The turning point of Bitcoin mining industry and the economic cycle behind it

Author: Little Parker

Editor: Hao Fangzhou

summary:

The bear market continues to squeeze the profit margins of all links in the cryptocurrency market, and the mining industry, which plays the role of cryptocurrency producers, is no exception. For example, Bitcoin, which accounts for half of the cryptocurrency market, had an overall mining revenue of only $190 million in February 2019, down 10% from the beginning of the year.

Even so, mining companies operating as enterprises still maintain high profits. According to Diar research, the profit margin of "big miners" remained at 59% in October 2018. In contrast to "big miners", "individual miners" have "no profit to mine" since September 2018.

From the above figure, we can see that the scale effect of Bitcoin mining has gradually emerged: large miners with abundant resources have stronger profitability, and small miners on the edge of breaking even either withdraw or form groups.

Behind the phenomenon, the Bitcoin economic system has its own cyclical nature. As a large number of miners exit due to lack of profitability, the computing power of the entire network decreases, the difficulty of calculation decreases, the mining cost becomes lower, and the profit margin increases, which in turn attracts new computing power to enter the market... This cycle continues until the last satoshi (the smallest unit of Bitcoin, 1 satoshi = 0.00000001 BTC) of Bitcoin is mined.

What is the essence of mining? How do the various links in the industrial chain transfer value? What core logic do they follow? What are the key factors that affect mining returns? What stage is the market in? How long can Bitcoin be mined?

These questions are crucial for those who still want to enter and stick to the mining industry. Odaily Planet Daily Research Institute will try to answer these macro questions in this article and focus on more details of the mining industry in subsequent reports. Please continue to pay attention and communicate with our analyst Xiao Pike (lmm662381). The pdf version of the report can be downloaded from the Odaily Planet Daily Research Institute website.

Table of contents

1. The basic principles of Bitcoin mining

2. Three major factors affecting Bitcoin mining

2.1 Computing Power

2.2 Computing Difficulty

2.3 Bitcoin Price

2.4 Feedback cycle between computing power, computing power difficulty and price

3. The main cost of Bitcoin mining: electricity price

3.1 Power industry relations and procurement channels

3.2 Minimum electricity price fluctuation space

4. Bitcoin Mining Hardware Foundation - Mining Machine

4.1 Miners’ perspective: cost-effectiveness of mining machines

4.2 Manufacturer’s perspective: Pricing rules for mining machines

5. Current mainstream mining method - mining pool

5.1 Mining Pool Operation Rules

5.2 Mining income settlement model

5.3 Classification of mining pools

5.4 Mining pool computing power distribution

6. Overview of Bitcoin Industry Chain

7. Bitcoin mining stage

8. References and Thanks

1. The basic principles of Bitcoin mining

Bitcoin is a peer-to-peer electronic cash system in which each transaction record is distributed and stored throughout the entire network, rather than in a traditional single central database.

In order to ensure that all participating nodes have consistent transaction records, the Bitcoin system stipulates competitive accounting. The specific rules are as follows:

1. Competitors use SHA-256 calculations to find the value that meets the requirements among a bunch of random numbers. Whoever has greater computing power and faster calculation speed has a greater chance of finding the "answer" first.

2. The Bitcoin system groups transaction data (i.e. packs them into a block) every 10 minutes on average. The competitor who finds the "answer" first obtains the right to keep accounts and gets Bitcoin as a reward. The system will adjust the mining difficulty (with a certain lag) according to the computing power of the entire network to ensure that a block is produced every 10 minutes on average.

3. The mining difficulty is adjusted every 2016 blocks (about 14 days). The block height is an integer multiple of 2016, which is the block corresponding to the system when the mining difficulty is adjusted.

4. Bitcoin mining income includes block rewards and transaction fees. The block reward was initially 50 BTC. The system stipulates that the block reward is reduced by half every 210,000 blocks (about 4 years) until it is reduced to 1 satoshi, the smallest unit of Bitcoin. Therefore, the block reward was adjusted to 25 BTC after 2012, 12.5 in 2016, and the next halving will be in 2020.

