Miners are frequently blamed for causing dips in the price of Bitcoin. These accusations are often unsubstantiated—worse still, they’re sometimes based on faulty metrics that conflate mining pool payouts with miner spending, ultimately misleading their users.
Accurately assessing the degree and impact of miner selling is crucial to understanding the market. In Following Flows: A Look at Miners’ On-Chain Payments, Coin Metrics unveiled a new methodology for estimating miner activity. This methodology factors in the way that pool wallets are structured, allowing users to distinguish between pool and miner activity.
In this piece, we’ll refine our miner activity estimates with data on the relationship between miners and exchanges, allowing us to determine when and where miners sell their coins. All of the metrics used in this article will be made available in our upcoming 4.9 release of Network Data Pro.
In line with the conventional wisdom, we find that miners tend to prefer Huobi and Binance to other exchanges. We also find that flows from mining addresses represent a small percentage of total exchange inflows, about 5.5% at time of writing, and are not a major source of market volatility.
Miner Marketplace Measurements
Coin Metrics’ miner-exchange flows are based on our existing estimates for miner and exchange activity. The existing inflow, outflow, and supply metrics provide useful context on market sentiment, exchange health, and network decentralization. Due to differences in their problem structures, the exchange flows and miner flows are built on two different clustering techniques, each of which has its own drawbacks.
Exchange flows are estimated using the common-input-ownership heuristic, which assumes that addresses that are inputs to the same transaction share an owner. This technique is precise, but requires at least one seed address for every exchange, limiting coverage to a predetermined universe of exchanges. The heuristic is also broken by CoinJoins and peeling chains.
Trading on a centralized exchange typically requires users to deposit coins onto the exchange, where they’re custodied by the exchange operator. Mining generally works similarly: miners share resources with one another to improve their chances of finding a block, coordinating through mining pools with centralized operators. Mining pool operators typically receive newly-mined coins to an address they control, and only later distribute them to miners through payout transactions.
Coin Metrics’ miner flows account for this architecture by basing clustering on an address’s distance in hops from the coinbase transaction. Addresses that have received a coinbase reward, or 0-hop addresses, are assumed to belong to mining pools. 1-hop addresses that have received payment from a 0-hop address are tagged as belonging to miners. This heuristic is less precise than the common-input-ownership heuristic, but roughly mirrors the structure of mining pool wallets and provides better coverage.
Neither type of clustering can detect the use of fresh addresses, and the results of each should be treated as rough estimates. Because exchange tagging is more precise, we resolve collisions in favor of that heuristic: addresses that are tagged as belonging to both a miner and an exchange are treated exclusively as an exchange.
By combining these two techniques, we can assess where miners are depositing their coins, which roughly corresponds to where they sell.
The flows from miners to exchanges are broadly similar to overall exchange inflows, but have several key differences. Binance and Huobi are by far the best-represented exchanges, with Huobi being significantly overweight in comparison to its share of overall inflows.
These results seem intuitive, since Huobi and Binance are the only two exchanges in the coverage universe that also operate mining pools. Both exchanges also have close relationships with miners, and have strong presences in Asia, where most miners are based.
Outflows from exchanges to miners are also dominated by Binance and Huobi. This suggests that miners are buying on these venues as well.
As with inflows, Huobi is significantly overrepresented in outflows compared to the exchanges that do not operate mining pools. Binance’s share of both distributions is roughly comparable.