Before we start allow me to clarify that these calculations do not apply to all pools, but to those that keep the miner transaction fee rewards for themselves opposed to sharing them among the miners using their pool. AntPool (BitMain), BW and F2Pool are the largest pools that do not share the reward. These three pools represent around 30% of the hashing power of the network. Pools that do share the transaction fee rewards can be found in this comparison of mining pools table. If you’re a miner it may be more profitable to use a pool that does share the reward.
People are often seen arguing about whether it is economical for Bitmain (Jihan Wu) to block and oppose a scaling solution because it is claimed that they stand to gain more from a Bitcoin price increase, so let us try and put this debate to rest. A lot of these individuals seem to be unaware of the reward structure of these mining pools, which needs to be factored into the calculations. To address this conflict we need to imply the following:
The pool owners only own a fraction of the hashing power of their own pool. In other words, only a fraction of it is their own hardware, the rest is the hardware of other Bitcoin users.
The blocks remain full and people are competing to have their transactions included in the blocks, thus raising the price of transaction fees and forming an inflated fee market.
All miner transaction fee rewards go directly to the pool and are not distributed to the miners using the pool.
Crunch the numbers
Let’s dive in. We’ll make the case for Jihan Wu’s pool here (AntPool), since it is the largest Bitcoin pool that does not share the transaction fee reward. AntPool currently owns 15.8% of the hashing power of the network. If we calculate the average daily USD total for the last 10 days of transaction fee rewards we come to the number $947,000. Given that AntPool controls 15.8% of the hash power of the network, if this inflated fee market was to persist this means that generally speaking AntPool is earning 15.8% of $947,000 in transaction fee rewards per day. That’s a lofty $149,000 per day, 4.47 million a month, and 53.64 million per year of revenue.
The fees earned through Bitmain’s own mining equipment out of this 15.8% will incur their own expenses because mining is an expensive operation, so let’s adjust the numbers to compensate. Let’s presume Bitmain directly owns 7.9% (half) of the mining hashing power of their pool, a seemingly reasonable figure. If we deduct 50% from the fee rewards we get the reward that AntPool would be earning if they had no mining power to contribute to the network themselves, an almost pure profit figure. This works out to be $74,500 per day, 2.23 million a month, and 26.82 million per year of profit (excluding pool management fees which are negligible in comparison). To recap, these are the earnings Bitmain would be making through the management of their pool alone, without mining Bitcoin themselves. Impressive figures.
Now let’s calculate how much money Bitmain is earning through their own personal mining equipment so that we can make a comparison and determine whether fees are more lucrative or block rewards are more lucrative. The current hash rate of the network is ~4800PH. This would put Bitmain in direct ownership of 380PH of the network. With the block reward currently at 12.5btc and a block being mined roughly every 10 minutes that works out to be 648,000 btc mined per year. Bitmain is thus entitled to 7.9% of that, coming to 51,000 btc by years end. At a Bitcoin rate of $2000 that brings the yearly profit to 102 million.
We now need to deduct expenses. The Antminer S9 is capable of around 13.5TH/s, so that 380PH works out to be around 28,150 S9’s. The S9 retail price is $1200, and if we presume a markup of 100% that brings the raw cost of the miners to about 16 million. We subtract that cost from the yearly profit, bringing it to 86 million. Next we have to factor in electricity costs. At 1375W of power consumption, 1375 multiplied by the number of miners gives us 38.7 MW of consumption, which works out to be around 13.5 million USD per year at 4c kWh. Subtract that from the profit total, we’re left with 72.5 million. Next we factor in data center costs which include staff management, routers, cooling, racks, and so forth. Roughly estimating that at 10 million brings the profit total to 62.5 million. This total is without ASICBoost which would actually increase the power consumption efficiency by around 20%, bringing the profit total to 65.7 million, but we’ll ignore that for now for simplicity sake.
So now we have our two figures. What do they tell us? 26.82 million per year profit if Bitmain was not mining using their own equipment and their business model was purely transaction fees, and 62.5 million per year profit if it is presumed Bitmain directly own 50% of their pool hash power. If scaling was addressed and the fee market was to burst we could assume fees to fall back to their historical level of around $40,000 USD per day. That works out to be a 2.27 million per year reward in contrast to 26.82 million per year. That means that Bitmain would need to recoup 24.55 million per year in profit through mining to make up for the collapse of the fee market. That would mean a ~28% increase in the price of Bitcoin for the year. The question then becomes whether we believe deploying a scaling solution to Bitcoin would allow it to gain a yearly 28% increase in price. My personal opinion is that this is a very reasonable figure to reach for and the resolution of the scaling debate would far exceed this mark.
One thing worth considering is that the above does not take into account the fact that large pools stand to gain substantial profits by resisting a scaling solution and accumulating cheaper Bitcoin, knowing that the value will soar once a scaling solution is deployed. If you have the power to artificially suppress the price and release that suppression at your leisure then you stand to make a sizeable fortune when the flood gates open.
Conclusion
By taking advantage of the transaction fee reward structure of these aforementioned mining pools they stand to gain a considerable profit. In the case of Bitmain, this works out to be around 26.8 million USD per year in profit through transaction fee rewards (excluding their own miners) at the current Bitcoin price and level of network congestion. This is in contrast to a 62.5 million USD per year profit from their directly controlled miners alone, through block subsidies.
As we can see, there is definitely incentive for these mining pools to maintain fees at the highest rate possible, but would it be more profitable for them to allow a scaling solution? The numbers seem to indicate so, considering that a 28% yearly increase in price would compensate for all lost fee rewards. The Bitcoin market usually pulls far more impressive numbers than that.
This then begs the question of motive, why is Bitmain so resistant to the proposed scaling solutions? As mentioned above, artificially suppressing the price allows an accumulation phase that will net immense returns once this suppression ceases. It’s also fair to say that reducing transaction fee rewards to a tenth of their current value plays into it as well. After looking at all of these figures, one thing is for certain, this is a highly lucrative business with seemingly plenty of room for competitors to enter the market.
Follow me on Medium for more stories like this: https://medium.com/@patrickob