5. Transaction fees (also known as mining fees) are paid by the transferor to the miner to reward the miner for providing sufficient computing power to ensure the security of the network system. Transaction fees generally fluctuate between 0.5% and 2% of the block reward of the day. Since the block reward is halved every 4 years, transaction fees will gradually become the main income of miners. Bitcoin is expected to be fully mined in 2140, when the mining income will be equal to the transaction fee.

6. The handling fee for each transaction depends on the size of the transaction (usually in kilobytes). When the network is congested, more people submit accounting requests at the same time. Users can increase the handling fee to encourage miners to prioritize packaging and shorten the transaction confirmation time. For example, at the end of 2017, the volume of Bitcoin transactions skyrocketed, and the average handling fee was as high as $40 per transaction. The miner's handling fee was as high as 30% of the block reward for that month.

7. The transaction size is also related to the amount of money transferred in and out, that is, the larger the transaction amount, the higher the fee/kilobyte.

2. Three major factors affecting Bitcoin mining

Hashrate, hashrate difficulty and Bitcoin price are the key factors that affect mining costs and revenues. Paying attention to these three indicators and understanding the relationship between them will help you determine market trends.

2.1 Computing Power

Hashrate refers to the number of times the hash value is calculated per second, which is used to measure the computing power of miners. The higher the hashrate, the greater the probability of mining a block.

More intuitively, the commonly used units of computing power have changed from the original hash value per second (H/s) to KH/s (kilo hash/s), MH/s (million hash/s), GH/s (billion hash/s), TH/s (trillion hash/s), PH/s (quadrillion hash/s), and EH/s (exabyte hash/s).

Currently, the total network computing power of Bitcoin has reached 47EH/s. This means that the computing power and electricity expenditure of mining hardware are required to be high. Once the computing power of the Bitcoin network exceeds 500EH/s, it will lead the hardware processing power into the zetahash era.

In general, the iteration of mining methods and hardware equipment has promoted the growth of Bitcoin network computing power and the rise of mining costs, and the growth of computing power and the rise of prices have forced the former to continue to evolve. Looking back at history, Bitcoin mining has gone through the development process of CPU - GPU - FPGA - ASIC - mining pool.

ASIC mining chips are the core of mining machines and the key to the entire device. The emergence of ASIC mining chips has led to the widespread application of ASIC mining machines. In addition to the original Butterfly Lab, more than a dozen different companies have emerged to provide customized ASIC mining machines, and mining machine hosting models have also emerged. Today, the growth of global computing power can be directly attributed to the application of ASIC mining machines.

2.2 Computing Difficulty

The difficulty of Bitcoin mining is a measure of the difficulty of mining. The greater the difficulty of mining, the more difficult it is to mine a block. The Bitcoin system controls the average time required to mine a block by adjusting the difficulty target value (Target, calculated by the value of the bits field) in the block header.

Target is a string of 256 bits in length. In other words, Target has approximately 2^256 possible values.

Adjusting Difficulty means adjusting the proportion of Target in the entire output space. The higher the Difficulty, the smaller the Target. For example: Mining is like shooting, all the bullets fired will fall on a large target. Target is a circle on this large target. The smaller the range, the higher the difficulty of being shot. Adjusting Target means adjusting the proportion of this circle on the entire target.

In addition, the range of difficulty target adjustment is limited to 4 times. For example: Assume that the 2016 blocks in the last difficulty target adjustment cycle were mined in only 7 days due to a surge in computing power. By adjusting the difficulty target, the difficulty target is halved, and the average block time can be maintained at about 10 minutes. However, if the computing power surges and the first 2016 blocks are mined in only 1 day, then the minimum difficulty target can be adjusted to a quarter of the original.

2.3 Bitcoin Price

Many people compare Bitcoin to digital gold. In addition to its limited supply similar to gold, its price is also affected by supply and demand like gold.

We believe that the size of computing power and mining difficulty determine the supply of Bitcoin, while the awareness of Bitcoin, its popularity, and the policies of various countries determine the demand for Bitcoin.

If the demand is high and the supply is not met, it will cause the price of Bitcoin to soar. Since the number of Bitcoins is limited, the supply is limited. Under the premise that the public is optimistic about Bitcoin, the price will continue to rise. However, the specific factors that affect the price of Bitcoin are mainly the following:

(1) Community consensus

Trust plays a vital role in the cryptocurrency environment, and the consensus of Bitcoin community developers is an important factor that affects price fluctuations. You can keep up with the community's ideas by paying attention to and studying Bitcoin forums, which helps to understand price trends.

