When Matthew Sigel, VanEck’s Head of Digital Assets Research, told the market that the AI infrastructure rally is “not a bubble” and that the crypto market’s coldness stems from institutional disappointment with Layer 1s, the reaction was a collective shrug. Yet the silence of that shrug is precisely where the alpha hides.
I’ve spent the last decade in the trenches of crypto research—auditing Zcash’s zero-knowledge proofs in 2017, mobilizing MakerDAO small-holders in 2020, counseling FTX victims in 2022, and reframing the Bitcoin ETF narrative in 2024. Through every cycle, the pattern is clear: when institutions speak, they speak in code. Sigel’s words are not a casual market update. They are a carefully calibrated signal of capital rotation.
Context: The VanEck Thesis and the Two Narratives
VanEck is not a random crypto influencer. As a $100 billion asset manager with SEC-registered Bitcoin and Ethereum ETFs, its research division carries weight among the institutional layer that actually moves markets. Sigel’s dual claim—that AI infrastructure is fundamentally demand-driven, and that institutional disinterest in L1s is the root cause of the current market “coldness”—is a framing that will influence asset allocation decisions for the next 6–12 months.
To understand why, we must separate the two narratives. The first is about real economic demand. The second is about broken promises.
Core: The Narrative Mechanism Behind the Rotation
Let’s start with the AI infrastructure claim. Sigel argues it’s not a bubble. Why? Because the underlying demand is real—Nvidia’s data center revenue, hyperscaler CapEx, and the energy buildout for AI workloads are all verifiable in traditional financial statements. This is not a speculative token ecosystem; it’s a physical infrastructure buildout that happens to have a crypto layer for coordination and settlement.
From my experience auditing the Zcash protocol in 2017, I learned that privacy is not just a feature; it’s a trust backbone. The same principle applies here: AI infrastructure tokens that can demonstrate actual usage—GPU hours rented, data stored, compute jobs executed—will sustain narrative momentum. The ones that rely on hype alone will collapse. The current market is doing a crude sorting mechanism, and VanEck is accelerating that process.
Now, the L1 disappointment. Sigel’s phrasing is precise: “institutional disappointment with L1s.” This is not a statement about Bitcoin or Ethereum. It’s about the long tail of Layer 1 projects that promised to be “the next Ethereum” but failed to deliver institutional-grade governance, regulatory clarity, or reliable performance.
During my 2020 MakerDAO governance mobilization, I saw how coordinated community action could prevent systemic risk. But today, many L1s lack even basic governance hygiene—anonymous founders, centralized sequencers, and opaque treasury management. Institutions don’t just want speed; they want auditability and accountability. The L1s that have disappointed are those that mistook technical capability for institutional readiness.
Sentiment Analysis: The Coldness Is a Rotation, Not a Recession
The market is not “cold” in a total liquidity sense. Bitcoin and Ethereum ETFs are still seeing net inflows. The coldness is a narrative vacuum—money that was previously allocated to L1 altcoins is now sitting in stablecoins or rotating into AI infrastructure plays. The vanity metrics of L1 TVL and transaction count have plateaued, while GPU network token prices have surged.
This is a textbook narrative displacement. The capital that was chasing L1 innovation is now chasing AI infrastructure innovation. But the key question is: Is the AI infrastructure narrative sustainable?
Contrarian: The Blind Spots in the Rotation
Here’s where the contrarian angle emerges. VanEck’s framing is seductive, but it has three blind spots that I, as a token fund manager, must flag.
First, institutional disappointment with L1s is not uniform. Ethereum’s layer-2 scaling, Solana’s reliability improvements, and Avalanche’s subnet architecture are all addressing specific institutional pain points. The disappointment is concentrated on L1s that failed to evolve. A blanket rotation away from all L1s could miss the ones that are actually becoming institutional-grade.
Second, AI infrastructure tokens face a “trust deficit” similar to what I saw in the 2022 FTX collapse. Many of these projects lack proper governance, have opaque tokenomics, and are run by anonymous or pseudonymous teams. In my 2026 work on AI-agent economic frameworks, I developed the “Human-in-the-Loop Consensus Framework” to ensure autonomous agents align with human ethical norms. This same principle applies to AI infrastructure tokens: without verifiable human oversight, they risk repeating the same trust failures that made institutions disappointed with L1s.
Third, the “not a bubble” narrative is a classic top tick signal. Just because the underlying demand is real does not mean token prices are correctly priced. During the 2024 Bitcoin ETF approval, I argued that ETFs were financial literacy infrastructure, not speculative instruments. But that didn’t prevent the price from correcting after the initial euphoria. The same will happen with AI infrastructure: a significant portion of the current price appreciation is speculative, and a correction is inevitable.
Takeaway: The Next Narrative Is Institutional-Grade L1s
The real alpha is not in blindly following the rotation from L1s to AI. It’s in identifying the L1s that will become the settlement layer for AI-driven economies. These L1s will need to offer regulatory clarity, auditable governance, and reliable performance. They will be the ones that pass the “trust and ethics” due diligence that I have applied to every investment thesis since the FTX collapse.
Read the docs. Question the whisper. The silence of the audit is where the next narrative begins.
Alpha hides in the silence of the audit. The market is not cold—it’s waiting for a signal. VanEck just gave one, but the real signal is not the destination. It’s the journey from L1 disappointment to AI infrastructure hope, and then back to a new generation of L1s that have learned from their failures.
Survival is the first strategy. But insight is the second.