It has been a week since Canaan Inc. (NASDAQ: CAN) was listed on the Nasdaq, and the price of Bitcoin has not changed. As the "first stock of blockchain mining machines in the world" and the world's second largest mining machine manufacturer, Canaan's listing is not a small matter in the industry. Especially in the context of Bitmain and Canaan's own failed attempts to go public, it is not easy for the CAN stock code to appear on the Nasdaq screen. However, the price of Bitcoin has not increased. Is the listing of mining machine companies not a good thing? Of course it is a good thing. Even though some people believe that listing means that there is no money, or that listing means that the industry is mature and the ceiling of the upward space is visible, in the eyes of most people, the listing of mining machine companies is a perfect promotion of Bitcoin for outsiders. Tens of millions of users reached by Nasdaq and media can learn about Bitcoin through Canaan Creative. From this perspective, this is definitely a good thing. However, despite this favorable situation, Bitcoin did not see any favorable market conditions. So, will the price of Bitcoin remain unchanged next time there is a positive trend? For example, the production capacity of Bitcoin will be halved in the middle of next year. There is a rule in Bitcoin's economic model that its output is halved every four years. From the time the founding block was mined in 2010 to the present, the output of each block of Bitcoin has been halved twice, from 50 Bitcoin rewards per block to 12.5 Bitcoin rewards per block. The next halving is expected to be in May 2020. That is, six months later, the output of each block will be only 6.25 bitcoins. From historical data, the reduction in production efficiency is good for Bitcoin, and all Bitcoin enthusiasts are looking forward to this good news that only occurs once every four years. Especially the miners. According to the current Bitcoin price, some mining machines have reached the shutdown price. That is, the income from mining with this type of mining machine can no longer cover the mining cost, and mining with this type of mining machine is completely losing money. Even the most mainstream and stable Antminer S9 mining machine on the market has approached the shutdown price. So if the price of Bitcoin does not rise when the Bitcoin production is halved in May next year, we have three questions: How many mining machines will become scrap metal? Will the Bitcoin computing power drop sharply? Will mining accidents occur? Will the mining machine become scrap metal? To estimate this change, you need to understand two key data: electricity cost ratio and energy consumption ratio. The electricity cost ratio of the mining machine is the most important link between the Bitcoin price and the mining machine's revenue. For mining machines, electricity costs are the main cost, and the electricity cost ratio refers to the ratio of the mining machine's daily electricity consumption to the daily mining revenue. According to the electricity cost ratio, the current revenue of the mining machine can be roughly calculated. For example, if the electricity cost ratio of a mining machine is 30%, then the remaining 70% is the mining machine's revenue. Energy consumption ratio is the ratio of the power of the mining machine to the computing power, which indicates how much power is required for each computing power. Looking at the computing power or power alone cannot tell the performance of the mining machine. The energy consumption ratio is the main data for comparing the performance differences between mining machines. The lower the energy consumption ratio, the stronger the mining performance of the mining machine and the more energy-saving it is. After understanding these two data, let's look at the halving of Bitcoin production. When the price of Bitcoin remains unchanged, then when the mining machine's profit is reduced by half, then the mining machines with electricity costs exceeding 50% will theoretically have to be shut down, and continuing to mine will mean losing money. According to the data from F2Pool, the energy consumption ratio of mining machines with an electricity cost ratio of 50% is about 60W/T. In other words, at the current price, after the halving occurs, all mining machines with an energy consumption ratio of more than 60W/T will theoretically have to be shut down. So, how many mining machines are going to be shut down in the market now? BlockBeats interviewed a number of miners and mine owners, and taking into account factors such as thermal power and hydropower, it was found that the current replacement rate of new mining machines is about 50%, that is, the proportion of mining machines with an energy consumption ratio of about 60W/T is about 50%. We can use the computing power growth slope of the past three months to estimate the total network computing power at the time of halving (the current total network computing power is 95.9E, the 180-day average computing power on December 10 is 82.06E, and the 180-day average computing power on September 10 is 59.48E). According to this estimate, the computing power at the time of halving is 82.06 x (82.06/59.48)*2=156E. However, a 50% replacement rate for mining machines does not equal the elimination rate of computing power. Taking the S9 (13T) and S17 (53T) mining machines among the new and old mining machines as examples, according to a 50% replacement rate, the elimination rate of computing power is 13 / (13+53) = 20%, which is 156E x 20% = 31.2E, equivalent to approximately 31.2 million T. So how much is 1 T? According to Canaan Creative's prospectus, the weighted average price of computing power for the whole year of 2017 and 2018 is (2114637 x 613 + 7158666 x 369) / (2114637 + 7158666) = 424 yuan/T computing power. That is to say, when the halving occurs, if the price remains unchanged, 424 x 31.2 million = 13.2 billion RMB worth of assets will be eliminated. Or we can try a different algorithm. BlockBeats learned that Canaan Creative's current market share is about 21%. According to the data disclosed in Canaan's IPO prospectus and the ideal situation of 100% utilization of the A7-A9 generation mining machines, the overall sales of Canaan's A7, A8, and A9 series mining machines in 2018-2019 were approximately RMB 2.922 billion, and the market size of mining machines with similar energy consumption ratios was RMB 13.914 billion. Considering mining machines as futures/rights that miners bet on the mid-term Bitcoin price, can we calculate how much money will theoretically go down the drain if the price remains sluggish? According to F2pool's electricity cost share data, Canaan Technology's A7, A8, and A9 series mining machines account for more than 50% of the total, which means that ideally, at the beginning of the next Bitcoin halving, the electricity cost share of these mining machines will be greater than 1, and the miners' floating income will become negative. Assuming that the mining machine sales market maintains a uniform growth rate before the halving, the