This round’s “bottom indicators” may have failed for the first time. The cause isn’t on-chain—it’s in the power grid.
One of the most commonly used valuation anchors for Bitcoin is “market cap / hash rate.” Hash rate represents the security cost: if the price falls below that security cost, it’s considered oversold. For this framework to hold, there’s an implicit premise—that miners do only one thing. Price drops → shutdowns → hash rate falls → the denominator shrinks → the indicator automatically mean-reverts.
This feedback loop has broken this year.
BTC is down from last October’s high by nearly half, hashprice (daily revenue per PH/s) has been pushed to a five-year low, but the network’s total hash rate has only retreated by a little over 20% from its peak. Q1 saw the first quarterly hash rate decline since 2020; in Q2 the drop widened to nearly 6%, while difficulty fell by about 19% from its peak. It sounds like capitulation—but against the backdrop of BTC trading at half its value, the resilience is abnormal.
The reason is that miners are no longer serving only one customer. With the same land, the same substation, and the same cooling system, the revenue per megawatt-hour generated by AI compute is far higher than what mining produces. So that facility’s mining machines get shut down, but **the site and power capacity haven’t exited**—some are even expanding. The denominator of “hash rate” is being propped up by AI demand, so “market cap / hash rate” is systematically skewed toward undervaluation—it’s now pricing the capital expenditures of two industries at once.
The second leg is loosening too. Treasury-like companies that issue premium stock to buy BTC already have roughly 40% of their shares trading below the net asset value of their BTC holdings. Once the premium disappears, this buying demand doesn’t just stop—it flips into selling pressure: issuing shares at a discount harms them, and to replenish liquidity they can only liquidate inventories.
So the current situation is this: outside of ETFs, the two major marginal buyers—miners and treasury companies—are contracting at the same time, while the indicator that tells you “you should buy the dip” has its denominator being supported by money from another industry.
The indicator itself isn’t broken. It’s measuring miners’ costs—only now that cost no longer corresponds solely to Bitcoin.
Once hash rate is tied to AI, will you still look at models like this? $BTC
One of the most commonly used valuation anchors for Bitcoin is “market cap / hash rate.” Hash rate represents the security cost: if the price falls below that security cost, it’s considered oversold. For this framework to hold, there’s an implicit premise—that miners do only one thing. Price drops → shutdowns → hash rate falls → the denominator shrinks → the indicator automatically mean-reverts.
This feedback loop has broken this year.
BTC is down from last October’s high by nearly half, hashprice (daily revenue per PH/s) has been pushed to a five-year low, but the network’s total hash rate has only retreated by a little over 20% from its peak. Q1 saw the first quarterly hash rate decline since 2020; in Q2 the drop widened to nearly 6%, while difficulty fell by about 19% from its peak. It sounds like capitulation—but against the backdrop of BTC trading at half its value, the resilience is abnormal.
The reason is that miners are no longer serving only one customer. With the same land, the same substation, and the same cooling system, the revenue per megawatt-hour generated by AI compute is far higher than what mining produces. So that facility’s mining machines get shut down, but **the site and power capacity haven’t exited**—some are even expanding. The denominator of “hash rate” is being propped up by AI demand, so “market cap / hash rate” is systematically skewed toward undervaluation—it’s now pricing the capital expenditures of two industries at once.
The second leg is loosening too. Treasury-like companies that issue premium stock to buy BTC already have roughly 40% of their shares trading below the net asset value of their BTC holdings. Once the premium disappears, this buying demand doesn’t just stop—it flips into selling pressure: issuing shares at a discount harms them, and to replenish liquidity they can only liquidate inventories.
So the current situation is this: outside of ETFs, the two major marginal buyers—miners and treasury companies—are contracting at the same time, while the indicator that tells you “you should buy the dip” has its denominator being supported by money from another industry.
The indicator itself isn’t broken. It’s measuring miners’ costs—only now that cost no longer corresponds solely to Bitcoin.
Once hash rate is tied to AI, will you still look at models like this? $BTC