Leopold’s Situational Awareness LP fund, the darling of U.S. IPOs, has completely transformed its Q1 2026 portfolio (see Figure 1)! 🧐 I double- and triple-checked the data—multiple times—thinking I’d misread it. The holdings in Figure 1 genuinely shocked me. I was so alarmed that I immediately sold nearly half of my U.S. AI stock positions tonight! This timing is highly suspicious. First, the 10-year U.S. Treasury yield has surged dramatically to 4.62%. Combined with NVIDIA’s upcoming major earnings report after Wednesday’s close, the sudden large-scale placement of put options feels like a harbinger of something big. Remember: Last Q4, Leopold was heavily invested in AI infrastructure. Now, he’s gone full bearish. What could he possibly be seeing? 🥸 🎯 First, examine the top holdings shift—it’s earthquake-level shocking: 📊 Q1 2026 Holdings (as of March 31): • SMH PUT (Semiconductor ETF short) – 14.94%, $2.04B market value • NVDA PUT (NVIDIA short) – 11.47%, $1.57B market value • ORCL PUT (Oracle short) – 7.84%, $1.07B market value • AVGO PUT (Broadcom short) – 7.36%, $1.00B market value • AMD PUT (AMD short) – 7.09%, $970M market value You read that right—the top five holdings are all put options. Honestly, I verified this data at least five times. What does this mean? Leopold is betting real capital against the entire tech and semiconductor sector. (It could also be a complex options hedging strategy to protect gains—but overall, puts significantly outnumber calls.) 🎯 Core Strategy Shift: Q4 2025 Holdings Logic: • Heavy allocation to AI infrastructure: Bloom Energy (power), CoreWeave (compute), Lumentum (optical communications) • Core thesis: The bottleneck in AI development isn’t chips—it’s power and compute infrastructure • Strategy: Sold NVIDIA, TSMC, and other chip stocks; shifted capital upstream to energy and data centers Now, Q1 2026 Logic: • Full-scale shorting of tech giants and semiconductors! • Shorted names: #SMH, #NVDA, #ORCL, #AVGO, #AMD, #MU, #TSM, #ASML, #INTC • Yet retains select call options for hedging: #MU, #SNDK, #TSM calls This is a complete 180-degree reversal! 🧐 My interpretation and thoughts: Leopold’s move signals three critical messages: 1️⃣ AI Bubble Theory? He may believe current AI stock valuations have wildly overpriced future expectations. While he still believes AGI will emerge by 2027–28, the market has already priced in that future. Look at NVIDIA, AMD, and Broadcom’s current P/E ratios—markets are pricing in perpetual growth. But Leopold may be seeing a turning point. 2️⃣ Shift from Infrastructure to Chip Shorting Last Q4, his thesis was: Chips aren’t the bottleneck—power is. So he exited chip stocks and doubled down on energy. Now he’s gone further: Not only are chips not the bottleneck—chip stocks are wildly overvalued! So he’s shorting them outright. Yet he still holds significant positions in energy and compute infrastructure: BE (Bloom Energy) at 6.42%, IREN (Bitcoin miner turned AI compute) at 2.93%, CORZ (Core Scientific) at 2.84%. This confirms his long-term belief in AI infrastructure remains intact—he simply believes chip stocks are being priced irrationally. 3️⃣ Hedging or Pure Short? Note this detail: Alongside his puts, he retains select calls: • MU CALL – 3.09% • SNDK CALL – 2.84% • TSM CALL – 2.59% This is a straddle strategy—buying both puts and calls. He’s betting on volatility: whether the market surges or crashes, as long as movement is large enough, he profits. But puts vastly outweigh calls—suggesting his directional bias is strongly bearish. 📉 Risks and Controversies: Leopold’s strategy carries enormous risk: • Timing risk: Options expire. If markets don’t drop before expiration, he loses all premium paid. • Counterparty risk: He’s shorting some of the world’s most fundamentally strong tech companies. • Market sentiment risk: The AI boom continues; the Trump administration is aggressively pushing AI development. Shorting tech stocks is going against the tide. But Leopold is sharp—he has exceptional situational awareness. I believe he sees things most of us don’t. Possible catalysts: • SpaceX’s landmark IPO in mid-June could drain significant market liquidity • AI capital spending may have peaked (Microsoft, Google, Meta’s AI investment growth may slow) • Geopolitical risks (escalating U.S.-China tech war, semiconductor supply chain restructuring) • Macroeconomic pressures (recent surge in Treasury yields;投行 forecasts suggest the Fed won’t cut rates this year—and may even hike—raising costs for capital-intensive AI projects) 🎯 What Should Retail Investors Do? Don’t blindly copy him! Leopold runs a hedge fund—he uses complex options strategies and has a professional risk management team. Retail investors lack that capability and resources. But you can learn from his mindset: • Stay skeptical: Don’t blindly believe “AI always goes up.” All assets cycle—even the best companies can become overvalued. • Monitor industry shifts: The AI supply chain is structurally evolving. Where are the bottlenecks? That’s where opportunity lies. • Hedge your risk: If you’re heavily invested in tech stocks, consider trimming positions on rallies or allocating to safe-haven assets like gold, Treasuries, or cash. • Think long-term, trade short-term: You can still believe in AI’s long-term potential while preparing for near-term corrections. Set stop-losses and take-profit levels—don’t hold through drawdowns blindly. In my view, Leopold’s move will likely be painful in the short term (1–3 months). If tech stocks keep rising, his puts will rapidly lose value. But over a 6–12 month horizon, he might be right. Market optimism around AI has reached euphoric levels. Any negative catalyst—a major tech firm’s AI ROI falling short, or an AI safety incident—could trigger a cascade. And don’t forget: Last H1 2025, his fund returned +47%, far outpacing the S&P 500’s +6%.He has this track record, and his judgment deserves our attention. To summarize: Situational Awareness LP’s Q1 2026 positions reveal the key signals: • Full-scale shorting of tech and semiconductor stocks (7 of the top 10 are PUTs) • Retention of AI infrastructure (energy, computing power, data centers) • Use of options strategies to hedge risk and bet on significant market volatility • Shift in logic from “chips aren’t the bottleneck” to “chip stocks are severely overvalued” — an extremely aggressive and bold pivot. The next 2–3 months will be the critical period to test Leopold’s thesis. Stay tuned! 🧐