The AI Trade Is Finally Cracking | Weekly Roundup
The hosts discuss a significant momentum factor collapse in tech and AI-related trades, driven by cracks in the AI narrative (Meta considering selling compute access, claims of memory efficiency breakthroughs) coinciding with yen intervention and weakening economic data. They argue the Fed should remain dovish given falling wage growth, declining labor force participation, and peaking inflation, while positioning for a shift away from mega-cap tech toward gold and debasement trades.
Summary
The podcast episode opens with discussion of a dramatic four-sigma momentum factor implosion occurring simultaneously with yen intervention by Japan's Ministry of Finance. The hosts explain how reflexive flows into AI and semiconductor trades, fueled by currency weakness and momentum positioning, are unwinding together as the fundamental narrative supporting the trade weakens.
Two key catalysts triggered the selloff: Meta's consideration of selling excess AI compute access (challenging the 'infinite demand' thesis) and a tweet from a reputable analyst predicting memory efficiency breakthroughs from an OpenAI spinout. The hosts emphasize that positioning matters more than the truth of these claims—with leveraged positioning in place, any narrative crack suffices to trigger liquidations across correlated trades in semiconductors, Korean won, and Taiwan exposure.
Regarding Fed policy, both hosts argue strongly against rate hikes. They cite weak jobs report data: while unemployment ticked lower, this resulted from declining labor force participation rather than strong hiring, and wage growth remains subdued with no signs of wage-price spiral. They compare the current setup to July 2024, when the Fed hiked despite deteriorating data, then reversed at Jackson Hole. The hosts note the Fed's recent communications have become more hawkish in tone while incoming data (ISM PMIs, wage growth, job report) suggests peak growth and peak inflation are occurring now.
On positioning, the first speaker favors selling bounces in mega-cap tech rather than buying dips, given the reflexive nature of hyperscaler valuations (which benefited from AI lab private valuations earlier in the year). The second speaker takes a longer-term view, having bought in the second or third inning and planning to sell portions during any blow-off top, while maintaining conviction that Jevons Paradox and new use cases will support the AI thesis over multi-year horizons.
Both hosts discuss gold and 'debasement trades' as preferable expressions of their thesis given the shift away from hawkish rate expectations. They also note Bitcoin's recent inverse performance to the semiconductor selloff, with MicroStrategy shorting pressure easing after extreme bearish sentiment. They caution that while Bitcoin may bounce from force deleveraging, significant gains beyond the current range would require MicroStrategy to resume buying.
The discussion concludes with commentary on the broader market environment being 'treacherous' into year-end, with extreme crowding on both sides of trades (home builders, biotech, software versus semis). They touch briefly on crypto market quality issues (poorly structured token offerings like the Venice Finance incident) but express optimism that capital will eventually flow to properly incentivized protocols. The hosts advocate for vigilance against being caught in the last-bear squeeze while avoiding the tourist hype that characterizes market peaks.
About this episode
Markets can look strongest just before their underlying assumptions begin to crack. This week, we examine whether the AI-led growth narrative is beginning to unravel as positioning, macro data, and market structure all shift at once. We discuss the AI momentum unwind, yen intervention, NFP & labor market weakness, Fed policy, gold, Bitcoin, and what we think is the next major trade. Enjoy! TIMESTAMPS: 00:00 Intro 01:40 AI Momentum Starts To Crack 07:14 What Triggered The Selloff? 10:21 Is The AI Boom Peaking? 12:24 Why Gold Still Looks Attractive 14:19 Breaking Down The Jobs Report 18:06 Could Inflation Stay Sticky? 24:56 Is The Market Misreading The Fed? 27:54 How We're Trading This 32:47 Is Bitcoin Turning? 38:42 Why Gold Beats Tech 40:36 The Next Phase Of Crypto 44:17 July 4th And The World Cup FOLLOW THE SHOW › Forward Guidance – https://x.com/ForwardGuidance › Felix – https://x.com/fejau_inc › Quinn – https://x.com/qthomp › Telegram – https://t.me/+CAoZQpC-i6BjYTEx › Blockworks – https://x.com/Blockworks RESOURCES › Weekly Roundup Charts – https://drive.google.com/file/d/14q53P0qQcFvyvvSBgVSZ-jM_TZU9HEyP/view?usp=drive_link EVENTS › Join us at Digital Asset Summit 2026 Asia October 7th & Digital Asset 2026 London November 10-11th https://blockworks.com/events Blockworks recently acquired Messari. For more information, please visit: https://blockworks.com/insights/blockworks-acquires-messari DISCLAIMER Nothing said on Forward Guidance is a recommendation to buy or sell securities or tokens. This podcast is for informational purposes only. Any views expressed are opinions, not financial advice. Hosts and guests may hold positions in the companies, funds, or projects discussed.
