The SEC filing landed without fanfare. Third Point LLC, Daniel Loeb’s hedge fund, reduced its stake in Lam Research. The market yawned. LRCX shares barely moved. But math doesn’t lie. This is not a portfolio trim. It is a deliberate signal about the semiconductor equipment cycle’s inflection point.
Lam Research is the undisputed king of etch and deposition. Its tools carve the billions of transistors in NAND, DRAM, and logic chips. For years, the company rode the AI wave: HBM, CoWoS, GAA. Revenue grew, margins held, and the stock traded at a premium. Yet Third Point, known for surgical exits, decided to take chips off the table.
Smart contracts execute. They don’t judge. But hedge fund managers do. They read the tea leaves of capital expenditure cycles, export controls, and competitive dynamics. The filing reveals a fund that has stress-tested the narrative of perpetual AI infrastructure buildout and found it wanting.
Let me break down what I see from the trenches. I’ve spent years auditing protocols, tracing on-chain liquidity, and analyzing the architecture of trust. Semiconductor equipment is different, but the same principles apply: verify claims with data, look for the hidden cost of centralization, and never trust a narrative that assumes linear growth.
Context: The Man Who Sells in a Boom
Third Point is not a passive holder. It is an event-driven, activist fund that rotates aggressively. In 2023, it piled into AI infrastructure plays. Now it is trimming. The filing does not specify the exact number of shares sold, but the market cap implies a significant position reduction.
Lam Research sits at the intersection of three megatrends: AI compute, memory scaling, and geopolitics. Its etching tools are essential for HBM’s TSV (through-silicon vias) and for 3D NAND’s 200+ layers. The company’s revenue in FY2024 is expected to recover to $16-17 billion after a 25% drop in FY2023. The stock trades at 30-35x trailing P/E, well above the 5-year average of 25x. The premium is the AI premium.
But a premium is a borrowed expectation. It must be repaid with future growth. If that growth slows, the valuation corrects. Third Point is betting that the growth narrative is about to hit a wall.
Core: The Seven Dimensions of a Cyclical Exit
1. Technical: The Machine Is Not the Problem
Lam’s technology is superior. Its high-aspect-ratio etching for 3D NAND is unmatched. Its TSV etch for HBM commands a dominant share. The company’s R&D pipeline (GAA, hybrid bonding, new deposition chemistries) is solid. The problem is not the technology. The problem is the order book.
Equipment orders lead fab capital expenditure by 12-18 months. If the global WFE market peaks in 2025 at $100 billion, then orders for Lam will decelerate in 2026. Third Point is not selling because Lam’s machines are ineffective. It is selling because the marginal buyer of those machines—the fab planner—will soon face a harder budget decision.
2. Supply Chain: The China Problem
Lam’s China revenue dropped from 29% of total in FY2021 to 20-25% in FY2023 and is still falling. U.S. export controls block the sale of advanced etch and deposition tools to Chinese fabs. The company can still sell service and spare parts, but those are lower margin and capped by installed base.
Meanwhile, Chinese equipment makers (AMEC, Naura, Yituo) are closing the gap in mature nodes. In three years, Lam’s market share in China for 28nm and above could drop by 20-30%. The rest of the world (US, EU, Japan, Korea) cannot absorb the slack. The net effect is a structural headwind on revenue growth.
3. Capex Cycle: The AI Deceleration
Cloud providers spent $200 billion on AI capital expenditure in 2024. That number is expected to grow 30% in 2025. But the marginal growth rate is slowing. The first $100 billion is easy. The next $100 billion requires proven returns. AI chip demand is still strong, but the equipment that makes those chips sees orders with a lag. If cloud capex growth decelerates from 30% to 15% in 2026, Lam’s order book will feel the pain first.
Third Point is pricing in that deceleration. The fund is selling the peak of the capex cycle, not the bottom.
