The SEC filing landed without fanfare. Third Point LLC, Dan Loeb's hedge fund, quietly offloaded its stake in Lam Research. The market barely blinked. But for those who track liquidity flows across asset classes, this is not a routine portfolio tweak. This is a signal. A macro signal that the semiconductor equipment cycle—the very engine of AI infrastructure—is peaking.
Liquidity doesn't care about your long-term thesis when the cycle turns.
Lam Research is not just any equipment maker. It's the dominant player in etching and deposition, the two processes that define advanced chip manufacturing. From 3D NAND with 200+ layers to HBM's TSV vias, Lam's machines are the picks and shovels of the AI gold rush. The company's revenue hit $17.4 billion in fiscal 2023, and its stock doubled in the two years following the AI boom. Yet Third Point, a firm known for macro-driven bets, chose to sell.
Why now? The answer lies in the seven dimensions of semiconductor risk—technical, geopolitical, cyclical, and financial. Let's break them down.
Context: The Lam Research Landscape
Lam sits at the intersection of two powerful forces: the insatiable demand for AI compute and the tightening grip of U.S. export controls on China. Its core products—high-aspect-ratio etching, atomic layer deposition, and electrofill for TSV—are critical for HBM, GAA transistors, and advanced packaging. Customers include TSMC, Samsung, SK Hynix, and Micron. The moat is deep: switching costs are high, and Lam's process know-how is embedded in the fabs. But the moat is not impenetrable.
Core: The Seven-Dimension Analysis
- Technical: Lam's etching technology is best-in-class for 3D NAND and HBM. No competitor matches its high-aspect-ratio capability. Yet the tech cycle is maturing. Next-generation HBM moves to hybrid bonding, which reduces the need for TSV etching. The unit value per wafer may decline.
- Supply Chain: China once accounted for 29% of Lam's revenue. After export controls, that dropped to 20-25% and is still falling. The U.S. Commerce Department's "presumption of denial" for advanced equipment licenses means Lam cannot serve China's AI chip fabs. The gap is not fully filled by U.S., Europe, and Korea expansions. The structural revenue headwind is real.
- Capital Expenditure Cycle: Global WFE spending is expected to hit $100 billion in 2025, but the growth rate is decelerating. Lam's orders lead capex by 12-18 months. If Third Point believes capex will plateau or dip in 2026-2027, selling now is rational. The stock's forward PE of 30-35x already prices in high growth. Any deceleration triggers multiple compression.
- Market Demand: AI demand is strong, but the marginal growth rate of cloud capex is slowing. Hyperscalers spent $200 billion in 2024, with 30% growth expected in 2025. But equipment orders reflect project starts, not just announcements. Once the HBM ramp matures, the equipment order pace will slow. The "AI capex supercycle" narrative is real, but it has a shorter half-life than most assume.
- Geopolitical: Export controls are permanent. The U.S. is unlikely to loosen them, and China's "de-Americanization" is accelerating. Large Fund III is pouring billions into domestic equipment. In the medium term, Lam will lose share in mature nodes. The only question is how fast.
- Competitive: Applied Materials and Tokyo Electron are encroaching. AMAT is investing heavily in hybrid bonding and advanced packaging. TEL has strong relationships with Korean fabs. Lam's market share in etching is stable, but the competitive intensity is rising.
- Valuation: At 30x trailing earnings, Lam is priced for perfection. The historical average is 25x. A 20% multiple compression is plausible even without an earnings miss. Third Point's exit may be a simple valuation arbitrage: sell when the PE is stretched, buy when it contracts.
Contrarian Angle: The Decoupling Fallacy
The conventional wisdom says Third Point is wrong. "AI capex is just getting started." "Lam's technology is irreplaceable." "Export controls are temporary." These are comforting narratives, but they ignore the liquidity cycle. The semiconductor equipment sector is cyclical, not secular. The AI boom lifted all boats, but the tide is turning. The real contrarian view is that Third Point is not selling because of Lam's fundamentals—it's selling because the macro environment is shifting. Rising interest rates, sticky inflation, and geopolitical fragmentation are squeezing capital expenditure across industries. Even AI will face a reality check.
Another rug? No, just a liquidity trap. The equipment sector is a liquidity trap: it looks safe because of the AI narrative, but the underlying capital flows are already rotating. Third Point is simply reading the tape.
Takeaway: Positioning for the Next Cycle
For crypto-native investors, this is a powerful lesson. The same forces that drive Bitcoin and altcoin cycles—liquidity, macro expectations, regulatory shifts—also drive semiconductor equipment. The sell signals that the easy money in AI hardware has been made. The next rotation will favor software, services, and perhaps decentralized infrastructure that can offer cheaper compute. Watch for the decoupling: when the market realizes that AI equipment demand is not infinite, the correction will be swift.
Liquidity doesn't care about your long-term thesis when the cycle turns. Third Point knows that. Do you?