Miner Capitulation describes financial pressure forcing some miners to cut operations, sell reserves or leave the market. It is an economic description, not a state recorded by Bitcoin’s protocol, the shared rules of the network. A fall in network computing power alone proves neither bankruptcy nor a future rise in bitcoin’s price.
Our imaginary farm might struggle with expensive electricity, lower revenue or excessive loan payments. From outside, we cannot see its accounts. Glassnode offers Miner Capitulation Risk, a model combining indicators of revenue stress and declining mining activity. It estimates elevated risk; it is not a list of companies that have stopped paying their bills. To establish trouble at a particular operator, we also need its financial information or a documented decision. A network chart cannot substitute for that evidence. [1]
Next to our indebted farm might be a healthy operation that switches off identical machines only during expensive hours. Braiins Price Adapt enables this kind of control using energy prices, equipment efficiency and Hashprice—the expected daily revenue per unit of computing power. An ASIC is a chip designed for a specific task; mining machines use such chips for the calculations involved in finding blocks. Switched-off machines can resume work when operating conditions improve. Deliberately reducing consumption, known as Curtailment, therefore need not mean insolvency, or inability to meet obligations. A quiet building alone does not reveal the reason. [2] [7]
Network Hash Rate, or hashrate, expresses the speed of computations used to find blocks. Coin Metrics estimates it from mining difficulty and the rate at which blocks arrive. It does not directly add up readings from every mining machine. Blocks also arrive at irregular intervals: even with unchanged computing power, they can appear more slowly for a while. A short dip in a chart therefore does not reveal how many machines disappeared. A longer measurement window smooths random fluctuations but reveals changes later. Both properties matter when comparing charts. [3]
If computing power really declines, finding blocks becomes slower on average. Difficulty Adjustment takes place every 2016 blocks, as described in the Bitcoin Developer Guide. It responds to the time taken during the preceding period, rather than immediately to a farm shutting down. Lower difficulty can increase remaining miners’ expected revenue from the same computing power, but other conditions can keep changing. Relief thus arrives with a delay and may not be enough. The protocol does not cancel loans, cut electricity prices or repair broken cooling equipment. [4]
Our farm could move bitcoin (BTC) to another wallet it owns, pay a partner or prepare a sale. All these situations can look like an outflow. Glassnode’s Miner Outflow measures amounts transferred from addresses identified as belonging to miners. The transfer alone proves neither an exchange sale nor financial distress. Metrics labelled entity-adjusted try to filter transfers within the same entity using estimated address ownership. Even this adjustment cannot reveal every transaction’s purpose. The metric’s definition, transfer destination and wider context all matter. [5]
Hash Ribbons tries to identify miners’ retreat and return from changes in computing power. Glassnode describes its Hash Ribbon metric using moving averages over 30 and 60 days. Each day, they are recalculated from the network hashrate estimates for the corresponding number of most recent days. The shorter average reacts faster; the longer one shows the slower trend. When the shorter crosses above the longer, the methodology interprets it as the worst phase of capitulation ending. This is an interpretation of smoothed data, not confirmation that every farm has recovered. Before reading another service’s chart, check that its time periods and signal rules are the same. [6]
In a historical chart, miner distress may coincide with a low bitcoin price. Glassnode uses this relationship in Miner Capitulation Risk. It does not follow that prices must rise after every signal. A signal can arrive late or fail, while hindsight makes successful examples tempting to cherry-pick. A fair evaluation uses only data available at the time and also tests periods outside those used to build the model. For beginners, the indicator is more useful as information about mining conditions than as an instruction to buy. [1]
Return to our two farms. The first may close because of loan payments, while the second resumes once electricity gets cheaper. Braiins shows why equipment efficiency and energy prices matter; the same fall in Hashprice still affects operators differently. Assessing a company also requires contracts, reserves, debts and the ability to restart. Even after an operation is sold, a new owner may restart the same machines under different conditions. One company closing, a temporary loss of computing power and BTC price movements are three different stories. [2]
For the clearest picture, read this entry together with Hashprice, Difficulty Adjustment, Curtailment, Hash Ribbons.