4. Demand: The HBM Mirage
HBM equipment is the darling of the story. Lam provides TSV etch and deposition for HBM stacks. The market grew 50% in 2024 and is projected to grow again in 2025. But HBM equipment demand is lumpy. Once the major fabs (Samsung, SK Hynix, Micron) finish capacity expansion, the equipment intensity drops. The next technology node (1β, 1γ) may reduce the number of TSV steps per wafer. The long-term demand for HBM-specific etch tools is not linear.
Liquidity is an illusion until it’s not. The same applies to order pipelines. A slowdown in HBM equipment orders would hit Lam disproportionately because its HBM exposure is high.
5. Geopolitics: The Permanent Fog
Export controls are not a short-term shock. They are a permanent feature. The Biden administration’s October 2023 rules created a presumption of denial for advanced equipment to China. The Trump administration is likely to maintain or even tighten them. Lam’s China revenue will continue to compress. The company’s ability to offset with non-China growth is limited by the slower pace of fab construction in the US and Europe.
Community governance in the semiconductor industry—the unwritten rules of the state-backed giants—means that Lam is caught between a U.S. government that wants to win and a Chinese government that wants to self-suffice. The result is a structural loss of addressable market.
6. Competition: The Squeeze
Applied Materials is a direct competitor in deposition and etch. Tokyo Electron is strong in memory. KLA is complementary. The competitive intensity is increasing, especially in advanced packaging and HBM. Lam’s lead in TSV etch is not unassailable. AMAT’s new hybrid bonding platform could reduce the need for TSV steps, directly threatening Lam’s equipment revenue per wafer.
If the total addressable market stops growing, competition becomes a zero-sum game. Lam’s margins will compress.
7. Valuation: The Math
At 30-35x P/E, Lam is priced for perfection. Even if the company hits its FY2026 revenue target of $18 billion, the earnings growth rate will decelerate from 20% to 10%. A 10% growth stock at 30x P/E is expensive. Historical mean reversion suggests a 20-25x P/E multiple, implying a 30% downside from current levels.
Third Point’s math is simple: sell the premium, wait for the cycle to reset, buy back later. This is not a bet against Lam’s technology. It is a bet against the consensus growth rate.
Contrarian: The Misreading of the Signal
Most analysts will frame Third Point’s move as a bearish signal on Lam. They will point to export controls, China slowdown, and competitive pressure. But the contrarian view is that the sale is a tactical rotation, not a fundamental rejection.
Lam’s technology moat is intact. The company generates $3 billion in free cash flow annually. It has a strong balance sheet. The AI-driven demand for advanced memory and logic is real and multi-year. The problem is timing. The stock’s valuation is pricing in the next two years of growth as if it will be linear. It won’t be. There will be a 12-18 month period of order digestion, inventory normalization, and capex pause. That is the window Third Point is selling into.
The real risk is not that Lam’s technology becomes obsolete. The real risk is that the market misreads the cycle and overcorrects. If the stock drops 20% on a growth slowdown, it will become a buy. But for now, the sell is rational.
In my years auditing DeFi protocols, I’ve seen the same pattern: a narrative-driven rally that pushes valuations beyond sustainable cash flows. The smart money rotates out before the narrative breaks. Liquidity is an illusion until it’s not. The same applies to Lam’s order pipeline.
Takeaway: The Cycle Is the Story
Third Point’s sale of Lam Research is a canary in the coal mine for the semiconductor equipment cycle. It signals that the beginning of the end of the AI capex super-cycle is near. Not the end of AI, but the end of the hyperbolic growth phase. For Lam, the next 12 months will be a test of whether it can maintain margins and revenue growth in a slowing environment.
What does this mean for the blockchain and crypto industry? The same capital cycle dynamics apply. Mining hardware, GPU cluster financing, and even DeFi liquidity bears all follow a pattern: early adopters capture the highest returns, then latecomers compress margins. The equipment sector is the ultimate “pick and shovel” play. And when the shovel sellers start to be sold, the diggers should be careful.
Math doesn’t lie. The SEC filing tells a story of a fund that has stress-tested the narrative and found it wanting. The next quarter’s earnings will tell us if the market agrees.