Goldman Sachs Warns: Forget AI Hype, Focus on the 'Halo Trade'

**Markets are hitting new highs, AI stocks are soaring—and Goldman Sachs is telling investors to sell into the rally.**  Last week saw U.S. stocks reach another peak, with AI-themed names leading the charge. But Goldman Sachs partner Bobby Molavi delivered a sobering message: the market is in a "delicate balance," and he'd "sell into strength" if forced to choose. This isn't just another Wall Street caution on overvaluation. Molavi's real insight cuts deeper: while everyone chases AI's "soft narrative," the durable money might be made in the **"halo trade"**—the physical, hard-to-replace assets that actually enable AI to function. --- ### **What's the Market Betting On? How Long the AI Story Lasts** Molavi puts it bluntly: market resilience has been "incredible." Oil is back to 1992 levels, rates have normalized (though cuts may come slower), European fiscal policy is supportive, and the AI capex story keeps rolling. The problem? **Most of this good news is already priced in.** The market is now betting on how far and how long the AI narrative can run. At these levels, volatility risk is building. Molavi isn't bearish on AI itself—he's bearish on chasing it at current valuations. --- ### **The 'Halo Trade': AI's Unseen Foundation** Molavi introduces a crucial theme: **"heavy assets, low obsolescence."** In areas where terminal value is hard to pin down (like some AI software plays), valuation gets fuzzy. But scarcity is being redefined. The world needs oil, critical materials, heavy freight—the "hard stuff" that was overlooked for years. This is AI's actual foundation. Think about it: chip manufacturing needs specialty gases, ultra-pure materials, precision equipment. These aren't software—they're physical assets with tight supply, limited capacity, and low substitutability. That's "low obsolescence." **The hotter AI gets, the scarcer these underlying assets become.** --- ### **The Real Split: 'Soft' Narrative vs. 'Hard' Reality** Molavi's logic highlights a deep market disconnect: AI's "soft story" has sprinted ahead, but the "hard assets" required to support it haven't kept up in price. The market is paying for AI's potential while overlooking the **real physical costs** needed to realize it. He notes that the value of these goods—"especially those that supply the chip-making ecosystem"—has been "validated" recently, and he expects this trend to continue. Key word: **"validated."** This isn't a prediction—it's already happening. Supply chain bottlenecks, geopolitical tensions, and capacity constraints are forcing a repricing of hard asset value. --- ### **What Comes Next? Watch the 'Hard Constraints'** For crypto and broader tech investors, this framework offers concrete takeaways: - **Liquidity narratives could flip.** If rate cuts delay and the focus shifts from "easy money" to "physical scarcity," Bitcoin's inflation-hedge narrative strengthens—aligning directly with the halo trade logic. - **AI speculation will bifurcate.** Pure-software, asset-light plays may see more volatility, but entities providing AI compute, storage, and energy (think mining hardware makers, data centers, power companies) could prove more resilient. - **Supply chain risk premiums rise.** Any industry reliant on global supply chains (including crypto mining) faces pressure from rising hard-asset costs, squeezing margins and raising barriers to entry. The unspoken subtext: when the story fades, what remains are the things you can't move, bypass, or easily replace. --- ### **The Bottom Line: Don't Just Watch AI Dance—Look at Who's Building the Stage** Goldman's warning is a reminder: don't get dazzled by the AI party. The market's delicate balance—AI optimism on one side, rate/inflation/geopolitical realities on the other—is fragile. Any shock could tip it. The halo trade matters because it doesn't rely on stories; it relies on **physics**. Chips need silicon, compute needs electricity, data needs storage. These demands won't vanish with sentiment swings. For investors, the key questions aren't about how much higher AI stocks can go, but: - Which assets represent **unavoidable costs** in the AI boom? - Which supply chain links are already under **real strain**? - If liquidity tightens, what actually holds value? Molavi's line says it all: **"Scarcity is being redefined."** In the age of AI, the scarcest resource might not be algorithms—it could be the power, hardware, and raw materials that make them run. The story may wobble, but the foundation won't collapse.