Data Reading 2020 Cryptocurrency Mining Panorama: Bitcoin Total Output Exceeds 5 Billion USD, Ethereum Miners Earn Excess Profits

This article is from PANews, authored by Carol.

In 2020, the "mining circle" experienced several major events.

First, Bitcoin completed its third halving. With the price performance shifting from stability to explosive growth before and after the halving, Bitcoin mining turned crisis into opportunity, maintaining revenue levels similar to 2019 throughout the year. Additionally, over a dozen other tokens, including BCH, BSV, BTG, ETC, DASH, ZEC, and XZC (now renamed FIRO), also successfully completed halving in 2020.

Secondly, with the comprehensive explosion of the Ethereum ecosystem in 2020 and the continuous accumulation of ETH asset premiums, Ethereum mining also entered a period of rapid development, forming the two major markets of mining alongside Bitcoin mining.

Moreover, the well-known mining machine manufacturer Ebang International successfully listed on NASDAQ after Canaan Creative, as mining machine manufacturers continued to "test the waters" in traditional capital markets.

PAData will comprehensively review the mining industry in 2020 through data analysis of the industry's fundamentals, blockchain network fundamentals, and the revenue performance of major market participants, restoring the true face of the mining circle.

Data Review:

• The estimated annual energy consumption of BTC mining was about 77.78 TWh, equivalent to the energy consumption of Chile; the estimated annual energy consumption of ETH mining was about 14.64 TWh, equivalent to the energy consumption scale of Tunisia.

• The total output scale of BTC mining was approximately $5.012 billion, while the total output scale of ETH mining was about $2.690 billion.

• The mining difficulty of the Bitcoin network was adjusted a total of 28 times throughout the year, with 17 increases, the highest single increase being about 14.95%; Ethereum's mining difficulty increased by 52.20% for the year, a significant increase compared to the 5.32% increase in 2019.

• In 2020, the growth rate of the total network hash rate for both Bitcoin and Ethereum was lower than the growth rate of their respective prices, suggesting that miners may have earned excess profits.

• In 2020, F2Pool was the mining pool with the highest BTC hash rate share, averaging about 17.53% for the year, followed by Poolin, BTC.com, and AntPool, each exceeding 10%; the hash rate share of ETH mining pools was relatively stable, with a higher concentration. The combined average hash rate share of the three major pools, SparkPool, Ethermine, and F2Pool, reached 75.51%.

• Among BTC pools, there were 6 pools with estimated service fee income exceeding $10 million, with F2Pool and Poolin exceeding $20 million; in ETH pools, only SparkPool and Ethermine had estimated service fee income exceeding $10 million.

• The rapid development of custom machines in 2020 marked the gradual maturity of ETH mining and is a development direction for the future of ETH mining. Among custom machines, brands like Innosilicon, Panda, and WolfGod are relatively mainstream, with Innosilicon launching 5 custom mining machines, Panda and WolfGod each launching 3 custom machines, together accounting for 47.83% of the custom machine market.

The total scale of the two major mining markets is approximately $7.6 billion, with the proportion of service fees continuing to increase

The industry scale is an important component of the industry's fundamentals, and energy consumption is an indicator reflecting the actual scale of the industry. According to the estimated energy consumption scale from Digiconomist, the energy consumption of BTC mining remained stable throughout the year, with an annual growth of about 6.37%. The highest estimated energy consumption for the year was about 77.78 TWh, equivalent to that of Chile, while the lowest estimated energy consumption was about 56.58 TWh, with a brief decline in estimated energy consumption after Bitcoin's third halving, gradually recovering only by early June.

Benefiting from the explosive growth of DeFi, the scale of ETH mining saw significant growth in 2020. The estimated energy consumption of ETH mining increased from 8.11 TWh at the beginning of the year to 14.64 TWh by the end of the year, with an annual increase of about 80.52%, showing significant growth throughout the year. The highest estimated energy consumption for the year was about 14.64 TWh, roughly 1/5 of the highest estimated energy consumption of BTC mining, equivalent to the energy consumption scale of Tunisia.

The annual changes in energy consumption scale are generally consistent with the changes in output scale of the two major mining sectors. According to PAData's earlier panoramic review, the total output scale of BTC mining in 2020 was approximately $5.012 billion, a decrease of 3.78% compared to 2019, remaining relatively stable, while the total output scale of ETH mining was about $2.690 billion, significantly increasing compared to 2019.

