VanEck Head of Digital Asset Research: Current AI Infrastructure Rally Is Not a Bubble; Crypto Market Sluggishness Stems from Institutional Disappointment with L1s

Compiled & Edited by: Deep Tide TechFlow

Guest: Matthew Sigel, Head of Digital Assets Research at VanEck, Portfolio Manager of the VanEck Onchain Economy ETF (NODE)

Hosts: Rob (Robbie Klages) & Andy, The Rollup's "AI Super Cycle"

Podcast Source: The Rollup

Original Title: VanEck Research Head: Why This Isn't The AI Bubble Everyone Fears (Here's Why)

Broadcast Date: August 10, 2026

Disclosure: Matthew Sigel is the Head of Digital Assets Research at VanEck and Portfolio Manager of the NODE ETF. Opinions expressed in this episode may align with the holdings of VanEck's funds.

Key Takeaways

The VanEck Onchain Economy ETF (NODE), managed by Matthew Sigel, has outperformed Bitcoin by nearly 100 percentage points over the past 15 months, primarily driven by an early bet on Bitcoin miners transitioning into AI data centers. He believes that the market was highly concentrated in the first five months of this year: companies spending the most on capital expenditures saw the best stock performance. However, this logic reversed after June, with software assets (including Bitcoin and crypto tokens) being sold off together, and capital expenditure itself becoming a penalized factor.

Sigel's stance can be summarized in two key judgments: First, AI infrastructure is not a bubble. Unlike the 19th-century railroad bubble, this cycle's funding comes from long-term leases in the private sector, rather than government land grants and speculative bonds.

Second, the real pressure in the crypto market isn't macroeconomic, but institutional disappointment with major L1s. Since the election, VanEck has reduced its positions in major L1s like Solana and ETH, shifting focus to enterprise chains like Circle, Stripe, and Robinhood. He believes that if the CLARITY Act passes and establishes a disclosure regime, affected tokens could experience significant relief rallies, but until then, he will remain cautious.

Key Highlights

On Market Structure Shifts

• "In the first five months of this year, the companies spending the most saw the biggest gains; but since June, the biggest spenders have been penalized the most."
• "Bitcoin is software, and coincidentally, open-source software. When software stocks are under pressure overall, it's hard for Bitcoin and crypto tokens to decouple."
• "I hope the market sees more dispersion rather than a single factor dictating everything. That way, we have opportunities to unearth alpha from individual assets."

On the AI Transition of Bitcoin Miners

• "ASICs aren't the miners' most valuable assets; power and land are."
• "Early miners' business model was constantly issuing shares to buy machines, with revenue halving every four years—it was a melting ice cube. Now they can access the debt market without diluting shareholders."
• "CleanSpark would need Bitcoin to reach $360,000 before it makes sense to return to pure mining, because only then would it be worth tearing up the AI leases they've already signed."

On Comparing AI Infrastructure to the Railroad Bubble

• "The U.S. poured 3% of its GDP into railroads for 20 consecutive years, while AI has only just reached that level for a single year."
• "The railroad bubble was government-driven: Congress passed the Act in 1862, promising land grants first, but only transferring ownership after the network was built. Treasury-issued construction bonds were also subordinated to private capital."
• "An AI factory, once connected to the grid, with fiber optics in place and chips installed, can immediately start training models and running inference. Railroads, on the other hand, only became truly useful once the East Coast connected to the West Coast."
• "The four major cloud providers hold over $2 trillion in contracted backlog, with many contracts featuring customer prepayments, customer-supplied GPUs, and terms exceeding five years. In 1870, no one was buying train tickets for a line that wouldn't open until 1885."

On L1s and Enterprise Chains

• "After the election, many tokens doubled, but no truly breakout application emerged, and no application has managed to attract global capital."
• "The winners now are enterprise chains: Circle has one, Stripe has one, Robinhood has one, and even Wells Fargo is building its own customized chain."
• "Banks and other regulated institutions don't want to put real money on open-source chains. Even if they support them, they'd need to support three or five chains simultaneously, which dilutes any single L1's winner-take-all dynamic."
• "If the CLARITY Act passes and brings true disclosure requirements, some tokens could see massive relief rallies. We remain cautious until then."

