The Miners Are Bleeding: Bitcoin's 7.7% Difficulty Drop Explained

On March 20th, Bitcoin experienced a large decrease in mining difficulty by 7.76%, falling from 145.68 trillion to 133.79 trillion for block 941,472. This is the second-largest drop since December and the most significant decline since February. Since the start of this year, Bitcoin's mining difficulty has had a steady decrease from approximately 147.99 trillion down to 133.79 trillion. The amount that miners need to process Bitcoin transactions has also decreased significantly over the past month due to the loss of many miners. Three key numbers illustrate this situation regarding the future of both Bitcoin and the mining community: $88,000 (the amount of energy an individual needs to mine 1 Bitcoin currently), $69,200 (the current market value of 1 Bitcoin), and the average time it took to mine a single block within the previous 10 minutes (it took 26% longer than planned). All three figures demonstrate that it currently costs $19,000 more to mine one Bitcoin than it is sold for on the open markets.

Why the Difficulty Fell and What the Protocol Did About It

Every 2016 block or approximately every two weeks, the Bitcoin Blockchain alters its mining difficulty based on how much time it took to create the blocks within that last epoch. If blocks come in quicker than the target time of 10 minutes per block, the mining difficulty increases. Conversely, if blocks come in slower than this target, the mining difficulty decreases. The March 20, 2026 adjustment was triggered because of CloverPool's reported average block time of 12 minutes and 36 seconds over the preceding epoch. There were enough miners going offline that they were able to reduce production significantly. The protocol worked as expected and reduced the block difficulty, thus restoring the 10-minute average interval between blocks.

When mining difficulty was adjusted, hashrates dropped to 915-943 Exahash per second which is well below the 1 Zettahash per second that the Network has maintained through 2025. The drop below 1 ZH/s — a level that Bitcoin first surpassed in September 2025 — is a permanent government retreat, rather than a temporary disruption.

The decrease in mining difficulty during February can be attributed to Winter Storm Fern causing many of the mining facilities in the U.S. to go offline. The mining difficulty dropped 11.16% but miners returned to full production very quickly as electrical power grids were restored to normal operation, resulting in a 15% increase in hashrate. The same does not apply to the March 2026 difficulty adjustment since there was no weather-related event causing miners to go offline; therefore, this adjustment can be attributed to continued economic capitulation.

The Cost Problem Is Getting Worse, Not Better

Checkonchain's model calculates the cost of producing 1 Bitcoin to be approximately $88,000, given today's difficulty of producing a Bitcoin and worldwide energy consumption. Bitcoin's market price hovers somewhere around $69,200; this produces a gap of approximately $18,800 per coin between the cost to produce and what Bitcoin is trading for. This gap is significant enough that many miners are forced to turn machines off, liquidate all their reserves, and rethink their entire business model.

In addition, the price of oil hovering above $100 per barrel is a catalyst for change. Energy prices are tied very closely to crude oil price movements in virtually every fossil-fuel-reliant mining community. Every extra week that oil trades at elevated levels means yet another increase in electricity costs with regards to the gas/fossil-fuel energy used to power the facilities that mine Bitcoin. Disruption in the Strait of Hormuz is still present, and the Saudi pricing signals provide little relief from these elevated prices. Hashprice was approximately $33.30 per petahash/second/day as of March 22; this is near the break-even point for most mining equipment and dangerously close to all-time lows of $28, which occurred on February 23.

In February, Bitdeer made a withdrawal of 943 BTC from its reserves and stated that it still has no Bitcoin holdings as of March 21. In contrast to this, Core Scientific has now stated that it plans to liquidate most of its BTC reserves by 2026 in order to fund the expansion of AI infrastructure. Additional companies that have declared similar plans to diversify their assets in recent quarters include MARA Holdings, Hut 8, Cipher Mining, Riot Platforms, TeraWulf and CleanSpark. As a result, the largest miners (who are traded publicly) are reducing their total mined BTC holdings by reallocating data center capacity away from BTC and toward AI workloads. The structural supply of hashrate committed to BTC will not recover in that next adjustment cycle.

AI Is Now Bitcoin Mining's Direct Competitor

Crypto trader Ran Neuner has even gone as far as to say that AI has effectively done away with bitcoin mining altogether. While that is certainly an extreme way of framing it, the underlying dynamic is indeed occurring. The demand for AI computational resources has exploded as generative models have gained significant popularity and now exist in direct competition with bitcoin mining for key resources such as cheap power contracts, accessible data centre space, and energy-efficient hardware. Miners looking to pivot away from bitcoin and into AI that have access to stranded energy in remote locations within the US have an amazing opportunity. Microsoft and Google are paying out significantly more than normal for their specific infrastructures. In relation to the current economics of bitcoin, the calculations to pivot from bitcoin to AI is clear cut.

What the Adjustment Actually Means for Surviving Miners

For operators that stay online, the 7.76% difficulty drop is immediate relief. Each petahash of deployed hashpower earns proportionally more Bitcoin until the next adjustment. Revenue per unit of computing power improves. Margins that were underwater move slightly less underwater.

The next difficulty adjustment is estimated for April 3. Current projections point toward another modest downward move of approximately 0.7% — meaning the relief isn't a floor, it's a step in a staircase that's still descending.

Historical patterns offer some comfort. Difficulty drops of this magnitude during the 2021 China ban and the 2022 bear market both preceded eventual stabilization and recovery. The difference is that those events had clear resolution timelines. This one sits inside a sustained macro environment — war, elevated oil, hawkish Fed — with no clear end date.

The protocol keeps producing blocks. The question is who's still hashing them.