Notably, in August and September 2020, as DeFi liquidity mining stimulated increasing user engagement and trading volume, the monthly output scale of ETH mining briefly surpassed that of BTC mining, especially in September, when the total monthly output of ETH mining was about $489 million, exceeding BTC mining's total output by $162 million, or 48.98%. The peak monthly output scale of ETH mining is a "snapshot" of the potential scale of future ETH mining.

The output composition of both BTC and ETH mining includes two main parts: block rewards and transaction fees. From the perspective of the proportion of these two parts, BTC mining is relatively more dependent on block rewards, while ETH mining has a higher proportion of transaction fees.

According to statistics, the total transaction fees for BTC mining in 2020 were approximately $326 million, an increase of about 108.97% compared to 2019, with a significant difference in the distribution of transaction fees throughout the year, marked by the third halving. After the third halving (including the day of the halving), the average daily transaction fee was about $1.2445 million, approximately 4.75 times higher than before the halving. As a result, the proportion of on-chain transaction fees in Bitcoin mining revenue increased from an average of 2.8% last year to an average of 6.69% this year, an increase of nearly 4 percentage points.

In contrast, the total transaction fees for ETH mining in 2020 were approximately $631 million, with a significant growth trend throughout the year, especially in the third quarter, when the booming DeFi liquidity mining caused the Ethereum network to become congested, leading to a rapid increase in transaction fees. The proportion of transaction fees in miner revenue also rose from 17.34% at the beginning of the quarter to 30.05% at the end of the quarter. Consequently, the average proportion of ETH mining transaction fee income in total revenue for the year reached 16.06%, nearly 10 percentage points higher than Bitcoin's proportion during the same period.

It can be expected that if the basic situation of the Bitcoin network in 2021 is similar to that of 2020, with the total network hash rate continuing to rise moderately while the number of transactions and block production remains stable, then transaction fee income will continue to be the main way to improve the marginal profits of mining, and the proportion of fees in revenue may continue to increase. On the other hand, whether the proportion of transaction fee income in Ethereum mining increases in 2021 will largely depend on the progress of Layer 2 scaling solutions. If mainstream DeFi applications migrate to Layer 2 networks, it may positively impact stabilizing or even reducing transaction fees.

Mining difficulty continues to rise, excess profits still exist

From the network status perspective, the mining difficulty of the Bitcoin network was adjusted a total of 28 times throughout the year, with only 9 decreases and 17 increases, the largest increase occurring on June 16, with a single increase of about 14.95%. Additionally, there were 4 instances where the difficulty was increased by more than 8%. The cumulative increase in difficulty for the year was about 43.79%, which is relatively moderate compared to the cumulative increase of 97.67% in 2019.

The adjustment of Ethereum's mining difficulty throughout the year was slightly higher than that of Bitcoin during the same period, with mining difficulty rising from 2456 T at the beginning of the year to 3728 T at the end of the year, an increase of about 52.20%, significantly higher than the 5.32% increase in 2019.

The changes in mining difficulty are closely related to changes in hash rate. In the mining cycle, the sequence of rising coin prices, increasing hash rates, and rising difficulty has formed a causal chain reaction. The moderate increase in BTC mining difficulty reflects the moderate increase in its total network hash rate. According to statistics, the total network hash rate of Bitcoin rose from 112.93 EH/s at the beginning of the year to 153.48 EH/s by the end of the year, with an annual increase of about 35.91%, a significant decline compared to the 143.59% increase in total network hash rate in 2019. Compared to Bitcoin's increase of 304.74% in 2020, the moderate increase in hash rate theoretically provides an opportunity for mining to earn excess profits.

On the other hand, the significant increase in ETH mining difficulty also reflects the substantial increase in the total network hash rate. According to statistics, the total network hash rate of Ethereum rose from 141.55 TH/s at the beginning of the year to 281.37 TH/s by the end of the year, with an annual increase of about 98.78%. Compared to the 10.54% decline in 2019, the total network hash rate of Ethereum saw a significant increase in 2020, indicating a substantial improvement in the mining fundamentals.

