The AI Panic Trade | What the Viral Doomsday AI Article Means for Markets

Episode Summary

Executive Summary: The episode launches Last Call with a humorous, research-driven format focused on AI, private credit, market flows, and global rotation. Guests argue that Citrini’s AI bear-case story is best read as scenario planning, not prophecy; private credit can break from classic credit-cycle stress even without an AI apocalypse; and options/flow data show differentiated selling in software rather than a uniform collapse. The show also highlights global asset rotation, gold miners, and the need to use probabilities and base rates rather than one-line narratives.

Main Topics: Launch of Last Call and the show’s premise: The hosts introduce the podcast as an alternative to standard market recap shows, emphasizing unique perspectives, data, and a lighter tone while previewing guests Ben Hunt, Brent Kachuba, and Kai Wu. Citrini’s AI bear-case as scenario planning: A central discussion centers on the widely discussed Citrini piece, which is described as a fictional, future-looking bear case that sparked reactions across finance. The hosts frame it as a useful exercise in thinking through AI extremes rather than a prediction. Private credit vulnerability in a classic credit cycle: Ben Hunt argues that private credit does not need an AI apocalypse to suffer; normal boom-bust credit dynamics, refinancing pressure, and leveraged structures can trigger defaults and contagion, especially after years of easy money. Market flows and options positioning in software stocks: Brent Kachuba explains that options prices and flows show real investor behavior, revealing extreme bearishness in certain software names around earnings and suggesting that some of the move may already be priced in. AI, labor replacement, and data limitations: The conversation repeatedly returns to uncertainty about AI adoption versus capability. The hosts argue that technology growth does not automatically translate into rapid enterprise adoption, labor displacement, or measurable macro effects yet. Global rotation, gold, and the ‘hard money’ trade: Rupert Mitchell and Meb Faber discuss U.S. underperformance versus international assets, government spending as a driver of equity returns, and the appeal of gold miners, foreign value, trend following, and diversification amid changing market leadership. Base rates, bubbles, and what ‘this time is different’ really means: Meb Faber ties together Schroders, Mobius, and Grantham-style research to argue that investors should use historical base rates, diversification, and rebalancing to navigate bubbles without missing long-term upside.

Key Arguments: Citrini’s piece is best understood as a manufactured scenario from the future meant to stress-test assumptions, not as evidence of imminent collapse. Extreme bear cases can be educational because they force investors to think through probabilities, ranges of outcomes, and what could happen to AI, labor, inflation, and valuations. Private credit does not require an AI-driven catastrophe to break; ordinary credit-cycle tightening, higher rates, refinancing stress, and leverage embedded in BDCs/CLOs can create losses and contagion. Lenders eventually stop refinancing when they prefer a certain manageable loss over a potentially existential future loss; that moment may already be appearing in private credit. Options data are more informative than headlines because they show what investors are actually paying to hedge or speculate; in software, the bearish move looks differentiated rather than universal. After earnings, volatility often compresses, so some software names may have already seen the worst of the short-term de-rating if the market was overly fearful. AI could both enhance labor productivity and replace labor, but no one knows the ultimate balance; adoption is likely to lag technical capability. U.S. market leadership is no longer the only game in town; global fiscal spending, weaker dollar dynamics, and cheaper foreign markets are drawing capital abroad. Gold miners remain attractive because of strong margins, low all-in sustaining costs, and relatively cheap valuations versus previous cycles. Investors should favor diversified portfolios, strategic rebalancing, and trend following rather than trying to time the top of a bubble. Base-rate thinking suggests that very high growth forecasts for AI companies should be treated skeptically unless historical comparables support them. Many macro narratives are backward-looking; corporate filings, employment data, AI adoption metrics, and energy/compute constraints are better forward-looking indicators. metric value context

Data Points: Private credit default rate scenario: 15% - Ben Hunt references UBS estimates of potential default rates in private credit as a sign that loan books can become stressed. AI/market commentary response: 1 Bloomberg headline - The hosts note that Citrini’s piece became significant enough to generate Bloomberg coverage and responses from major figures. Software ETF support level: 6,800 to 7,000 on the S&P 500 range - Brent Kachuba says the market has largely traded in this range over the last two months. Salesforce buyback: $50 billion - Mentioned as a notable corporate action amid concerns about AI pressure on software. Software options fear ranking: IV rank 95% and risk reversal rank 9% - Brent describes Adobe/Autodesk as extremely bearish in options pricing. Earnings-related volatility expectation: 8% to 10% moves - Brent notes that some software names are pricing in unusually large earnings moves. Market-cap concentration: Above dot-com levels in companies trading at 10x sales - Meb cites Grantham-related work showing valuation extremes reminiscent of prior bubbles. OpenAI revenue expectation: 100% annual growth through 2029 - Referenced in Meb’s discussion of Mobius’s base-rate analysis. Gold miners all-in sustaining cost: About $1,500 per ounce - Rupert Mitchell cites AISC as a key support for mining profitability. AI data centers electricity consumption: Mid-to-high single digits, heading toward double digits - Discussed as a potentially inflationary input to the AI buildout. U.S. vs. global market cap: U.S. around two-thirds of world market cap - Meb uses this to illustrate how leadership can swing dramatically over decades. Software hiring surge: 2 to 3 times national average for four years - Referenced when discussing how pre-COVID/early post-COVID hiring distorted labor comparisons. Gold mining valuation comparison: Massive discount vs. 2011 peak - Rupert says miners remain cheap relative to the last gold peak.

Pivotal Quotes: "This is not some piece of like new evidence. This is literally a manufactured story from the future." — Host discussion of Citrini: Describing Citrini’s AI bear-case article and why it went viral. "The dirty little secret is that equity markets go up where governments are spending money, right?" — Rupert Mitchell: Explaining the global rotation and why foreign markets may be attractive. "There comes a time in every credit cycle where the money... they say, No more." — Ben Hunt: On why private credit can crack without an AI apocalypse.

Implications: Listeners are urged to think in scenarios, not certainties: watch credit refinancing, options/flow data, AI adoption and labor trends, and global capital rotation. The episode suggests the next market leadership phase may favor select software winners, gold, and non-U.S. assets rather than a single broad U.S. AI trade.