Hook
Daniel Loeb’s Third Point LLC has quietly shed a significant stake in Lam Research, according to a recent SEC filing. The timing is precise. Lam Research, the etching and deposition equipment giant, sits at the intersection of AI infrastructure, US-China tech decoupling, and global semiconductor capital expenditure cycles. The consensus will frame this as a simple profit-taking move. It is not. This is a signal on the macro horizon — a warning that the AI capital expenditure narrative is approaching an inflection point. For those who track liquidity flows, the message is clear: the equipment cycle is peaking, and the liquidity tide that lifted all AI-related assets is about to recede.
Context
Lam Research is not a household name, but it is the backbone of advanced chip manufacturing. Its etch and deposition tools are essential for building 3D NAND, HBM stacks, and leading-edge logic chips down to 2nm. The company commands roughly 20% of the global wafer fabrication equipment market. Its revenue is tied directly to the capital expenditure plans of foundries like TSMC, Samsung, and SK Hynix. Since 2023, Lam has been a beneficiary of the AI boom: HBM memory expansion, advanced packaging, and GAA transistor adoption drove its stock to all-time highs. However, Third Point’s exit suggests that the best of this cycle is already priced in.
Core
From my experience auditing smart contracts during the 2017 ICO boom, I learned that the highest returns come from identifying structural fragility before the crowd. The same principle applies to macro analysis. Third Point’s move is not about Lam’s technology — its etch-dominant position in high-aspect-ratio etching for 200+ layer NAND and TSV etching for HBM remains unassailable. The fragility lies in the demand side. Global WFE (wafer fab equipment) spending is expected to reach $100 billion in 2025, but the growth rate is decelerating. Cloud providers — the ultimate drivers of AI capex — are already signalling a shift from “build at any cost” to “optimise return on invested capital.” When the marginal dollar of AI capex stops accelerating, equipment orders will be the first to feel the pinch. Collateral is just debt wearing a mask of trust. The equity premium on Lam Research is backed by the trust that AI capex will grow indefinitely. That trust is fraying.
Furthermore, US export controls on advanced semiconductor equipment to China have structurally capped Lam’s addressable market. China once contributed 29% of Lam’s revenue; now it is below 20%. The company cannot recapture that growth elsewhere. The CHIPS Act subsidies in the US and Europe are slow to materialise, and the build-out of fabs in Arizona and Germany is plagued by labour shortages and cost overruns. The net effect is a permanent reduction in Lam’s growth ceiling. We do not ride the wave; we engineer the tide. The tide of AI-driven equipment demand is no longer rising — it is entering a plateau.
On valuation, Lam trades at 30-35x trailing earnings, well above its historical average of 25x. The market has priced in a “super-cycle” that may not materialise. A 20% multiple compression is consistent with a mid-cycle slowdown, and that is exactly what Third Point is hedging against.
Contrarian
The immediate reaction from crypto-native investors will be to ignore this as a “traditional finance” issue. That is a mistake. The decoupling thesis between crypto and macro is real, but only for assets that have genuine utility independent of centralised capital flows. AI-related tokens — Render (RNDR), Akash (AKT), and others that depend on the narrative of “decentralised compute for AI” — will be directly impacted. If hyperscalers slow their GPU procurement, the demand for decentralised compute-as-a-service will also soften. The irony is that the most vocal proponents of AI-crypto convergence are often the ones who fail to see that the underlying infrastructure (Lam’s equipment) is the canary in the coal mine. We do not ride the wave; we engineer the tide. The tide is turning, and the froth on AI tokens will recede first.
However, the contrarian opportunity lies in the opposite direction. The slowdown in centralised AI capex could accelerate the adoption of alternative compute models. When the cost of marginal GPU cycles rises due to supply constraints, decentralised compute networks that aggregate idle capacity become more viable. Projects like Akash, which offer spot-like pricing for GPU compute, could see increased demand from AI startups that cannot afford premium cloud rates. The key is to distinguish between narrative-driven tokens and those with genuine unit economics. My 2020 work on compound liquidity during the DeFi crisis taught me that bear markets separate the wheat from the chaff. The same is happening now in AI-crypto.
Takeaway
Third Point’s Lam Research exit is a macro signal that should not be ignored. It points to a peak in the AI capital expenditure cycle, a structural headwind from export controls, and a valuation correction in the semiconductor equipment space. For crypto investors, the implication is twofold: first, reduce exposure to AI-crypto narratives that are priced for perfection; second, begin accumulating positions in decentralised compute projects that offer genuine cost advantages over centralised cloud. The liquidity cycle is shifting, and those who engineer the tide will be ready when the next wave forms.
Collateral is just debt wearing a mask of trust. The trust in AI’s infinite growth is being tested. Trust is the most volatile asset.