Stimulated by the favorable developments in Ethereum ecosystem applications, ETH rose by 468.64% throughout the year, far exceeding the increase in total network hash rate, which theoretically made ETH mining more likely to earn excess profits compared to BTC mining in 2020.

F2Pool and SparkPool dominate the two major mining sectors, with 8 pools estimated to earn over $10 million

Mining pools are major participants in the mining industry and are also one of the signs of industry maturity. Based on the average hash rate share throughout the year, F2Pool was the mining pool with the highest hash rate share in 2020, averaging about 17.53%, followed by Poolin and BTC.com, with average hash rate shares of about 14.81% and 12.30%, respectively. Additionally, AntPool's average hash rate share also exceeded 10%, at approximately 10.97%.

If we observe the changes in hash rate shares of major mining pools over shorter time periods, we can see that competition among large BTC mining pools remains very intense. This year, aside from F2Pool maintaining a stable highest hash rate share, the rankings of other pools have been constantly changing. For example, the hash rate shares of Poolin, BTC.com, and AntPool, ranked second to fourth, have alternated, while the competition among pools with lower hash rate shares has been even more fierce.

However, similar to exchanges, the scale effect of large BTC mining pools has become evident, making it increasingly difficult for relatively smaller pools to compete with large pools. Even a roadside pool that "parachuted" into the market in May with over 5% hash rate share failed to maintain strong momentum for further share acquisition.

Another noteworthy development in the mining pool sector in 2020 was the emergence of exchange mining pools. Huobi.pool, OKExPool, and Binance Pool all established BTC mining pools, with their hash rate shares at the end of the year being approximately 9.39%, 3.57%, and 11.48%, respectively. The trend shows that Huobi.pool and Binance Pool's hash rate shares have been continuously increasing, especially Binance Pool, which saw a significant increase of over 4 percentage points in hash rate share during the fourth quarter when BTC prices were at their best, rising 4 ranks in market position.

The hash rate share of ETH mining pools was relatively stable, with a higher concentration. The combined average hash rate share of the three major pools, SparkPool, Ethermine, and F2Pool, reached 75.51%, with SparkPool's average hash rate share accounting for 32.69%. Looking at the monthly changes in hash rate share, aside from F2Pool suddenly occupying 73.47% of the total network hash rate in June, SparkPool consistently held the top position among ETH mining pools during other times.

From the correlation between block production and hash rate share among major BTC mining pools, Binance Pool was the "most rhythmic" pool in 2020, with a high positive correlation between hash rate size and block production throughout the year, with a Pearson coefficient exceeding 0.9. Other pools with good "rhythm" include Huobi.pool and ViaBTC, which exhibited a weak positive correlation between hash rate and production, with Pearson coefficients around 0.6 or higher. A common feature of these pools is that in the fourth quarter, when BTC prices began to rise rapidly, their hash rate shares also increased.

In contrast, the hash rates of Poolin and SlushPool were negatively correlated with block production, with Pearson coefficients of approximately -0.81 and -0.63, respectively. Their common characteristic was that in the fourth quarter, when BTC prices began to rise rapidly, their hash rate shares decreased, leading to a decline in block production.

Other large pools, such as AntPool, BTC.com, and F2Pool, showed relatively stable performance, with hash rates and block production not forming a statistical correlation.

In Ethereum mining pools, Spider Pool was the only pool with a high positive correlation between hash rate and block production, with a Pearson coefficient of about 0.77, indicating that when Spider Pool's hash rate share increased, ETH prices also rose. In contrast, UUPool and BWPool exhibited a high negative correlation between hash rate and block production, with Pearson coefficients around -0.7.

Similar to BTC mining pools, large ETH mining pools like SparkPool, Ethermine, and F2Pool also did not show a statistical correlation between hash rate and block production, with block production remaining relatively stable throughout the year.

Theoretically, the hash rate of a mining pool is directly related to its income; the higher the hash rate, the greater the block production, and thus the higher the income. On the other hand, income is also related to coin prices, but since the distribution of coin-based income does not necessarily convert to fiat-based income, the impact of coin prices on mining pool income has a significant degree of randomness. To uniformly observe the income of mining pools, this article will estimate mining income based on the daily coin price.

