Investors should exercise extreme caution with overvalued semiconductor leaders like NVIDIA (NVDA), Broadcom (AVGO), and AMD (AMD), as massive bearish put options suggest a potential short-term top in chip valuations. Consider rotating profits away from hardware and into the "AI energy" trade, focusing on companies that control power assets and data center infrastructure. CleanSpark (CLSK) represents a high-conviction long position, pivoting from Bitcoin mining to essential AI infrastructure. Look for opportunities in specialized energy and grid providers like CoreWeave and Iron, as electricity has replaced GPUs as the primary bottleneck for AI growth. The most actionable strategy is to prioritize the "electrons" over the "chips" by investing in the power grid and nuclear energy sectors required to fuel the next phase of the AI supercycle.
NVIDIA (NVDA), Broadcom (AVGO), AMD (AMD), Oracle (ORCL), ASML (ASML)
• Leopold Ashenbrenner has taken a massive bearish stance against the major semiconductor and cloud infrastructure players. • His recent 13F filing reveals $7.5 billion in put options, which are essentially bets that these stock prices will decline. • This is not necessarily a bet against AI technology itself, but a bet against "chip multiples"—the idea that these stocks have become overvalued relative to their actual earnings potential.
Takeaways
• Exercise Caution in Semi-Cap: The "easy money" in the primary chip trade may be over. Investors should be wary of high valuations in the semiconductor sector. • The "Michael Burry" Risk: While these are massive bets against the market leaders, Ashenbrenner is hedging his portfolio. Retail investors should look at these puts as a signal of a potential short-term "top" in the AI hardware cycle.
CleanSpark (CLSK)
• This is a significant recent addition to Ashenbrenner’s portfolio, where he added the majority of his long positions. • While traditionally known as a Bitcoin miner, the company is being positioned within the framework of the "AI energy" trade.
Takeaways
• The Pivot to Infrastructure: Look for companies that own power assets or high-density data centers. • Strategic Accumulation: Ashenbrenner is aggressively long on this asset, suggesting he sees it as undervalued compared to the overextended chip manufacturers.
Energy & Power Infrastructure (Lumenergy, Sandisk, CoreWeave, Iron)
• The core thesis of the "Ashenbrenner Framework" is that the primary bottleneck for AI is no longer GPUs (chips), but "electrons" (energy). • He remains "all in" on energy, maintaining long positions in companies like Lumenergy, Sandisk, CoreWeave, and Iron. • The transition from a "chip-centric" AI trade to an "energy-centric" AI trade is the key driver of his $13.7 billion portfolio.
Takeaways
• Follow the Bottleneck: In any massive tech cycle, the most profit is made by identifying the scarcest resource. Currently, that resource is the power required to run AI models. • Sector Rotation: Consider rotating some profits from hardware/chips into energy providers, grid infrastructure, and specialized data center operators (like CoreWeave). • Long-Term Bullishness: Despite the bets against chip stocks, the overall sentiment on the AI "supercycle" remains extremely bullish; the investment focus is simply shifting to a different part of the supply chain.
Investment Framework: The "AI Supercycle"
• The transcript highlights that copying specific trades from a 13F filing is risky due to the multi-month reporting lag and high market volatility. • The "Alpha" (market-beating edge) lies in adopting the framework: AI growth is inevitable, but the infrastructure to power it is currently underpriced compared to the chips that run it.
Takeaways
• Don't Chase Lags: Avoid blindly buying what a hedge fund bought three months ago. Instead, look for new opportunities within the "Energy for AI" theme. • Focus on Scarcity: As AI models grow, the demand for electricity will skyrocket. Investments in the power grid, nuclear energy, and renewable energy storage are the logical "next step" for the AI trade.