(2) Technology Update

Technological innovation will also affect the price of Bitcoin. For example, the implementation of Segregated Witness and Lightning Network solutions will improve the transfer efficiency of the Bitcoin system.

(3) Policies of various countries

Because Bitcoin is not regulated by any government, it has become the object of government regulation efforts. Whenever there is an official announcement regarding digital currency regulation, Bitcoin prices fluctuate. For example, in Japan, cryptocurrencies have been recognized as legal payment since April 2017, and several stores have begun accepting Bitcoin, which has triggered a rise in Bitcoin prices.

(4) “Giant Whale” Activities

Sometimes the reason behind price fluctuations is because of well-funded holders, also known as whales. These whales are large in size and small in number, and their actions can affect the rise and fall of Bitcoin prices, but preventive measures can be taken by tracking their actions.

(5) Security incidents

If hackers attack an exchange, it will undermine users' confidence in cryptocurrencies and cause prices to plummet.

(6) Influence of public opinion

The attitude of the mass media towards the crypto industry influences potential investors and merchants. For example, hype about Bitcoin can cause the price to soar, while negative news can cause the price to drop.

2.4 Feedback cycle between computing power, computing power difficulty and price

Because of the self-regulating mechanism of the Bitcoin system and the laws of market economics, there is a feedback cycle between computing power, computing difficulty and price.

On the premise that the infrastructure remains unchanged, the higher the price of Bitcoin, the more attractive mining becomes, the computing power of the entire network increases, the competition among miners becomes more intense, the difficulty of mining becomes higher (the system will adjust the difficulty in the next cycle with a certain lag), and at the same time the cost increases, mining becomes no longer profitable, and small miners withdraw or are merged.

As the computing power of the entire network decreases, the difficulty of mining also decreases (plus the gain effect of the system itself adjusting according to the difficulty of the previous cycle). At this time, miners who have exited the market tend to sell, and the price of Bitcoin decreases accordingly (which is beneficial to the remaining large miners). As the difficulty decreases, the profit of mining increases, the attractiveness of mining becomes greater again, new miners join, and the competition becomes fierce again, and the cycle continues.

As can be seen from Figure 9, the trend of the computing power curve and the difficulty curve is basically the same. The greater the computing power, the greater the difficulty. In 2018, the price of Bitcoin fell, but the computing power added to mining increased, resulting in an increase in difficulty.

We can use the indicator of "computing difficulty/price" to compare the changes in "income factors" and "cost factors" to determine the profit margin of mining.

In Figure 9, when the price dropped in 2018 and both the hashrate and difficulty increased, the "difficulty/price" curve showed an upward trend. In December 2018, some hashrate was withdrawn due to high costs, the hashrate of the entire network dropped, the mining difficulty dropped, and the Bitcoin price also dropped. This in turn caused the "difficulty/price" curve to drop and mining profits to increase.

In the long run, the "difficulty/price" curve is generally upward. But in the short run (Figure 10), this line fell in 2017, indicating that the price growth rate in 2017 exceeded the growth rate of computing power difficulty, ushering in a boom for miners. In 2018, the trend turned upward, and the growth rate of computing power difficulty far exceeded the growth rate of Bitcoin price, ushering in a recession for miners. It is worth noting that before the beginning of 2018, this coefficient was always below 0.5. From September 2018, the coefficient began to be greater than 1, until it reached its peak in December, ushering in the worst moment of the mining industry, that is, mining accidents. From December 2018 to the end of 2019, this coefficient fell, Bitcoin mining became profitable again, and miners rejoined mining.

3. The main cost of Bitcoin mining: electricity price

Bitcoin mining requires a lot of computing power, and computing power requires machines and electricity to operate and maintain, so we can regard Bitcoin as a digital product converted from electrical energy. Therefore, the price of machines and electricity actually constitutes the main cost of Bitcoin mining. Of course, in addition to this, there are also fixed costs such as infrastructure.

