Linas Beliūnas Breaks Down Coatue’s $12 Trillion AI Spending Framework: Sellers vs. Buyers of Sc…

In a recent LinkedIn post, Linas Beliūnas dives into a new report from Coatue that offers a comprehensive framework for understanding the burgeoning AI market and its projected $12 trillion in capital expenditures. Beliūnas highlights that Coatue’s analysis divides the market into two distinct categories: “Sellers of scarcity” and “Buyers of scarcity.” This distinction, according to Beliūnas, is crucial for comprehending the true dynamics beyond the surface-level “AI bubble” debate.

Beliūnas emphasizes Coatue’s core thesis, as he presents it, by stating:

“If you own the choke point, AI is an accelerant 📈 If you rent the choke point, AI is a margin tax 📉”

This powerful dichotomy, shared by Beliūnas, suggests that companies controlling essential AI inputs are poised for significant growth, while those dependent on these inputs face margin pressures.

The Sellers of Scarcity: AI’s Essential Infrastructure

According to Linas Beliūnas, the “Sellers of scarcity” are the companies that possess and provide the fundamental resources that AI development and deployment cannot function without. He lists these critical components, as identified in the Coatue report and shared on LinkedIn, as:

• Compute
• Memory
• Power
• Data centers
• Frontier models
• Distribution

Beliūnas points out that these are the bottlenecks AI is currently unable to circumvent, making them strategically vital assets in the current technological landscape. He notes that the immense capital being poured into AI infrastructure underscores this point.

Quantifying the AI Investment Wave

To illustrate the scale of investment, Beliūnas relays figures from the Coatue report, highlighting the substantial financial commitments already underway and projected. He writes:

“OpenAI + Anthropic are now at ~$55B in combined annualized revenue run-rate. Hyperscalers are on track to spend $700B+ in capex this year alone. Coatue estimates $12T in AI capex between 2026 and 2031.”

These numbers, as presented by Beliūnas, demonstrate the massive economic forces at play and the critical role of infrastructure providers in enabling the AI revolution. The significant increase in operating margins for companies like Micron, which Beliūnas cites, further supports the idea that owning these scarce resources is highly lucrative.

The Buyers of Scarcity: Navigating Margin Pressures

Conversely, Beliūnas explains that the “Buyers of scarcity” encompass the vast majority of companies operating within the AI ecosystem. These entities are consumers of the compute, data, and models provided by the sellers. As Linas Beliūnas articulates:

“The buyers are everyone paying for those inputs and hoping the unit economics eventually work.”

This group, he suggests, faces the challenge of managing costs associated with AI inputs while striving to achieve profitable applications and services. The “margin tax” concept, as shared by Beliūnas, reflects the potential for these input costs to erode profitability if not carefully managed or if the end products do not command a sufficient premium.

Broader Market Implications and Founder Guidance

Beliūnas also touches upon the wider market implications discussed in the Coatue framework, including the performance of major tech stocks like the “Mag 7” and the underlying reasons for market movements, such as the Nasdaq’s resilience. He mentions that the Coatue playbook delves into the “agentic AI architecture shift” and explains why memory, storage, power, and data centers have become “suddenly strategic assets.”

Furthermore, Beliūnas indicates that the breakdown he shared offers guidance for founders on where to build next and for investors on areas requiring caution. He posits that the “AI bubble” narrative is an oversimplification of a more complex market structure defined by scarcity.

Finally, Beliūnas draws a parallel between Coatue’s thesis and the investment strategy of Leopold Aschenbrenner’s hedge fund, Situational Awareness, noting a significant shift in their semiconductor positions. This connection, as highlighted by Beliūnas, suggests a broader industry consensus around the importance of AI infrastructure and the associated scarcity dynamics.