Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
Every cycle hands contrarians the same temptation: wait for the miners to break, then back up the truck. The logic is seductive because it has worked before. When the people who literally manufacture the asset can no longer afford to keep the machines running, the marginal seller is exhausted, the float tightens, and price tends to find a floor. The problem with treating that as a mechanical buy signal in 2026 is that the conditions which made it reliable have quietly eroded, and the data this month says the squeeze is genuine but unfinished.
This past weekend the network executed one of the steepest downward recalibrations of its mining difficulty in seventeen years, a roughly ten percent cut that reset the target from the high-130-trillion range back toward the mid-120s. It is the kind of adjustment that only happens when meaningful hashrate has already walked off the network, and average block times had indeed stretched well past the ten-minute target in the run-up. The protocol does what it always does: when miners unplug, the math gets easier for whoever remains. What makes this episode worth a closer look is not the cut itself but the company it keeps. This is now the third significant difficulty reduction of the year, following an 11 percent drop in February and a further decline in March. As reporting on the adjustment notes, only a handful of calendar years in Bitcoin’s entire history have ever logged three or more top-tier downward moves, and the year is only halfway done.
When the marginal cost catch is no longer marginal
The cleanest read on miner stress is to compare spot price against what it actually costs to produce a coin. On that score the picture is stark. Capriole Investments founder Charles Edwards pegs the all-in production cost, including depreciation and amortization, at roughly $62,650, with the bare electrical floor sitting near $50,000. Bitcoin spent this week trading directly on top of that production figure. In his framing, shared via his analysis of the production-cost model, the most durable long-term value windows have historically opened in the band between the production cost and the electrical cost. By that logic the asset has merely arrived at the top edge of the zone, not the bottom of it.
Here is where the contrarian needs to be honest about history rather than nostalgic about it. Production cost has acted as a gravitational pull, not a hard tripwire. The same Capriole data shows Bitcoin traded below estimated production cost for more than six months in both 2019 and 2023. “At cost” is not a bottom. It is the price level at which the least efficient operators begin bleeding, which is a different and much slower process than a flush.
The efficiency divide is hiding the damage
What complicates the classic miner-capitulation thesis is that hashrate has stayed deceptively firm even as marginal operators suffer. The reason is a widening gulf in hardware economics. Secondary-market prices for rigs have collapsed over the past year, which sounds like distress but functions as a subsidy for the strong: well-capitalized firms are buying next-generation machines cheaply and retiring legacy fleets. A current-generation unit can deliver more than double the hashrate of a three-year-old machine at similar power draw, cutting energy per terahash by well over half. The aggregate network looks healthy because the efficient are absorbing the share vacated by the inefficient, even as the inefficient quietly switch off.
This is the structural wrinkle that breaks the old playbook. In 2018 and 2022, capitulation was visible precisely because there was nowhere for stranded capacity to go. Today there is somewhere to go, and it is not Bitcoin at all. With AI compute demand pressing against the limits of available power, several listed mining firms are converting infrastructure toward high-performance computing for artificial intelligence. As coverage of the margin collapse points out, Bernstein analysts have framed access to electricity, not chips, as the binding constraint on AI data-center growth. Miners sitting on power contracts are an obvious arbitrage. That escape valve dampens the forced-selling reflex that once made miner stress such a clean bottoming signal.
The on-chain gauges say “stressed,” not “broken”
The metrics that historically marked cycle bottoms are all leaning in the right direction without actually arriving. The Puell Multiple, which measures daily miner revenue against its annual average, has slid below one, signaling income running under trend. But its 30-day reading near the recent low is nowhere near the 0.45 it touched at the 2022 bottom or the 0.33 of late 2018. The price-to-miner-revenue multiple has cooled from peaks around 160 to roughly 80, yet the 2022 capitulation low sat at 33. A separate gauge tracking price drawdown since the last difficulty bottom has crossed into a stress zone at around 21 percent, but the 2022 reading reached nearly 39 percent.
Every one of these is the same story told three ways: pressure is building, and none of it has hit the extremes that historically marked exhaustion. As the on-chain breakdown of the difficulty drop summarizes, the market premium over miner revenue has narrowed but not vanished, and a true capitulation signal would likely demand either a deeper move or a longer stretch of depressed income.
The cycle that contrarians grew up with may be gone
The most important point for anyone positioning around miner stress is the one easiest to miss. Edwards himself has argued that the miner-driven bust of prior cycles is structurally dead, because post-halving issuance has fallen to a rounding error against aggregate demand. When public companies and ETFs are absorbing several multiples of daily mined supply, the miner is simply no longer the marginal price-setter he was in 2018. Forced miner selling can still cause volatility, but it can no longer single-handedly carve out the 80-to-90 percent drawdowns that trained a generation of buyers to wait for the capitulation low.
That reframes the entire trade. The contrarian instinct to wait for miners to break and then buy assumes miners still set the floor. If liquidity and institutional flow now set it instead, then “wait for capitulation” risks waiting for an event that the new market structure may never deliver in its old, unambiguous form. The difficulty cut will help the survivors by handing each remaining unit of hashrate a better shot at block rewards. But the relief arrives while price sits at cost, fees languish at multi-year lows, and the subsidy keeps shrinking. The honest position is neither “miners are breaking, back up the truck” nor “nothing to see here.” It is that the squeeze is real, the bottom signals are not yet flashing, and the framework that once told you exactly when to buy is no longer the framework the market is running on.
The next few weeks decide which way the strain resolves. A reclaim above the production-cost zone, a rebound in fees, or a stabilization in miner revenue would argue the worst is absorbed. Another leg down with hashprice still depressed would switch more machines off and put miner reserves under fresh scrutiny. Watch the gauges, but do not assume they will ring the bell as loudly as they used to.
This article is for informational purposes and does not constitute investment advice.
