Bitcoin miners are quietly doing the math and defecting. The power, the cooling, the fiber, the substations. All the stuff that took a decade to build for hashing SHA-256 turns out to be exactly what a hyperscaler will pay a premium for. The premium is substantial. Which is why the question hanging over crypto right now is whether the AI compute boom is slowly hollowing out Bitcoin’s security model, or just rearranging the furniture while the network keeps producing blocks.
VanEck’s Matthew Sigel, appearing on The Wolf Of All Streets with Preston Pysh, gave the most useful framing yet. He thinks Bitcoin will be “materially higher in one year” even while conceding leveraged proxies may outrun spot in the short term. He offered something rare in crypto commentary, a falsifiable threshold. If Bitcoin fails to surpass its all-time high by Q1 2028, that would “force maybe a thesis rethink.”
The miner exodus is real, and the price action shows the strain
Bitcoin is trading at $63,373 as of Monday, down 27.57% year to date and 41.98% over the past year. Ethereum has been worse, off 40.07% YTD, so this is not solely a Bitcoin story. But Bitcoin has a specific problem the rest of crypto does not. Its miners have a better paying customer down the street.
The EIA’s Annual Energy Outlook 2026 projects data center server energy use could climb to 818 billion kilowatthours by 2050 in the High Electricity Demand case, more than 16 times the 2020 level. Grid trade groups told Congress in April that the average individual data center load doubled from 150 to 300 megawatts between 2023 and 2024, and that data centers could reach 9.1% of all U.S. electricity consumption by the end of the decade. A miner sitting on a 100 megawatt interconnect in West Texas has to weigh a volatile block-subsidy revenue stream against a signed 15-year contract from a company with a trillion-dollar market cap. You know how that ends.
Institutional flows echo the shift. Coverage in March flagged that institutional capital is shifting from Bitcoin to AI protocols, and a January piece described American elites moving away from self-mining toward cloud computing arrangements. That is what the abandonment looks like in practice. It looks like a slow re-labeling of the rack.
Sigel’s conditional bull case leans on sovereigns, not hashrate
Sigel’s catalysts have almost nothing to do with mining economics. He pointed to a US Bitcoin Strategic Reserve the White House has been “teasing since Vegas” without delivering, and to 22 countries now mining or holding Bitcoin at the sovereign level, growing by 1 to 2 countries per year. Pysh agreed. Sigel added that investors are comfortable “missing the first 10% or 20%” while waiting for a clear catalyst.
His closing line, “Just stay the course. Don’t get too excited either which way,” reads as a shrug, but it does real work. The thesis rests on a floor of sovereign and institutional demand deep enough to absorb whatever supply the AI defectors dump on the way out, whether or not miners stay loyal.
What to watch between now and Q1 2028
Prediction markets are less optimistic than Sigel about the next six months. Polymarket puts the odds of Bitcoin reaching $70,000 by December 31, 2026 at 67%, $100,000 at 10%, and $200,000 at 1.5%. The $100,000 contract has drawn $2.18 million in all-time volume, the deepest book in the complex, and the implied probability has slipped 2.5% over the past week. Bitcoin’s previous all-time high above $109,000 in July 2025 is the number Sigel is measuring against, and the market currently rates that as a stretch even on an 18-month horizon.
WTI crude at $68 per barrel, down from an April high of $114.58 has bought miners some breathing room on operating costs. That helps at the margin. It does not change the underlying trade of hashrate revenue versus AI lease revenue.
The existential-threat framing is overheated. Quantum computing panic already got debunked when Google’s Willow chip triggered a 5.3% single-day drop in BlackRock’s iShares Bitcoin Trust (NASDAQ:IBIT), followed by expert notes that breaking SHA-256 would require roughly 1 million high-quality qubits, possibly 1.9 billion, against Willow’s 105. AI competes for kilowatts, which is the actual pressure point. If Bitcoin cannot make a new high by early 2028 with sovereign adoption compounding and ETF plumbing fully built out, the problem will be that the buyers never showed up, whatever the miners did.
Contact [email protected] for any questions or corrections.
Top Gaining Stocks
Top Losing Stocks