Among them, fixed costs such as infrastructure are allocated to each machine, about 700-800 yuan per mining machine (including converting high voltage electricity into low voltage electricity, mining machine cooling, dust removal, mining machine power supply, mining machine container, manpower, etc.); mining machines are one-time investments, and you only need to find the mining machine with the highest "cost-effectiveness" on the market (detailed in the next part); and electricity prices are the largest variable cost, and therefore become the main factor affecting Bitcoin mining.

According to Bloomberg BNEF (Bitcoin in Energy Crisis as China Cracks Down, Lu et al., 10 Jan 2018) data, 78% of Bitcoin's computing power comes from China. Therefore, China's electricity price fluctuation space, procurement channels and related policies are the focus of Bitcoin miners.

3.1 Power industry relations and procurement channels

From the perspective of production process, power production is divided into five links, but from the perspective of the entire power industry, it is generally believed that the power system is divided into four links, namely generation, transmission, distribution and sales. Power generation is the generation of electricity by power plants (power plants include thermal power, hydropower, wind power, solar power and nuclear power, and thermal power is also divided into gas, coal, biomass, waste incineration, etc.), transmission is the long-distance transmission of electricity (mainly using AC and DC ultra-high voltage technology), distribution is the link that directly connects to users and distributes electricity to users, and sales of electricity, as the name implies, means that if you want to buy electricity, I will sell it to you.

These four links constitute a huge power system. The power generation link is completed in the power plant, and the other three links need to rely on the power grid to ultimately deliver electricity to users.

Electricity procurement channels include power plants, power grid companies, power sales companies, and power sales intermediaries.

It should be noted that the targets of direct power supply transactions from power plants are large users with annual power consumption of 10 million kWh, power sales companies and power grids. The targets of power sales companies are users with annual power consumption of less than 10 million kWh and large users who do not choose power generation companies on their own. The users of the two are different, and the bidding is also fundamentally different.

Based on a rough calculation of the Antminer S9j model, an annual electricity consumption of 10 million kWh requires a mining farm with 850 mining machines, so mining farms with a scale of more than 850 mining machines can theoretically purchase electricity directly from power plants. However, direct power purchases are highly competitive and involve government relations, so government resources are a very important factor.

3.2 Minimum electricity price fluctuation space

For miners, the most important issue is how to find the cheapest electricity price. Knowing the floating space of the lowest electricity price can give you an idea of ​​the profit margin of mining.

Electricity prices vary greatly depending on the generation mode, but China's electricity is mainly coal-fired and hydropower, of which coal-fired power is the main power output. In 2018, the market traded electricity volume was 1045.9 billion kWh, accounting for 76% of the total market sales, and hydropower was 205.6 billion kWh, accounting for 15% of the total market sales. Hydropower is mainly generated during the flood season (from March to April to August and September each year), and the output power is small and not continuous, but it is cheaper because of the low cost of power generation.

The power generation industry involves multiple links, which results in different electricity prices, such as grid-connected electricity price, transmission electricity price, distribution electricity price, and sales electricity price. The grid-connected electricity price refers to the price at which the power grid buys electricity from the power plant; the transmission electricity price and distribution electricity price refer to the price of electricity that is increased due to losses when transporting electricity to various substations; the sales electricity price refers to the price of electricity purchased from the power sales company. The grid-connected electricity price is less than the sales electricity price, so miners are mainly concerned about the grid-connected electricity price.

In the Bitcoin mining industry chain, most miners will host their mining machines at large miners (mine owners). Therefore, in addition to the on-grid electricity price (generally referred to as the bare electricity price by miners, which refers to the contractual electricity price signed between the mine owner and the power plant), there is also the managed electricity price (the contractual electricity price signed between the miner and the mine owner after the miner hosts the mining machine at the mine). Below we will analyze the minimum electricity price fluctuation space of the on-grid electricity price and the managed electricity price.

(Note: The pricing policy for electricity prices is mainly based on the document "Notice of the National Development and Reform Commission on Printing the Implementation Measures for Electricity Price Reform")

(1) On-grid electricity price:

According to public information, the on-grid electricity price of coal-fired power is generally between 0.27 yuan and 0.47 yuan per kilowatt-hour, and the on-grid electricity price of hydropower is generally between 0.2 yuan and 0.4 yuan.

