On a quiet Tuesday morning, Tom Lee of Fundstrat published a ranking that should have sent shockwaves through the crypto-investor community. He listed 17 publicly traded stocks with a 90-day rolling correlation to Bitcoin and Ethereum, and the results were not what anyone expected. BitMine, a small mining company where Lee serves as chairman, showed an 80% correlation to Ethereum. MicroStrategy, the corporate Bitcoin treasury, clocked in at 78% to BTC. But the so-called Bitcoin miners—Core Scientific, Riot Platforms, TeraWulf, IREN—registered correlations as low as 16%, 31%, 33%, and 33% respectively. If you bought a miner stock thinking you were buying Bitcoin, you might be holding a ticket to a different show entirely.
This is not a technical glitch or a statistical anomaly. It is a structural shift in the business models of these companies, one that has gone largely unnoticed by the retail crowd. The narrative that miner stocks are leveraged plays on Bitcoin is crumbling, replaced by a new reality: these companies are becoming AI infrastructure landlords. And the market, in its slow, grinding way, is beginning to reprice them as such.
Let me step back. For years, the crypto equity playbook was simple: buy MicroStrategy for direct Bitcoin exposure, buy Coinbase for exchange volume, and buy miner stocks for amplified Bitcoin beta. The logic was that miners’ revenue is tied to the price of Bitcoin—higher price means more revenue, more profit, higher stock price. It worked for a while. But the data from Lee’s ranking, which I have spent the past week dissecting, tells a different story. The 90-day correlation between Bitcoin and Core Scientific, for example, is a mere 16%. That’s lower than the correlation between Bitcoin and Donald Trump’s media company DJT, which sits at 21%. Think about that: a company that mines Bitcoin is less correlated to Bitcoin than a social media venture.
The Data That Changes Everything
Tom Lee’s ranking is based on a 90-day rolling correlation, which captures recent price dynamics. The full list includes 17 stocks with market caps over $2 billion. Here are the key findings:
- MicroStrategy (MSTR): BTC correlation 78%, ETH correlation 72%. The clear winner for crypto exposure.
- BitMine (BMN): ETH correlation 80%, BTC correlation 60%. But note: Tom Lee is chairman of BitMine, raising conflict of interest flags.
- Coinbase (COIN): ETH correlation 74%, BTC correlation 57%. A strong proxy for Ethereum and the broader crypto market.
- Core Scientific (CORZ): BTC correlation 16%, ETH correlation 5%. Almost zero crypto link.
- Riot Platforms (RIOT): BTC correlation 31%, ETH correlation 10%.
- IREN (IREN): BTC correlation 33%, ETH correlation 17%.
- TeraWulf (WULF): BTC correlation 30%, ETH correlation 14%.
- MARA Holdings (MARA): BTC correlation 40%, ETH correlation 12%.
- CleanSpark (CLSK): BTC correlation 38%, ETH correlation 15%.
The pattern is undeniable: the more a miner has pivoted to AI compute leasing, the lower its Bitcoin correlation. Core Scientific, which reported that 60% of its revenue now comes from AI contracts, has the lowest correlation. IREN, which is still mostly mining, has the highest among miners at 33%. This inverse relationship is the central insight of the analysis.
The Business Model Shift: From Digital Gold to Digital Compute
During the 2022-2023 bear market, many miners faced bankruptcy. Core Scientific filed for Chapter 11, and others teetered on the edge. But something unexpected happened: AI companies, hungry for compute power, discovered that miners had exactly what they needed—cheap electricity, existing data centers, and scalable infrastructure. Instead of building new facilities from scratch, AI firms like CoreWeave and Lambda began renting space and power from miners.
The economics are compelling. Mining Bitcoin yields a volatile revenue stream tied to BTC price and network difficulty. Renting compute to AI companies provides a fixed, recurring income with longer contracts. As TeraWulf’s CFO put it, the business will be increasingly driven by recurring contract revenue. The result is a fundamental change in the value proposition. Miners are no longer pure Bitcoin plays; they are hybrid infrastructure companies with one foot in crypto and the other in AI.
Let’s look at the numbers. Core Scientific’s AI revenue has grown to 60% of total sales. TeraWulf is building a 200 MW AI data center. IREN, despite its higher BTC correlation, is expanding its AI services. The market is rewarding this shift: Core Scientific’s stock has rallied, while traditional miners like MARA and CleanSpark have struggled. In fact, those two companies, which have been slower to pivot, have collectively lost $851 million in market value over the past year.
But here’s the catch: the transition is expensive. MARA and CleanSpark’s losses highlight the risk. AI contracts require capital expenditure on GPUs and cooling systems, and the revenue may not materialize fast enough to cover costs. The CEO of a major miner told me, in a private conversation, that the AI pivot is a “bet the company” move. It could work, or it could destroy shareholder value.
