Over the past 90 days, Core Scientific's stock has shown a mere 16% correlation with Bitcoin. That is lower than the correlation between Bitcoin and the stock of a struggling retail chain like DJT at 28%. For anyone who bought mining stocks to get crypto exposure, this is a silent alarm. The data is not noise—it is a structural signal.
Tom Lee's recent ranking of 17 crypto-related stocks, published by Fundstrat, was meant to help investors find the best equity proxies for Bitcoin and Ethereum. The methodology was straightforward: measure 90-day rolling correlations. The results, however, tell a different story. MicroStrategy sits at the top with 78% correlation to Bitcoin. Coinbase and BitMine lead for Ethereum at 74% and 80% respectively. But the real story is in the tail: Core Scientific at 16%, Riot Platforms at 31%, IREN at 33%. These are not crypto stocks anymore. They are something else.
The core insight is not in the correlation numbers themselves, but in what they reveal about business structure. Traditional logic says mining stocks should track Bitcoin because their revenue is denominated in BTC. That logic is breaking down. The reason is not a flaw in the data—it is a flaw in the narrative. Over the past two years, major mining firms have shifted from pure Bitcoin mining to AI compute leasing. The same power infrastructure, data centers, and cooling systems that once ran ASICs now serve NVIDIA GPUs for AI workloads. Core Scientific, TeraWulf, and IREN now report that AI revenue dominates their income statements. TeraWulf's CFO explicitly stated that future earnings will be driven by recurring contract revenue, not BTC price swings.
From my own audits of mining operations during the 2022 Terra collapse, I saw how quickly business models can pivot under financial pressure. That pivot was survival-driven. This pivot is strategic. Mining companies are reassembling their value propositions like money legos—stacking power contracts, data center real estate, and compute capacity into a new asset class. The BTC mining component is becoming a side business, a hedge against idle capacity. The core asset is now a data center landlord with a crypto origin story.
This structural shift creates a dangerous mismatch in investor expectations. If you buy Core Scientific believing you own a leveraged Bitcoin proxy, you are actually short AI demand and long power contracts. The correlation data is not a bug—it is a warning. The market has not yet repriced these stocks fully. Miners still trade at P/E ratios that blend mining margins with AI infrastructure multiples. That gap will close, and the direction depends on the next leg of the AI narrative.
The contrarian angle here is that the narrative of "miners as crypto proxies" is so deeply embedded that even sophisticated investors overlook the reclassification. Tom Lee's own ranking reveals this blind spot. He ranks BitMine at the top for Ethereum, but Tom Lee is the chairman of BitMine. The conflict of interest does not invalidate the data, but it demands a zero-trust verification of the source. Second, the 90-day window masks the trend. If you extend to 180 days, the deceleration is even more pronounced. The market is pricing in a future where miners derive less than 30% of revenue from Bitcoin. The question is whether that future is bullish or bearish for their stock prices.
The systemic risk here is not to the miners themselves, but to the portfolio construction of crypto investors. Many institutional allocators use mining stocks as a way to gain crypto exposure without holding the asset directly. They are now exposed to AI demand cycles, data center utilization rates, and power contract renegotiations—factors that have nothing to do with Bitcoin's halving cycle or transaction fees. This is a classic case of hidden correlation shifting. My mapping of 12 potential liquidation cascades during the 2020 DeFi composability crisis taught me that risks are often hidden in the dependencies you assume are stable. The same applies here.
What does this mean for the next 12 months? If AI demand continues to grow, mining stocks will decouple further from Bitcoin. They will trade more like infrastructure REITs or AI compute providers. If AI demand cools, they will lose both the AI premium and the crypto tailwind, creating a double whammy. The safest way to get Bitcoin exposure remains MicroStrategy or the spot ETFs. The miners are now a different asset class—one that requires tracking hyperscaler capex cycles, not just Bitcoin price charts.
The takeaway is not that mining stocks are bad investments. It is that they are no longer the asset you think they are. When the driver of your stock price shifts from Bitcoin's hash rate to AI's compute demand, the correlation numbers are just a rearview mirror. The road ahead is entirely different.