According to statistics, the mining pools with the highest hash rate shares, F2Pool, Poolin, BTC.com, and AntPool, also had the highest estimated annual outputs, each expected to exceed $500 million in 2020, with F2Pool's expected output being the highest at $888 million. If we estimate based on a 3% service fee, there are 6 pools with service fee income exceeding $10 million, with F2Pool and Poolin exceeding $20 million.

Compared to 2019, under the same fee rate, different pools have seen changes in income due to variations in hash rate shares. For instance, BTC.com had the highest average hash rate share and income in 2019, approximately $26.22 million, while F2Pool had the highest average hash rate share and income of about $26.64 million in 2020. However, overall, BTC mining income has remained relatively stable. According to statistics, a pool with 1% of the total network hash rate was expected to earn about $1.54 million in service fees in 2019 and about $1.49 million in 2020, which is related to the BTC mining sector maintaining a similar total output scale over the past two years.

To estimate the total annual output of ETH mining pools, PAData equates the hash rate share of each pool to its share in the total output of ETH mining, thereby estimating the total annual output of each pool. However, this annual output total does not consider uncle blocks and empty block rewards, so the actual total output may exceed the estimated value.

According to statistics, the three pools with the highest hash rate shares, Spark Pool, Ethermine, and F2Pool, are also the pools with the highest output revenue, with estimated annual outputs of approximately $772 million, $582 million, and $257 million, respectively. Notably, the annual output total of the leading ETH pool, Spark Pool, has already matched that of the leading BTC pools. Additionally, the estimated annual output total of Spider Pool and Nano Pool also exceeded $10 million. If we estimate the mining service fee income based on a 2% service fee rate, only Spark Pool and Ethermine's estimated service fees exceed $10 million.

Bitmain still dominates BTC mining machines, while the Ethereum custom machine market is on the rise

In addition to mining pools, mining machine manufacturers are also important participants in the mining industry, and compared to mining pools, they are positioned further upstream in the industry. According to F2Pool's monitoring of the mining machine market, as of the end of January 2021, there were a total of 101 BTC mining machines on the market. By brand classification, Bitmain, Bitfury, and Canaan Creative are the three manufacturers producing the most mining machines, with Bitmain producing a total of 25 BTC mining machines, Bitfury and Canaan Creative producing 19 and 17 BTC mining machines, respectively. One of the traditional mining machine giants, Ebang International, only produced 10 mining machines, comparable to the number produced by Innosilicon.

Among the 15 mining machines with the best profit performance on the market, Bitmain and Bitfury each have 4 models, while Canaan Creative and Innosilicon each have 2 models. The average hash rate of these 15 mining machines reached 89 T, and at a cost of 0.34 yuan per kilowatt-hour, the average daily net profit of these machines is about 94.33 yuan. Four models have daily net profits exceeding 100 yuan, namely the Shennong M30S++, Antminer S19 Pro, Shennong M30S+, and Antminer S19, with the highest profit being Shennong M30S++ at 124.5 yuan.

ETH mining machines can be divided into assembled machines and custom machines. The rapid development of custom machines in 2020 marked the gradual maturity of ETH mining and is a development direction for the future of ETH mining. According to F2Pool's statistics, as of the end of January 2021, there were a total of 60 ETH mining machines on the market, with 35 assembled machines, accounting for 58.33%, and about 25 custom machines, accounting for about 41.67%.

Among the assembled machines, there are 13 machines with RX graphics cards and 10 machines with GTX graphics cards, together accounting for 65.71% of the total assembled machines. Among custom machines, brands like Innosilicon, Panda, and WolfGod are relatively mainstream, with Innosilicon launching 5 custom mining machines, Panda and WolfGod each launching 3 custom machines, together accounting for 47.83% of the custom machine market.

From the revenue performance of ETH mining machines, among the 15 machines with the highest daily net profits, there are 8 assembled machines and 7 custom machines. Among custom machines, Innosilicon has 4 models with good revenue performance, making it the strongest in revenue capability among all custom manufacturers.

The average hash rate of the 15 mining machines is about 747 M, with the highest being the Linzhi Phoenix at 2600 M, far exceeding other ETH mining machines on the market. If we calculate the cost at 0.34 yuan per kilowatt-hour, the average daily net profit of these 15 mining machines is about 418 yuan, with the Linzhi Phoenix's daily net profit reaching 1477 yuan, also far exceeding other mining machines on the market.