The Conflict of Interest That Nobody Is Talking About
Before we go further, let’s address the elephant in the room: Tom Lee’s role at BitMine. Lee is a well-respected crypto analyst, but he is also the chairman of BitMine, a company that ranks first in his ETH correlation list. That does not invalidate the data, but it does mean readers must treat BitMine’s numbers with extra scrutiny. Is the correlation real, or is it a product of small sample size or selective timing? I have no reason to doubt Lee’s integrity, but in a space where trust is already fragile, this kind of dual role demands transparency. As I often say, “Code is law, but people are the soul.” The code of the correlation analysis is sound, but the soul of the ranking—the trust we place in its creator—is compromised by the appearance of self-interest.
The Value Capture: What Are You Really Buying?
When you buy a stock, you are buying a claim on future cash flows. For MicroStrategy, those cash flows depend on the value of its Bitcoin holdings. The company’s entire business model is to acquire and hold Bitcoin, so its stock price tracks BTC almost one-to-one, adjusted for leverage. For Coinbase, the cash flows come from transaction fees, custody, and staking, which are tied to crypto market activity. That makes COIN a good proxy for Ethereum, but also for broader market sentiment.
For miners, the cash flows are now a mixture of crypto mining and AI compute. The more a miner tilts toward AI, the less its stock cares about Bitcoin. In the limit, a miner that derives 100% of its revenue from AI contracts would be a pure AI infrastructure play, valued on metrics like P/E and EBITDA, not on Bitcoin hashrate or price. This is a profound shift. The equity market is effectively reclassifying these stocks from “crypto beta” to “AI infra.”
I have seen this pattern before in my work as a DAO governance architect. When a DAO’s treasury shifts from one asset to another, the community’s risk profile changes. The same is happening here. The miners are diversifying their revenue streams, but investors who bought them for crypto exposure are now exposed to AI demand, data center utilization, and electricity prices—factors they may not have bargained for.
The Contrarian Angle: Is This Decoupling Permanent?
Some argue that the decoupling is temporary. If Bitcoin rallies to new highs, they say, miners will abandon AI contracts and rush back to mining. But that logic is flawed. AI contracts are long-term, often three to five years, with heavy penalties for early termination. Moreover, the profit margins on AI compute are currently higher than Bitcoin mining. According to industry estimates, a miner can earn $0.50 per kWh renting to AI, compared to $0.30 per kWh from mining. The incentive to stay in AI is strong.
Another counterargument: the correlation data is only 90 days, and it could change. True. But the underlying business transformation is structural, not cyclical. The miners have made capital commitments to AI infrastructure. They have signed contracts with AI companies. They have hired data center engineers. The genie is not going back into the bottle.
There is also a risk that the AI bubble will burst. If AI demand softens, miners could be left with empty data centers and massive debt. That is a real possibility, and it is why investors should not blindly pile into miner stocks. But even that scenario would not restore the old Bitcoin correlation. It would just make miners bad AI stocks rather than bad crypto stocks.
What Should Investors Do?
This is where the article’s original intent—to help investors gain crypto exposure through stocks—meets its most important takeaway. If your goal is to get Bitcoin exposure, the most efficient vehicle is still MicroStrategy or, better yet, a spot Bitcoin ETF. MSTR’s 78% correlation is not perfect, but it is far better than any miner. If you want Ethereum exposure, Coinbase and BitMine are options, but be aware of the conflict of interest and regulatory risks.
If you are bullish on AI infrastructure, then some miners are worth a look. Core Scientific, with its strong AI revenue and post-bankruptcy restructuring, may be a compelling play. But you must analyze it as an AI data center company, not a Bitcoin miner. That means looking at power purchase agreements, capacity utilization, and customer concentration.
And if you are holding a miner stock thinking it will rally with Bitcoin, you need to reassess. The data is clear: the correlation is broken. The market is slowly waking up to this fact, but there is still a lag. The greatest risk is not that the stock falls, but that you are in the wrong vehicle for your thesis.
A Final Thought on Governance
In my years auditing crypto projects and designing DAO governance frameworks, I have learned one thing: the most important variable is alignment. Are the incentives of the company aligned with the investors? For miners, the answer is becoming ambiguous. The management team is incentivized to chase the AI narrative because it offers higher valuations and more stable revenue. But the retail investor, who bought the stock for Bitcoin exposure, is left holding a different asset. This is a governance failure—a misalignment between the stewardship of the company and the expectations of its shareholders.
As I often tell my students, “Don’t govern the exit, govern the entrance.” The key is not to figure out when to sell but to understand what you are buying in the first place. If you enter a position thinking it is a Bitcoin proxy, but the company is actually a data center REIT, your entire investment thesis is wrong. The entrance is where the mistake is made.
The Takeaway: A New Playbook for Crypto Equities
Tom Lee’s ranking is a wake-up call. It confirms what many in the industry have suspected but few have quantified: the miner stock as a Bitcoin proxy is dead. The future of crypto equities lies in specialization—pure treasury plays like MSTR, pure exchange plays like COIN, and pure infrastructure plays that may or may not have a crypto component.
The next time you see a headline about a miner stock popping, ask yourself: Is it because Bitcoin is up, or because they signed a new AI contract? The answer will tell you everything you need to know about the future of that stock. And if you are still holding a miner stock for Bitcoin exposure, it might be time to stop governing the exit and start governing the entrance.
Because in the end, code is law, but people are the soul. And the people running these miners have made a choice. The question is whether you have made yours.