Wayward Block Times and Their Impact on Bitcoin Mining
For those unfamiliar with the business of Bitcoin mining, one of the benefits of mining is a predictable revenue stream — especially for miners who store their wealth in Bitcoin. There are only three variables which determine mining revenue and the first is the percentage of the world’s hashing power (also known as computing power or processing power) by the miner. Over time, miners receive a share of the global mining revenue equivalent to their share percentage of the world’s hashing power.
The price tag for the service provided by miners is called the Block Reward and it is the second variable in determining mining revenue. The Block Reward is itself the sum of two variables: the Block Subsidy and Transaction Fees. The Block Subsidy is new Bitcoin being released into circulation and it is given to the miner that processed (won) the block. Presently the rate of dispersal is ₿6.25 per block, and it will remain fixed at this rate until block 840,000 at which point it will drop to ₿3.125. (Block height at the time of this writing is 691,153.) Transaction fees vary from block to block in a range of about ₿0.05 to ₿0.75. For purposes of simplicity, the assumption will be that that typical transaction fees are ₿0.25; therefore, the typical Block Reward is ₿6.50.
The third variable is the number of blocks per day and exploring this in detail is the main theme of this article. Part of Satoshi’s genius in developing Bitcoin is a difficulty adjustment that examines the average amount of time required to process the previous 2016 blocks. Each of these periods is called a difficulty adjustment epoch, or block processing epoch. At the end of each epoch the Bitcoin protocol increases/decreases the difficulty of processing a block in hopes that the average block processing time in the upcoming epoch will be as close to 10 minutes as possible. So generally, anyone looking to estimate the number of blocks in a day will simply take the number of minutes in a day, 1440, and divide by 10 minutes. The result is 144 blocks per day and dividing that by 2016 blocks per epoch results in an average block processing epoch of 14 days, or two weeks.
This means that a simple projection of daily global mining revenue can be made by taking the typical block reward of ₿6.50 and multiplying by 144 blocks per day for a total ₿936 per day. This then extrapolates to annual global mining revenue of ₿341,640. Many find it easier to think in terms of $USD, so at $35,000 = ₿1 the daily revenue is about $33M and annual revenue is $12B.
Looking deeper though, 144 blocks per day is not a constant. As block processing time varies from the 10-minute goal so do the number of blocks that get generated each day. Table A below illustrates that for most Bitcoin’s lifetime the block processing time has been faster than 10 minutes.
This is logical because for most of Bitcoin’s history there have been continual additions to the world’s overall hashing rate and since the algorithm is always looking into the past, it has typically underestimated the processing power available in the upcoming epoch. As Table A shows, on average (all-time) block processing times have been running about 30 seconds, or 5%, faster than the goal. On the surface that may not appear material but it really is — on average, 7 more blocks per day than the target have been processed throughout Bitcoin’s history. Over the past 18 months though, the network has stabilized dramatically, and the average block processing time has been remarkably close to 10 minutes.
Even before China’s ban on Bitcoin mining in mid-June, there had been some flight of hashing power from the network starting in early May. This may have been because of the softening in Bitcoin’s price and/or because some Chinese miners received an early warning of the upcoming ban. Regardless, block processing times have been well above average for the last three epochs, and the current epoch is trending toward about eleven minutes. The great mining
exodus from China appears to be almost over though, and when it does it will almost certainly mark a bottom in global hash rate. Precise real-time measurements of global hash rate are very difficult to make, but a reasonable guess of this low point is about 90 EH/sec, and it establishes a very important infliction point for the Bitcoin mining industry.
This drop in Bitcoin mining’s global hash rate has made almost all remaining ASIC based mining equipment profitable — especially those built after 2017. Typically, 4 or 5-year-old systems would be the “buffer” units that are turned on or off as Bitcoin’s price moved up or down. These would be the units at the bottom of quadrants 1 and 2 in Chart A below, and the ones that played a key role in the high historical correlation of Bitcoin’s price and hash rate. This relationship no longer exists though. (For more on this please see “The Unforeseen and Tragic Breakup of Bitcoin’s Price and Hash Rate” available at https://bit.ly/3heWhxy.) Instead, we are about to enter an extended era where hash rate will go up continually regardless of price. This is because mining systems, even old mining systems, will stay active even as new systems will be come online because they remain profitable. One could say that the Chinese government has given old systems sitting outside of China a drink from the fountain of youth.