Main Content

Chapter 1: The Market Shift from "Chasing Capex" to "Punishing Capex"

Rob: Let's start with the market conditions you're seeing recently. There was a period where positive news like DTCC or the CLARITY Act didn't elicit any positive reaction from digital assets. Now Saylor is selling, CLARITY seems unlikely to pass, and the market is neither rising nor falling—completely indifferent. Meanwhile, AI and US equities seem to be sucking up all the liquidity. Is that what you're seeing?

Matthew Sigel: Yes. The first five months of the year were very concentrated—companies spending the most saw the best stock performance. But starting June 1st, that relationship inverted. Excluding the recent bounce from the Situational Awareness liquidation bottom, that period saw the highest capex companies get hit the hardest.

Bitcoin and crypto tokens are actually being classified under "software." Many are watching the relative strength between semis and software, and Bitcoin is software—open-source software, at that. AI capabilities like Claude and Codex are significantly impacting many open-source software projects, but their upgrade cycles aren't top-down like Web2 companies where changes can be implemented on command. You can't force users to upgrade. This overall software underperformance weighs heavily on Bitcoin and crypto tokens.

Add to that the psychological weight of the four-year halving cycle, and the market holds this pattern in high regard. I'd like to see more dispersion and more sources of return diversification so we can find alpha in individual stocks and individual assets, rather than just betting on a single factor. I don't think it has to be either AI up, crypto down, or vice versa—reality will be more nuanced.

As for my own positioning, I'm still exposed to the AI infrastructure theme through Bitcoin miners. I'm optimistic about Bitcoin bottoming in Q4 and look forward to the market rebalancing. However, I prefer to act on positive catalysts accompanied by volume; choppy short-term trading isn't my style.

Chapter 2: Why NODE Made a Heavy Bet on Miners' AI Transition

Rob: So what's your current specific positioning? I see some tickers like MARA, Riot, APLD, WULF—essentially Bitcoin mining stocks transitioning into AI data centers. They've already built the infrastructure, have power and energy capacity, often in remote locations. These are precisely the raw materials AI data centers need, and they can even pivot existing hardware. How long do you think this trend can last? Also, if so much hashrate flows from the Bitcoin network to AI, could that create systemic issues for Bitcoin?

Matthew Sigel: The NODE ETF I manage has been running for 15 months and has outperformed Bitcoin by nearly 100 percentage points. The biggest source of return came from an early realization: every megawatt of power controlled by Bitcoin miners was severely undervalued relative to the valuation multiples of the few data center REITs at the time.

Early miners' business model was constantly diluting shareholders, issuing stock to buy ASICs, just to stay ahead of competitors. But revenue halves every four years—it was a melting ice cube, capital-intensive with razor-thin margins. Now, we're seeing financing costs for construction leases around these data centers drop significantly, not just for hyperscalers. These companies can access the debt market without continuing to issue stock and dilute shareholders. And with every new lease signed, the economics are almost better than the last. Combined with falling interest rates, this is creating significant value.

So the market style shift since June has hurt the most leveraged players the most, and we didn't use leverage and tried to underweight highly leveraged companies. When we saw funds like Situational Awareness, which heavily overlapped with our holdings, being force-liquidated by prime brokers, we chose to buy the dip against the trend. Last Thursday at the open, we made the largest single-day trade since the fund's inception: we exited nearly 10% of low-volatility, low-beta positions and doubled down on our highest-conviction miners.

We haven't seen fundamental deterioration in the capital returns on this hyperscaler investment. In fact, reading earnings calls from companies like Amazon, these returns are better than initially expected. Old GPUs that were leased at $2 an hour are now seeing customers request significant price increases to renew contracts upon expiry. So we look at fundamentals first, and fundamentals are still improving.

At the lows, some of our highest-conviction companies were priced only for the value of existing leases, with no value assigned to the terminal value of data centers, platform value, or potential future leases. Even assuming the 10-year yield rises another 100 basis points, that work is already done. So at the lows, we saw a significant margin of safety.

Regarding the Bitcoin network, I don't see systemic issues. Quite the opposite—after the hashrate drops, remaining miners earn more. The companies we added to, like Bit Deer and MARA, still have strong optionality: continue mining, or convert facilities to AI. At some future Bitcoin price, people might revisit whether to switch back to mining, but we're not calling for facilities already converted to AI to switch back. We estimate CleanSpark would need Bitcoin to reach $360,000 to make it worthwhile to tear up the AI leases they just signed. So this optionality itself is very valuable.