In addition, as for thermal power (Inner Mongolia, Xinjiang, Shanxi, Shaanxi), there is also a kind of "mine-mouth power", that is, coal is directly put into the power plant next to it after it is mined from the mine for power generation. The marginal cost of mine-mouth power is low, and the on-grid power price is equivalent to the price of hydropower. According to the survey, the cost price of mine-mouth power is generally 0.16 yuan to 0.18 yuan per kilowatt-hour, and the on-grid power price is 0.26 yuan to 0.28 yuan per kilowatt-hour.

Mine-mouth power stations are mainly distributed in areas rich in coal resources, such as Inner Mongolia and Shanxi. Taking Inner Mongolia as an example, the eastern region of Inner Mongolia began to organize direct power trading in 2014. At present, the access conditions for large users require a voltage level of more than 10 kV and an annual power consumption threshold of 100 million kWh. The transaction is relatively mature, and the government provides generous subsidies, so the electricity price can be 0.06 yuan , and it is not open to power sales companies.

Still roughly calculating based on the Ant S9j mining machine, a scale of more than 9,000 mining machines is needed to build a factory next to the pit-mouth power station in the eastern part of Inner Mongolia. In addition, corresponding government resources are also needed.

The western part of Inner Mongolia has been conducting electricity trading for nearly 8 years. In 2017, the entry conditions for large users required users to consume more than 10 million kWh of electricity per year, and all transactions were made by the government, and power sales companies were not allowed to enter the market. Still roughly calculating based on the Antminer S9j model mining machine, a mining farm with a scale of more than 850 mining machines can be built in the western part of Inner Mongolia, and government resources have also become a core competitiveness.

Based on the above situation, the minimum price fluctuation range of thermal electricity that the mine can get is 0.06 yuan to 0.47 yuan per kilowatt-hour.

The cost electricity price of hydropower stations is 0.04-0.09 yuan per kilowatt-hour, while the sales price of hydropower that some state-owned power companies do not collect can be as low as 0.12 yuan per kilowatt-hour, that is, the minimum price fluctuation range of hydropower is 0.12-0.4 yuan, but legal risks and power stability need to be considered.

(2) Managed electricity price

The hosting electricity price is the contract price signed between miners and mine owners when they host their mining machines in a mine. The hosting electricity price changes according to the price of Bitcoin. When the price of Bitcoin drops, the electricity price increases, and when the price of Bitcoin rises, the electricity price decreases, but in any case it will not be lower than the on-grid electricity price. When the price of Bitcoin drops to the marginal cost, the electricity price will rise to the shutdown price, which is currently 0.38 yuan per kWh on average.

In fact, the miners who chase the flood season every year are small miners on the edge of Bitcoin mining costs. The large miners (mining farms) who find the pithead electricity price are in a state of continuous profitability. Therefore, electricity resources are the core competitiveness of large miners.

4. Bitcoin Mining Hardware Foundation - Mining Machine

In the above, we introduced that in addition to electricity prices, the cost of Bitcoin mining also includes the investment in mining machines. Mining machines not only conform to the basic logic of industrial components, but also have financial attributes due to their strong correlation with the cryptocurrency market. We will take the mainstream ASIC mining machines on the market as an example to introduce them.

4.1 Miners’ perspective: cost-effectiveness of mining machines

4.2 Manufacturer’s perspective: Pricing rules for mining machines

As the price of Bitcoin rises, the demand for mining machines is driven up. Under the oligopoly, upstream mining machine suppliers have absolute pricing power over mining machines, and the price of mining machines has also risen. At the beginning of 2018, the Antminer S9 was once sold at a futures price of 30,000 yuan in Huaqiangbei. So, how should mining machines be priced?

According to the model of Guosheng Securities Research Institute: the current mining machine price P = fixed payback period D × current mining daily income (mining income R-mining cost C). That is, the current mining machine price is positively correlated with the current mining daily income.

Based on the above formula and linear regression results, Guosheng Securities believes that the pricing strategy of mining machine suppliers can be derived: the payback period is fixed at about 180 days, and the price of mining machines is dynamically adjusted according to the current mining income.

It is worth noting that such a pricing strategy is based on the assumption that the daily mining income will remain unchanged in the future, so that the payback period can be guaranteed to be 180 days. If the price of the currency rises in the future, resulting in an increase in daily mining income, the actual payback period will be shortened, otherwise it will be longer.

