While everyone is watching Bitcoin ETF flows, the real liquidity signal is coming from a sector most crypto natives ignore: AI infrastructure.
Three Wall Street analysts from BofA, JPMorgan, and Oppenheimer just named their top AI stock picks — Palantir, Amazon, and Lam Research. Their target prices imply 30-50% upside. But for a macro watcher, the numbers behind these picks tell a far more important story about where global capital is heading, and how crypto will be affected.
Over the past 14 years, I’ve built a framework that tracks liquidity flows across traditional and digital assets. The AI investment cycle is not just a tech story — it’s a macro liquidity event that will redraw the map of capital allocation for the next three years. And if you’re only watching crypto order books, you’re seeing the shadows, not the source.
Context: The Three Layers of AI Infrastructure
Let me deconstruct what these three stocks actually represent. They are not just companies; they are proxies for three distinct layers of the AI economy:
- Palantir (Application Layer): The enterprise AI deployment platform. Its U.S. commercial revenue grew 149% year-over-year, and it raised guidance to 134%. This is the demand signal — companies are actually spending real money on AI applications that deliver measurable ROI.
- Amazon/AWS (Cloud/Compute Layer): The cloud infrastructure that powers AI workloads. AWS grew 37% and reported $496 billion in backlog (remaining performance obligations) — nearly 2.5x the previous year. This is the capacity signal — enterprises are committing to multi-year cloud contracts for AI compute.
- Lam Research (Physical Layer): The semiconductor equipment maker that builds the factories for AI chips. Its NAND revenue doubled, and it raised its 2026 WFE (wafer fab equipment) spending forecast to ~$150 billion. This is the supply signal — chipmakers are building capacity to meet AI demand.
This three-layer stack forms a coherent macro narrative: AI demand (Palantir) → cloud compute (AWS) → semiconductor manufacturing (Lam). The capital flows are cascading down the chain, and each layer’s growth reinforces the next.
But here’s the critical insight for crypto investors: This entire capital allocation chain is a direct competitor to crypto’s liquidity narrative.
Core: The Liquidity Drain and the Counter-Cyclical Opportunity
Based on my experience auditing liquidity flows during the 2020 DeFi Summer and the 2022 bear market, I can tell you that the current AI infrastructure boom is siphoning capital that would otherwise flow into crypto. Here’s how:
1. Venture Capital is Switching Tracks
In 2024-2025, crypto VC funding dropped 60% from its peak, while AI infrastructure deals surged. The same institutional investors who once poured money into L1s and DeFi protocols are now buying Palantir stock and funding GPU clusters. The numbers from the analysts’ report confirm this: Palantir’s $255 target implies a P/S ratio of 80x, which is only possible if the market is pricing in years of hyper-growth. That optimism is drawing capital away from risky crypto assets into “safer” AI equities.
2. The Semiconductor Cycle is a Macro Bellwether
Lam Research’s $150 billion WFE forecast is a historic high. To put it in perspective, that’s roughly the total market cap of all Layer 1 blockchains combined. Every dollar spent on fab equipment is a dollar not spent on crypto mining rigs, GPU-backed tokens, or DePIN infrastructure. During the 2022 bear market, the semiconductor capital expenditure cycle turned down, and crypto mining stocks collapsed. Now, the cycle is turning up, but this time it’s driven by AI, not crypto. The implication: GPU supply will remain tight, and crypto mining will be the marginal buyer, not the primary driver.
3. Enterprise ROI Demand is a Crypto Headwind
Palantir’s success is built on “measurable ROI” — enterprises are demanding that AI deployments show clear P&L impact. This is the opposite of the crypto narrative, which often relies on speculative value and long-term optionality. As long as enterprises are laser-focused on ROI, they will allocate budgets to proven AI stacks like Palantir and AWS, not to experimental blockchain solutions. The 149% growth in Palantir’s commercial revenue is a direct reflection of this budget rotation.
Contrarian: The Decoupling Thesis — Crypto’s Real Asymmetric Bet
Here’s where the contrarian angle comes in. While the mainstream narrative is that AI is draining crypto’s liquidity, I see a different story: The AI infrastructure buildout is creating a massive physical asset base that will eventually need to be tokenized, and the crypto stack is the only system capable of efficiently managing that tokenization.
