S. U.S. internet penetration: 93% of population - Sponsor read comparison with China. Nvidia intraday decline: almost 20% - Market reaction to DeepSeek fears and AI capex concerns. Semiconductor ETF decline: down 10% - Broad selloff in chip stocks during the AI shock. S&P growth vs. value: growth down almost 4%, value up 1% - Shows rotation away from growth into value during the selloff. S&P 493 performance: up while the index fell - Breadth improved even as major tech names sold off. Retail Nvidia buying: $562 million - Retail investors bought the dip in Nvidia, per Vanda Research. Minimum credit-card payment share: 11% - Share of cardholders making only minimum payments, near the highest since 2020. Brooklyn household take-home pay: $25,000 per month - Example budget from The Purse discussion. Brooklyn household mortgage: $6,500 per month - Part of the debated household budget. Brooklyn household retirement contributions: almost $2,500 per month - Used to judge saving rate for a high-income household. Egg prices: up about $4 per dozen year over year - Example of inflation people notice even if the dollar impact is modest. Current rent inflation proxy (NTRI): -2.5% - Colin Roche cited this as leading indicator suggesting inflation cooling. Insider buy/sell ratio: 98 companies with at least one insider buy vs 447 with at least one insider sell - Kevin Gordon's note that insider buying is at a record low pace. Nvidia 3x levered ETF move: down over 50% - Illustrates the danger of leveraged single-stock ETFs. UPRO 3-year return: 70% - Compared with S&P 500's 45% over the same period. S&P 500 3-year return: 45% - Benchmark for UPRO comparison. U.S. GDP volatility: much lower than historical levels - Goldman chart on rolling 20-year volatility of real GDP growth. Crypto token creation: 1 million new coins per week - Brian Armstrong's claim on Coinbase regulation challenges. Work preference poll: Hybrid 53%, remote 33%, in-office 15% - YouTube audience survey on preferred work environment. This month coverages of AI market breadth: more than 300 stocks rose despite S&P down nearly 1.5% - Jason Goepfert sentiment stat on unusual breadth. Hedge fund fees collected since 1969: $1.8 trillion - Reuters stat referenced in discussion of industry economics. Klarman annualized return since 2014: 4% per year - Bloomberg piece cited on Baupost's recent performance. Klarman pre-crisis annualized return: 20% over 26 years - Context for Seth Klarman's long-run reputation. Classical daily commute savings: hours per week saved - Qualitative but emphasized as a major benefit of remote/hybrid work.

Pivotal Quotes: "A world where Microsoft gets to provide inference to its customers for a fraction of the cost means that Microsoft has to spend less on data centers and GPUs." — Ben Thompson (quoted by hosts): Explaining why cheaper AI inference may benefit large platform companies. "I think that DeepSeek has provided a massive gift to nearly everyone." — Ben Thompson (quoted by hosts): Summarizing the long-run bullish case for consumers and big tech. "A bubble is a temporary mania in which people are so agog at things that they throw over all discipline, all caution." — Howard Marks (paraphrased/quoted by hosts): Used to argue that current market excess may be pockets of mania rather than a full bubble.

Implications: Listeners should expect continued volatility as markets price a possible AI disruption, but also a broader rotation opportunity beyond tech. Cheap AI, strong breadth, and falling rent inflation could support consumers and large platforms, while speculative excess in crypto and leverage remains risky.