market size of mining machines with similar energy consumption ratios in May 2020 is 139.14*([10 months + 5 half-years * 6 months]/10 months) = 17.4 billion yuan. At the same time, according to the above estimate of the computing power elimination rate, 174*(1-20%) = 13.9 billion yuan. Then, without considering the fixed costs of existing mining machines, the mining machine market will lose 13.9 billion yuan. This is almost the same as the result of our first method based on the computing power of the entire network. Regardless of the algorithm used, it is theoretically foreseeable that most mining machines, including the S9, a flagship machine that has survived a round of bull and bear markets, will become stones under the wheels of history in this round of halving. However, the actual situation is always different from the theory. BlockBeats interviewed a mining farm owner, who had different views on the relationship between Bitcoin and price. The electricity fee is not a unique value. The shutdown price is calculated based on 0.35 yuan. Mining is a global activity, and the electricity fee is not the same around the world. That is to say, even if the mining machine reaches the shutdown price after the halving, it is still possible to start up in another place. "Someone shuts down at 0.39 yuan, and I can take over his computing power because my electricity fee is less than 0.2 yuan." Will the computing power drop sharply? The rise and fall of computing power is closely related to whether the mining machine is turned on. The most obvious example is LTC, whose output has just been halved this year. BlockBeats has found that compared with the highest point of computing power before the halving this year, due to the price drop, LTC's total network computing power has dropped by 70%, and the mainstream LTC mining machines have been shut down. According to the conclusion we have just drawn, even if the price of Bitcoin remains unchanged, the mining machines will not be shut down due to electricity costs. Therefore, the Bitcoin computing power will not drop sharply after halving. However, the above-mentioned mine owner told us that there is another factor that must be considered, which is the production cycle of the mining machine manufacturer. If before the halving, new high-computing machines from mining machine manufacturers such as Bitmain or Canaan Creative are produced and the production capacity is sufficient, the overall computing power of the entire network will not drop sharply, but will only rise. In other words, if everyone switches to new machines, the computing power is unlikely to plummet. However, if new high-computing power machines are produced but the production capacity is insufficient, that is, only some high-computing power mining machines appear on the market, and most people are still using low-computing power mining machines, then the Bitcoin computing power may plummet. Will the 2020 halving mean a mining disaster? Recently, there is a popular view in the market that the computing power will continue to grow steadily on the existing basis. After the Bitcoin block revenue is halved in May 2020, if the price remains around US$7,200, it will be equivalent to the revenue level of US$3,600 before the halving. Since the computing power in May 2020 is 2 to 3 times higher than that in early 2019, a super "mining accident" will occur. At first glance, this logical deduction makes sense, but in fact it is untenable. We found evidence from historical data. Bitcoin has experienced two halvings, in November 2012 and July 2016. Since the Bitcoin market in 2012 was too small and immature, the reference value was limited, so we focus on the last halving. The above-mentioned "mine disaster prediction" has already occurred in the halving cycle in July 2016. In October 2015, the price of Bitcoin was close to the bottom of the cycle, at $240, and the total network computing power was 400P. In July 2016, when the halving actually occurred, the price rose to 680USD, but the computing power increased at a faster rate to 1.5E. Therefore, after deducting the reduction in revenue due to halving, when the halving occurred in July 2016, the revenue of each Bitcoin fiat block increased by about 42% from the bottom of the cycle, while the year-on-year growth of computing power was as high as 275%, which was much smaller than the computing power. In other words, the growth rate of people grabbing the cake is much higher than the proportion of the cake growth. And we know that in the bear market of 2015, the mining industry was also very difficult. Does that mean that the mining industry after the halving in July 2016 is more difficult than the bottom of the cycle in 2015? Obviously, this is not the case. The mining industry achieved rapid development in 2016 and 2017. Similarly, at the beginning of 2019, the price was around 3,600 USD and the mining industry was in a difficult situation with 45E hashrate. After the halving in May 2020, assuming the price remains at 7,200 USD, minus the reduced revenue due to the halving, the fiat revenue of a single BTC block will be equivalent to 3,600 USD, with a growth rate of 0. At this time, the estimated hashrate of the entire network is 150E, an increase of 233%. This does not mean that a larger-scale stampede will definitely occur in the mining industry. What is the reason behind this? The above deductions about block revenue and computing power are all at the macro level. In actual combat, the micro-decision-making of mining machine manufacturers, mine owners, retail miners and investors is much more complicated. Mining behavior is not a game of a time point, but a game of a medium-term period. For mining machine manufacturers, the price of each unit of T computing power is dynamic, each batch of orders has a certain schedule, and the value of computing power transmitted to the entire network is also changing. For powerful mining farm owners, when retail miners give up first, they will take over mining machines at a lower price, thereby expanding their share of computing power. In essence, this is a process of continuously reducing the marginal cost of the Bitcoin mining market. For investors, if institutions unanimously have bullish expectations for halving, and some miners also hold on to their coins because of bullish sentiment, this will lead to changes in supply and demand in the market, and the high price growth will offset the reduction in block fiat currency revenue caused by halving. Therefore, this kind of reasoning logic is not accurate. Judging from historical data, this logic cannot deduce that mining accidents will occur after halving. According to our calculations and interviews, when the halving occurs, if the production capacity and cycle of the mining machine manufacturers are normal, the Bitcoin computing power will most likely not fall on a large scale. Some mining machines that are theoretically going to be eliminated may also be moved to other places with lower electricity costs to continue mining. Therefore, mining accidents during halving may not occur.