In addition, the actual calculation of the static payback period by the mining machine often deviates from 180 days, mainly due to:

(1) The price adjustment of mining machines is relatively low frequency compared to the mining income and Bitcoin price changes, which causes the static payback period to deviate. Therefore, the same mining machine may be shipped three times in a month, so the price is adjusted three times in a month.

(2) When setting actual prices, mining machine suppliers will also take into account the expected changes in future mining revenues. Especially when the market is rising, they will make optimistic judgments about future mining revenues, thereby increasing the price of mining machines and lengthening the static payback period.

(3) Mining machine suppliers are unwilling to reduce their own profit margins when mining revenues decline. When the price reduction is smaller than the decline in mining revenues, the static payback period becomes longer.

To further refine the above view, mining machines can be regarded as "call options" on the price of Bitcoin. When miners buy mining machines, they obtain a relatively certain number of Bitcoins in the future at a certain price (although the future computing power of the entire network is uncertain, the approximate range can be estimated). The motivation for miners to buy mining machines comes from their prediction that the price of Bitcoin will be higher than now, which is equivalent to buying Bitcoin at a price lower than the market price in the future, thereby obtaining excess returns.

5. Current mainstream mining method - mining pool

As computing power increases, the probability of mining Bitcoin becomes smaller and smaller. At present, the computing power of the entire network has reached 47Eh/s. In order to pursue sustained and stable income, mining pools have become the mainstream mining method. Simply put, a mining pool is a collection of computing power. When everyone concentrates computing power on a mining pool, the probability of mining a block will be greatly increased, and then the income will be distributed according to each person's computing power ratio. The core work of the mining pool is to assign tasks to miners, count the workload and distribute the income. Compared with the Solo mode, the expected value of miners' income has not changed, but the income is more sustained and stable.

5.1 Mining Pool Operation Rules

We will introduce how mining pools work through some basic concepts:

(1) Mining Pool Agreement:

The "mining pool" coordinates hundreds or thousands of miners through a proprietary protocol. After creating a mining pool account, miners set up their mining machines to connect to the mining pool server. When the mining machine runs mining online, it needs to maintain a connection with the mining pool server and work synchronously with other miners. Common protocols include: Stratum (STM protocol) & GetBlockTemplate (GBT protocol) and the outdated GetWork (GWK protocol).

(2) Miners in the mining pool:

Earn relatively stable rewards based on your computing power contribution.

(3) Mining pool administrator:

By charging a certain percentage of handling fees, mining pool administrators can also participate in computing power contribution as Solo miners.

(4) Mining pool entry conditions:

The mining pool is open to all miners. After a miner in the mining pool successfully mines, the block reward is paid to the fixed Bitcoin wallet address of the mining pool. The reward does not belong to the miner who mines the mine, but to the entire mining pool.

(5) Reward distribution mechanism:

The mining pool will set a "threshold" for reward distribution. This threshold is the difficulty target for the TargetHash value calculated each time, which is usually less than 1/1000 of the difficulty of the Bitcoin network. For example, the mining difficulty of the entire network is the Hash value with 10 consecutive zeros at the beginning, and the difficulty threshold set by the mining pool is the Hash difficulty with 7 consecutive zeros at the beginning; after a miner in the mining pool successfully mines the ore, those miners who successfully perform Hash calculations and meet the difficulty threshold of the mining pool can share the rewards.

The rewards are not distributed immediately after a mine is mined, but when the mining rewards accumulate to a certain amount set by the mining pool, the rewards are distributed once, or settled at a fixed time (such as daily).

(6) Luck value of the mining pool:

Bitcoin mining is essentially probabilistic, and the speed of block generation is sometimes fast and sometimes slow. There is a certain probability factor in the theoretical value of luck in the mining pool, and the speed of block generation is sometimes fast and sometimes slow. The luck value fluctuates around the theoretical value of 100%. When the luck value is high, the income of the mining pool will increase, otherwise it will decrease, but this will only affect users who choose the PPLNS income model.

5.2 Mining income settlement model

There are various revenue settlement models in the market, and the mainstream mining models include: PPS, PPLNS, PPS+ and SOLO models.

