Third Point LLC just filed an SEC disclosure. They sold their stake in Lam Research. The timing is everything. Lam is the dominant supplier of etch and deposition equipment for the most advanced semiconductor nodes. This is not a random portfolio trim. This is a macro signal.
Context: The Equipment Bellwether
Lam Research sits at the intersection of two megatrends: AI infrastructure and global chip fabrication. Their equipment is critical for HBM (high-bandwidth memory) TSV etching, 3D NAND stacking, and GAA transistor deposition. When a sophisticated macro fund like Third Point reduces exposure, it implies a shift in the expected capital expenditure cycle. The semiconductor equipment sector is the 'pick-and-shovel' of the AI era. But picks and shovels are cyclical. The question is: are we at the peak of that cycle?
Core: Decoding the Sell-Off
The analysis of the SEC filing reveals three layers. First, valuation. Lam’s trailing PE has been running 30-35x, well above its historical average of 25x. The AI premium is baked in. Second, export controls. The United States’ restrictions on advanced equipment to China have structurally reduced Lam’s addressable market. Its China revenue share dropped from 29% to below 20% in two years. Third, the AI capex marginal growth rate is slowing. Cloud providers have spent over $200 billion on AI infrastructure in 2024, but the incremental dollar is now being scrutinized. If returns on AI compute disappoint, the next wave of orders will be deferred.
From my macro framework, I see this as a liquidity-first signal. The Federal Reserve’s balance sheet is still contracting in real terms, and global M2 growth is tepid. Equipment stocks are long-duration assets. When liquidity tightens, the discount rate on future cash flows rises. Lam’s stock price already reflects this. Third Point is simply front-running the compression.
Contrarian: The Decoupling Thesis
The conventional take is that Third Point is bearish on AI. I disagree. They are rotating from the manufacturing layer to the compute layer. The move from 'sell picks and shovels' to 'buy the gold miners' is a classic sector rotation. Lam’s equipment is essential, but the bottleneck is shifting from hardware fabrication to software and algorithms. In crypto terms, this mirrors the transition from L1 infrastructure to L2 application scaling. The true AI-crypto convergence will be driven by decentralized compute markets, not by the number of fab tools. The lab experiment of AI agents using blockchain for data availability (Filecoin, Arweave) is still in its infancy. The institutional capital is already positioning for that next phase.
Takeaway: Positioning for the Cycle
Watch the quarterly capex guidance from TSMC, Samsung, and SK Hynix. If they guide lower, Lam’s order book will turn. But for the crypto-native analyst, the real signal is about the timing of the next AI liquidity wave. The Federal Reserve will eventually pivot. When that happens, equipment stocks will re-rate. But until then, the chop is for positioning. The sell-off is not a death knell for AI. It’s a tactical retreat. From the lab experiment to the global standard, the path is never linear. Yields attract capital, but security retains it. The macro shift is clear. The micro panic is overdone.