Let me explain.
The Crisis Capitalist Play
In 2022, when FTX collapsed and most funds were liquidating, I directed our fund to acquire distressed debt from Celsius at 10 cents on the dollar. That position returned 300%. The same logic applies here: when everyone is piling into AI equities at 80x P/S, the real asymmetric opportunity is in the underappreciated infrastructure that will support the AI economy’s next phase — namely, decentralized compute, data verification, and asset tokenization.
Three Blind Spots
- AWS’s self-designed AI chips (Trainium/Inferentia) are a direct threat to NVIDIA’s GPU monopoly. If Amazon’s chips become cost-effective for inference, they will reduce the demand for high-end GPUs, which in turn lowers the cost of decentralized compute networks like Akash or Render. This could be a catalyst for DePIN, not a headwind.
- Palantir’s high customer concentration is a risk that the market is ignoring. The analyst report notes that Palantir has only 653 U.S. commercial customers but an average revenue per customer of $3.5 million. This is a “land-and-expand” strategy that works well in a bull market but becomes fragile in a downturn. The same fragility applies to crypto protocols with concentrated whale holdings.
- Lam Research’s $150 billion WFE forecast is based on the assumption that China’s fab construction won’t be blocked by further export controls. If the U.S. tightens restrictions, those capital expenditures will be delayed, creating a ripple effect that hits GPU availability and crypto mining hardware. This geopolitical risk is a blind spot in the analysts’ bullish thesis.
The Decoupling Thesis
Crypto is not following the AI bubble. The correlation between Bitcoin and the Nasdaq has been weakening since mid-2025. Why? Because crypto is becoming a monetary asset, not a tech stock. The AI infrastructure boom is a real economic event, but it’s also creating inflation in the real economy — more data centers, more energy consumption, more debt issuance. That inflation is exactly what Bitcoin is designed to hedge against. So while AI stocks are pricing in a perfect disinflationary growth story, Bitcoin is positioning for a world where central banks have to print more money to finance the AI buildout.
My institutional bridge-building experience in Zurich showed me that traditional finance is starting to see this. The Swiss private bank I partnered with after the 2024 ETF approval was not buying Bitcoin because of AI; they were buying it because they saw the fiscal consequences of the AI investment cycle. The $2.1 billion in ETF inflows I tracked in early 2025 were not just from tech investors — they were from macro funds hedging against the very infrastructure buildout that the AI stocks represent.
Takeaway: Positioning for the Liquidity Reallocation
The single most important question for crypto investors right now is not “Which L1 will win?” or “When will the bull market return?” It’s this: Are you positioned for the liquidity reallocation that the AI infrastructure boom is creating?
Here’s my framework:
- Short-term (6-12 months): The AI capital expenditure wave will continue to suck liquidity out of crypto, keeping Bitcoin in a range-bound bear market. Altcoins will suffer most. Survival is more important than gains. Focus on protocols with strong balance sheets and real revenue — the same way I analyzed DeFi protocols in 2020.
- Medium-term (12-24 months): As the AI buildout matures, the need for tokenized physical assets, decentralized compute, and verifiable data will create a new wave of crypto demand. Palantir’s own data integration challenges — which I’ve seen firsthand in my work with enterprise clients — are solvable only by blockchain-based off-chain verification. This is the contrarian thesis that no one is talking about.
- Long-term (24-36 months): The macro liquidity cycle will turn. Central banks will ease to offset the inflationary impact of AI capex, and crypto will be the primary beneficiary. The same investors who bought AWS at 3x revenue will rotate into Bitcoin at 10x book value.
Watch the order book, not the headline. The AI boom is not a threat to crypto; it’s a macro forcing function that will separate the survivors from the speculators. The market is always right, but narratives are often wrong. The real signal is in the $150 billion of fab equipment, the $496 billion of AWS backlog, and the 149% growth in enterprise AI spending. Those numbers tell me that the liquidity is flowing into the physical world first — and then it will cascade into the digital asset space.
I’ve been through four crypto cycles, and I’ve learned that the best entries come when everyone is looking the other way. Right now, everyone is looking at AI. That’s exactly when you should start building your crypto position for the next cycle.