The SEC filing landed like a cold diagnostic readout. Third Point LLC, Dan Loeb's hedge fund, quietly offloaded its stake in Lam Research. No fanfare. No press release. Just a data point buried in a 13F. But for anyone who reads the gas of a capital cycle, this isn't noise. It's a signal. A signal that the AI infrastructure buildout—the very engine driving Lam's $17.4B revenue—may be hitting a velocity wall.
Lam Research sits at the choke point of the semiconductor equipment food chain. It's the dominant supplier of etch and deposition tools for 3D NAND, HBM, and advanced logic. When a foundry or memory maker expands capacity, Lam ships the machines that make the chips. Its revenue is a direct readout of global wafer fab equipment (WFE) spending. And WFE spending has been on a rocket ride fueled by AI: cloud capex hit $200B+ in 2024, with another 30% growth expected in 2025. But Third Point is saying: the fastest part of the climb is done.
Let's cut through the narrative. The fund's move isn't a bet against Lam's technology. The etch-deposition fortress is real. Lam's high-aspect-ratio etch for 200+ layer NAND and its TSV etch for HBM are best-in-class. Switching costs are brutal. The moat is deep. What Third Point is betting against is the valuation of a cyclical peak. Lam trades at 30-35x trailing earnings, a premium to its historical 25x average. The AI premium has been baked in. Every future growth assumption is already in the price. The problem is that growth assumptions are about to be tested.
Here's the core mechanic: equipment orders lead fab capex by 12-18 months. Right now, the order book is fat. But the marginal drivers are shifting. U.S. export controls have structurally compressed Lam's China revenue—from 29% of sales in FY2021 to ~20-25% today. The China gap is being partially filled by U.S., Korea, and Europe expansion, but the replacement is not one-to-one. The incremental growth from ex-China markets is lower-margin and slower to ramp. The result is that Lam's revenue growth rate is likely to decelerate from 2025 onward, even as AI demand remains strong in absolute terms. The gas isn't the problem. It's the friction of poor architecture. In this case, the architecture is the global semiconductor supply chain, fractured by geopolitics.
But there's a contrarian angle most retail analysts miss. The real risk isn't Lam's product line—it's the shape of the AI capex curve. Cloud providers are pouring billions into GPU clusters. But equipment sales are a second-derivative signal. When cloud capex growth slows from 30% to 15%, equipment orders will contract faster than chip sales. Vulnerabilities aren't always in the code. Sometimes they're in the timeline. Third Point is front-running a potential 2026 capex correction. If AI returns on investment disappoint—a scenario that's entirely plausible given the enormous upfront spend—the capex haircut will hit equipment names first. Lam will feel it before NVIDIA does.
And yet, the bear case is not a death sentence. Lam's installed base generates a high-margin service revenue stream (60%+ gross margins) that is less cyclical. Even if new equipment sales slow, the service annuity provides a floor. The company also has a fortress balance sheet and a strong buyback program. But Third Point's exit suggests the fund sees the risk/reward skewing negative over the next 12-18 months. Optimization isn't just about making things faster. It's about respecting the user's limits. In this case, the user is the market, and its limit for speculative equipment premiums has been reached.
What does this mean for the broader tech narrative? The same AI exuberance that lifted Lam to 35x PE is inflating the entire crypto-and-AI infrastructure stack. Projects that pitch themselves as "the GPU for AI" or "the decentralized compute layer" are riding the same wave. When the equipment cycle turns, those narratives will be stress-tested too. If you can't explain the mechanism, you're not ready for mainnet reality. Third Point's filing is a reminder that all cycles have a peak. The question is whether you're positioned for the descent.
Takeaway: Lam Research is a world-class company with a durable moat. But its stock price has already priced in the next two years of AI growth. Third Point is selling because the risk of a 2026 capex slowdown is underpriced by the market. Watch the next round of cloud capex guidance. If the marginal growth rate decelerates, equipment stocks will correct. The signal is already in the filing.