Since for the foreseeable future hash rate will be continually increasing, the difficulty will always be under-adjusting for the upcoming epoch and there are some important ramifications of this.
· Block processing times will consistently run under ten minutes.
· The timing of the next halving is being sped up
· Significantly more than 144 blocks will be processed per day.
· Miners will see a material increase in revenue flows.
· Bitcoin’s money supply will inflate faster than expected.
Forecasting the specifics of these numbers is difficult but a reasonable approximation is possible. Assuming the exodus from China is complete by epoch 345 (beginning on approximately August 1, 2021), then global hash rate will climb steadily from there.
Table C estimates the complexion of the mining units that were in use in China at the time of the ban. As described in my earlier articles, China’s mining industry faces a material challenge redeploying these systems as energy sources and facilities are not readily available. Given a limited ability to expand, these companies will certainly bring the newest machines online first as they are the most efficient and most profitable. Because the redeployment will be slow and will occur youngest to oldest, and because there is a significant cost of redeploying the systems, the units from the class of 2017 and 2018 (18.1% of the mix) are likely to never return to operation. This is about 830K systems that will be retired, but the remaining 81.9% back still represent over 1.1M units that need to be decommissioned, then shipped from China, and then reinstalled in a new facility. Getting these systems back in action will not happen overnight.
Table D estimates the rate at which the 1.1M units will come back online and the resulting increase in hash rate. Also, an examination of the historical expansion of Bitcoin hash rate (since 2018) shows that it has grown an average of 2.41% (compounded) per month. When this growth rate is applied to the current non-China based hash rate it shows an additional 81 EH/Sec being added to the world’s mining capacity over epochs 345 to 371. Combining these two results in hash rate growing 3.8% compounded per epoch over this 27-epoch span.
Looking even deeper, Table E shows that beginning with that epoch 345 (about August 1, 2021) block processing times will begin running about 26 seconds faster than the expected 10-minute block time. This will accelerate the epochs to as fast as 13 days and 8 hours, or 16 hours ahead of plan.
This also means that as many as 7.4 extra blocks per day will be processed and, in some epochs, this will result in global mining revenue increases as high as 5% with an average of 4% over expectations. Since miners have essentially fixed operating expenses, this extra revenue will float directly to the bottom-line. In total, this will increase mining industry revenues (and profits) by $469M (₿1 = $35K) to as much as $1.3B (₿1 = $100K). It should be obvious that anyone with an operating mining business is going to have a very good year ahead.
As mentioned earlier, it is anticipated that epoch 345 will begin on August 1, 2021, and Table E projects that epoch 371 will be complete about July 30, 2022 (363 days). A normal 27-epoch span would take 54 weeks (378 days) to complete but the China mining ban has the impact of beating this by over 15 days. An interesting observation is that this also pulls forward the projected timing of the halving to late April 2024 from May 14, 2024 (according to BitcoinBlockHalf.com).
Another consequence of the China Mining ban is that the supply of Bitcoin will be increasing at a faster than normal rate. Technically, this means the rate of inflation for Bitcoin will be slightly higher than projections over the next year. Ultimately, the 21M supply cap still applies so this is of relatively miniscule impact, but since the Bitcoin community is traditionally ultra-focused on any change in money supply, be it fiat, Bitcoin or an alt coin, awareness of this increase is noteworthy.
Finally, if the mining community were to get in a “sell mode” at some point during epoch span 345 to 371, the market would have to absorb a higher supply of Bitcoin, and this would have downward price pressure; however, this is unlikely. The accelerated block processing speeds are not associated with any additional production/operational costs. So, the ₿13,426 (see Table E) flowing into the money supply are pure profit for miners and, most miners, especially those in North America, are in the business because they wish to accumulate and hodl Bitcoin. Selling Bitcoin to buy fiat is not well aligned with their philosophy.
While the first order effect of China’s mining ban was clearly disastrous for China’s mining community, the second and third order effects are overwhelming positive for Bitcoin’s ecosystem and the remaining Bitcoin mining industry. Bitcoin proves once again that it is just like a starfish, chop off an arm and it’ll “just keep on, keeping on,” while it regrows its arm.
Tick Tock Next Block.
Bob Burnett
CEO, Barefoot Mining
Block height 691,153 (or as some might say — July 15, 2021)