Key Insights
- The hosts argue that momentum factor collapses happen simultaneously with geopolitical events (yen intervention) because the underlying trades are highly correlated on both the way up and down, meaning multiple dominoes trigger together rather than sequentially.
- The speakers contend that the truthfulness of catalysts matters less than the positioning they enter into—Meta's compute sale consideration only moved markets because of existing leverage, not because infinite AI demand was actually uncertain.
- Both hosts claim the Fed is currently committing the same policy error as July 2024 by maintaining hawkish guidance despite deteriorating labor data (falling participation, weak wage growth) and peaking economic indicators, risking another reversal.
- The first speaker argues that hyperscaler profits are reflexively tied to AI startup valuations on their balance sheets, so any AI trade unwind simultaneously harms both the 'payers' (hyperscalers) and 'receivers' (AI labs), creating correlated downside.
- The hosts contend that core inflation's stickiness is primarily due to uncontrollable acyclical components (supply shocks), while the cyclical components the Fed can influence (wage-driven demand) show no signs of overheating.
- The second speaker argues that the current market environment exhibits characteristics of a bubble, with two-to-three months of price action reversing in days across international semiconductor markets, plus pink-sheet pump-and-dumps and extreme profit expectations across all sectors.
- The hosts assert that Bitcoin's rally potential depends on MicroStrategy's forced deleveraging removing downward pressure rather than on positive fundamental developments, limiting upside unless the company resumes active buying.
- Both speakers contend the next two months until the mid-year earnings reset and fiscal stimulus fade represent the 'most treacherous' period, with crowding into risk-on positioning at exactly the moment when growth and inflation signals are peaking.
Topics
Transcript
There's a lot of cracks forming where the things that were pumping this to the upside aren't really working anymore. That's all you needed. These two headlines and suddenly you just have this like momentum factor implosion that we're seeing. Yeah, I would be very cautious here if I was loaded to the gills with tech risk. You're reaching this crescendo, what I think is peak growth and peak inflation for the economy. There's just no reason for the fed to act in either direction right now because of the labor market a lot of things are coming together to me that it's just like i want to be focused on trades that express nothing said on for guidance is…
Full transcript available for MurmurCast members
Sign Up to AccessMore from Forward Guidance
The Growth Strategy Trapping The Fed | Darius Dale
Darius Dale discusses the current macroeconomic landscape, highlighting the transition to a 'paradigm C' growth strategy characterized by running the economy hot. He emphasizes the challenges investors face amidst market volatility and explains the implications for asset allocation and risk management.
Washington Is Suppressing Volatility To Keep The AI Boom Alive | Weekly Roundup
The hosts discuss how Washington is actively suppressing market volatility to maintain the AI boom and economic stability through coordinated central bank and Treasury interventions. They analyze how policy coordination between the Fed and Treasury has shifted from Fed-driven liquidity to state-directed capitalism, with implications for everything from AI infrastructure investment to currency markets and geopolitical dynamics.
The Portfolio Built To Survive Every Crash | Jared Dillian
Jared Dillian discusses his new book 'The Awesome Portfolio,' a diversified 20/20/20/20/20 split (stocks/bonds/gold/cash/real estate) that has delivered 9% annual returns since 1971 with only 12% maximum drawdown and half the volatility of traditional portfolios. He also shares his contrarian views on Fed Chair Warsh's recent decision to hold rates steady, which he interprets as intentional yield curve steepening to tighten monetary policy while letting markets take the lead.
The AI Unwind And Warsh's Long-End Gamble | Weekly Roundup
The hosts discuss Leo Leopold's AI trade liquidation forced by Citadel, analyzing the role of leverage and market structure in the collapse. They then pivot to Kevin Warsh's recent Fed meeting, debating his hawkish communication on long-end bond repricing and balance sheet normalization, which triggered market selloffs and credibility concerns despite arguably achieving his intended tightening effects.
AI Efficiency Is Repricing The Compute Market | Steve Hou
Steve Hou from Silicon Data discusses how AI efficiency and model substitution are repricing the compute market. He explains that token expenditure has plateaued not due to declining demand, but because enterprises are rationally substituting expensive frontier models with cheaper alternatives, while overall inference demand remains robust and growing across all GPU tiers.