Chapter 3: This Isn't a Bubble; It's the Anti-Thesis of 19th-Century Railroads

Rob: You mentioned historical analogies. You wrote an article comparing AI infrastructure to 19th-century railroad construction. Can you quickly walk us through why this comparison holds?

Matthew Sigel: Many people use the railroad bubble to counter my argument, saying it was also an economy-changing capex cycle, but early capital was massively destroyed. My response is: you can compare the scale, and you can compare the funding structure.

On scale, the U.S. invested 3% of its GDP in railroads for nearly 20 consecutive years. On the AI side, from GPT's emergence to now is about five years, and this is just the first year we've hit the 3% GDP level. Given current AI company valuations, the market is definitely not pricing in a twenty-year buildout. Most analysts think it peaks by 2030. That's the first disconnect.

The second is the funding structure. The railroad bubble was actually government-led. In 1862, Congress passed the Pacific Railroad Acts, promising hundreds of millions of acres of federal land to railroad companies, but land ownership only transferred after the entire route was completed. The Treasury also issued construction bonds, which were subordinated to private capital. The biggest railroad companies went overseas selling bonds marketed as safe assets, but these bonds depended on land grants when the companies didn't even hold the land titles yet. That's the bubble.

Now compare that to AI factories. Railroads needed to be connected to the network to realize true utility—their value was limited until the East Coast connected to the West Coast. But an AI factory, once connected to the power grid, with fiber optics in place and chips installed, can immediately train models and run inference. There's no need to wait for a global network to sync up.

The final key difference is the contracted backlog. In 1870, no one was buying train tickets for routes that wouldn't open until 1885, nor was there a futures freight market—everything was built on land sale speculation. But today's data centers, with the four major cloud providers alone, have over $2 trillion in signed contract backlog, with Microsoft and Oracle accounting for roughly half combined. Many contracts include customer prepayments, customer-supplied GPUs, and terms exceeding five years. The compute output from these factories is backed by real purchase orders, and financing is based on these contracts, not government subsidies.

Rob: So the conclusion is this cycle is more sustainable.

Matthew Sigel: Yes, more sustainable. It's supported by long-term private-sector contracts and backed by multi-year backlog as collateral.

Chapter 4: Major L1s Are Losing to Enterprise Chains

Rob: Someone in the chat mentioned that in 2021 or 2023, you said Solana needed to pause its L1 development. What would Solana need to do to win you back? Also, we're seeing many L1s reducing inflation and decreasing incentives for validators. Ethereum just released a new proposal this morning, and so did Solana and Near. How important are these inflation adjustments to L1 market conditions?

Matthew Sigel: After the election, when many tokens doubled but real adoption didn't accelerate and no truly breakout application emerged, we reduced our company-wide L1 exposure. Since then, we've witnessed the rise of enterprise chains. Circle has one, Stripe has one, Robinhood has one—these are semi-permissioned chains that allow public companies to customize user experiences and extract some of the economic value.

I understand this goes against the spirit of open source and crypto fundamentalism, but those who want to use these networks at scale need predictable fee streams. You asked about Solana, but I want to mention ETH first, because ETH is a prime example of excessive transaction fee volatility, and many large institutional players want more stable costs. So enterprise chains are capturing significant market share, and that's a key focus for our investments.

Solana's fees aren't as volatile, but until recently, you couldn't even buy USDC on Solana in New York, and this morning we saw Wells Fargo researching its own tokenized deposit chain. Banks and regulated entities don't want to put real money on open-source chains. Even if they participate, they'd need to support three or five different L1s, which dilutes the winner-take-all dynamic of any single chain. So market share is eroding, and the winner-take-all characteristic is weakening, which keeps us significantly underweighted in these tokens.

If the CLARITY Act passes—though the probability has dropped to its lowest point this year—I believe some tokens would see massive relief rallies. It would establish a disclosure regime, letting us know who the true beneficial owners are. There'd be no more KOLs shilling tokens without disclosing their holdings. We'd also see how many tokens labs and foundations hold. It's precisely the lack of a disclosure regime that makes many institutional investors ignore this space. I'd like to see this change before I can seriously become bullish on many L1s again.