5.3 Classification of mining pools

There are mainly managed mining pools and P2P mining pools on the market. Due to the low efficiency of P2P mining pools, this model has gradually faded out of the market.

5.4 Mining pool computing power distribution

The computing power of mining pools can be divided into Bitmain and non-Bitmain. BTC.COM, AntPool, BTC.TOP, and ViaBTC all belong to Bitmain, accounting for 46% of the total network computing power, and other non-Bitmain systems account for 54%.

At the beginning of last year, Bitmain's mining pool accounted for 53% of the total network computing power, but now it has dropped to 39%. In addition, according to Dir data, unknown miners are changing the distribution of mining pool computing power. In December 2018, mysterious miners mined 22% of the blocks in the Bitcoin network, while at the beginning of last year this proportion was only 6%. This change has reduced the dominance of mining pools. Currently, mysterious miners control more than 22% of BTC's computing power.

The most basic function of a mining pool is to gather the computing power of miners to mine together. The difference in technology is very small. At this stage, the competition is about computing power and service quality. Bitmain-affiliated mining pools have inherent advantages such as favorable prices for mining machines and fast acquisition of mining machines; non-Bitmain-affiliated mining pools have gained the support of miners with their own reputation and differentiated services, represented by F2Pool in China and SlushPool abroad.

6. Overview of Bitcoin Industry Chain

At this point, we have a basic understanding of the mining machine and mining pool industry. The following figure will take mining as an industry and depict the overall picture of the participants and the relationship between the links.

7. Bitcoin mining stage

Finally, we have to try to answer the question about "endgame", that is, how long can Bitcoin be mined?

Some people say that Bitcoin transactions are very similar to commodities, which is not entirely accurate. Because when the production of commodities stops, the transactions of commodities can still be carried out, but when miners stop mining, Bitcoin will die immediately, because no one verifies the transactions, Bitcoin cannot circulate, and currency that cannot circulate is worthless, that is, the "death spiral". That is to say, if the mining profit drops to zero, the value of Bitcoin will also drop to zero.

At present, all links are reserving profit margins for enough miners to ensure that the Bitcoin system is protected from attacks. However, when block rewards are completely converted to transaction fees, the raised transaction fees will prevent users from using Bitcoin to transfer money. At that time, stabilizing transaction fees at a reasonable balance point will become a new focus. The mining industry will also usher in a new turning point and become an industry that provides basic computing power to other fields. Opportunities will tend to large enterprises with power resources and operational strength.

Theoretically, this turning point is in 2140, but from the Bitcoin halving schedule (Figure 4), we can see that in 2048, the Bitcoin block reward will be less than 0.1 BTC, which is less than the average transaction fee. By then, either the Bitcoin price will rise enough to support the mining cost, or there will be cheaper electricity. Otherwise, Bitcoin may really face a "death spiral".

In addition, according to Figure 2, small miners have lost their profits since September last year. The price of $7,193 at that time can be regarded as the marginal price. The current block reward is 12.5 BTC. It is roughly estimated that if the price of Bitcoin cannot increase by more than 125 times, or about $900,000, after 2048, Bitcoin mining will no longer be attractive.

8. References and Acknowledgements

References:

Bitcoin Developer Reference, Target nBits

How is difficulty calculated?

Mastering Bitcoin Chapter 8 Mining and Consensus

Introduction to blockchain | 10 numbers related to Bitcoin

What determines the price of Bitcoin?

What is the relationship between power plants, power grids, power companies, power supply bureaus, and power generation groups?

What is the difference between the grid bidding price of power generation enterprises and the electricity sales price of power sales enterprises?

Computer Industry: Cryptocurrency Mining Market Special Report

In-depth report on the Bitcoin mining industry: A new battlefield for chip competition, a new carrier for consensus creation

Decrypting Cryptocurrencies: Technology, Applications

And Challenges

Bitcoin network computing power exceeds 1 Petahash per second

Special thanks to:

Wang Xiaoyi, COO of Mining and Ocean University, David Li, CEO of Rawpool, Geng Lei, Marketing Director of Rawpool, Yan Danlin, Ph.D. in Industrial Economics from China University of Geosciences, Aaron Yuan, CEO of ENT Public Chain, and Yang Xin, Product Manager of Bitmain.