This post was guest-authored by Crypeli and published here with permission.
As some of you may know, Ether mining’s profits are falling due to several factors such as:
• Decreasing market price;
• Increased mining difficulty;
• (in some cases) increased energy prices.
In turn, mining difficulty is increasing due to more hashrate (i.e. mining capacity) being inserted into Ethereum’s blockchain. Right now, Ethereum mining reward’s cake is fixed at 3 ETH per block, which means the one who finds the correct nonce value for a block gets those 3 ETH. This implies the following:
Formula 1: Probability of earning the block reward given a fixed hashrate X.
Thus, if players such as Bitmain enter into the arena (as they did), retail miners will find it harder to keep up with energy costs, as their probability of earning block reward gets dilluted. This in itself would not pose a problem for Ethereum’s blockchain, as it provides more hashrate to the network and thus more transaction throughput. Except it does: 51% attacks.
A 51% attack is a kind of cryptographic manouvre performed into a blockchain such as Bitcoin and Ethereum that allows such performer to double-spend the tokens of the blockchain in question; in order to perform it, it is necessary to have more than half of a blockchain’s hashrate (see this for more info). Now, as more retail miners are propelled out of business, there are less and less miners, and bigger miners increase their hashrate’s share, making it easier for them to potentially collude to perform the so-called 51% attack.
In order to tackle this, the ETH team plans and strive to transition from PoW (Proof-of-Work, i.e. current mining scheme) to PoS (Proof-of-Stake, i.e. nodes process transactions and need a confiscable deposit -stake- to operate). The problem is, the transition is not done yet, and no hard deadline seems plausible in the short term.
Thus, they most likely need to do something in the meantime. Why? Because of the following:
Figure 1: Breakeven energy price and ETH global hashrate.
As it can be appreciated in Figure 1, hashrate has been increasing steadily, even though the baseline has not kept pace! How come can this happen? Let’s analyze how the breakeven price is calculated.
Formula 2: Breakeven price per KWh.
Your hashrate and power draw (i.e. the power your mining rig(s) consume(s)) are fixed, as they are attached to your mining operation’s technology and size. The total issued ETH per day is fairly constant, as it is calculated based on the block reward (currently 3 ETH) and average time between blocks (currently about 14-15 seconds), both of which are given by the structure Ethereum blockchain. The remaining variation possibilities lie on the USD-ETH exchange rate and the total ETH’s blockchain hashrate. What happens when we analyze them both together?
Figure 2: ETH price and ETH global hashrate.
As ETH price is decreasing and global hashrate is increasing, and as per Formula 2, it is quite obvious that the breakeven point is getting tighter and tighter as time passes by. But, wait… why is hashrate getting higher if such an important driver such as price is against miners? The answer is PROFESSIONALIZATION.
As Vitalik Buterin (a.k.a. ETH’s leader) said: “Unfortunately, most mining is now professionalized, even with GPUs, so it’s not clear that mining increases decentralization of ETH holdings…” Simply put, if there’s more money on the table, big actors will increase their capacity so as to absorb that money, most likely in a way that is faster than that of the retail miners (thanks to economies of scale and increased mining expertise, among other factors). This leads to big miners increasing their share and… again, we all know what happens when big actors get relatively bigger: increased likelihood of a 51% attack.
Regarding professionalization, the biggest single actor in the mining scene is Bitmain. With its ASICs expertise and fabrication capabilities, it is undoubtedly the authority in ASIC mining. Now, as per what Riccardo Spagni, Monero’s lead developer said, “It takes AT LEAST 5 months to design, tape out and take delivery of ASICS…”. Thus, if prices are high, it is expected that they start their ASIC inception cycle and the hashrate to significantly increase in about 5-7 months if ETH prices keep high enough. What if we find a spot when global hashrate started to spike, go 6 months before and see when it falls into? Well, the steepest hashrate spike the ETH blockain had began on mid-december 2017 (when Bitcoin had its all-time high), and the mid-may high (where ETH started to become more popular) was 6 months before that. This may suggest that some sort of actor that is able to 1) Develop mining capabilities within a 6 month period or so, 2) Ramp up mining hashrate quickly. The most likely known actor that can do this is Bitmain:
Figure 3. ETH price and ETH global hashrate (with ASIC milestones).
It is clear that some sort of big actor such as Bitmain entered into the scene. Now, as it has been mentioned before, that is a problem for ETH Blockchain’s integrity, even more when the baseline is getting tighter every day. The question is: how big of a problem it is? No one really knows for dead sure, but some approximations can be done.
As of today, the proposals (Ethereum Improvement Proposals or EIP’s, in ETH’s jargon) for mitigating the aforementioned economic and security issues while PoS is implemented are as follows:
a) EIP-858 – Reduce block reward from 3 ETH to 1 ETH per block (a.k.a 66% less mining revenues and asymptotically decreasing to nil);
b) EIP-1227 – Delay bomb and change rewards to 5 ETH (i.e. increase mining revenues by 66%);
c) EIP-1234 – Delay bomb and change rewards to 2 ETH (this is reducing mining proceeds by 33%).
Basically, it’s a question of 1) Direction and 2) Magnitude of block reward change. As for direction, we can go upwards or downwards. Right now, worldwide profitability overview looks like this (please note that the energy rates used here are most likely residential ones, i.e. what what home miners would get; most of the times, most of the times, professionalized miners get better rates. For more details: https://www.globalenergyprices.com/en/data-collection-method/):
Figure 4. Current worldwide ETH mining profitability overview.
Right now, Western Europe, Australia, New Zealand, Japan and some parts of America are in the red zone. Now, what would happen if we go downwards like in EIP-858? This:
Figure 5. Worldwide ETH mining profitability overview with EIP 858 on board.
As it can be seen, All of America, China, India and other large regions will add up to the red list. This suggests that reducing rewards significantly increases the odds for a 51% attack. What if we go upwards? Let’s see what will happen with EIP-1227:
Figure 6. Worldwide ETH mining profitability overview with EIP 1227 on board.
Everyone will be theoretically happy, as almost all countries will be in the green zone. Nevertheless, and as it has been mentioned before, when more money is on the table, big actors begin to ramp their capabilities up faster than small ones and thus outpower them. We all know what happens when big actors/groups of interest increase their power.
Other than the Ethereum Foundation, there are 4 main groups of interest:
• Retail miners: these are non-professional, mostly amateur actors, who don’t have access to economies of scale. Their main interest is to survive professionalized miners’ consequences and, most of the times, try to recover their investment. They provide the highest benefit to ETH network’s security, as the hashing power these provide is dilluted across a lot of entities. Their drawback is their weakness to decreasing breakeven energy prices;
• Professionalized miners: these actors have access to specialized equipment, facilities and enough financial resources to survive decreasing breakeven energy prices, with aims to maximize their profits. These provide a lot of benefit to the ETH network, as they have a lot of hashrate power, which is highly concentrated, though, and bear increased 51% attack risks;
• Investors: these are entities that tend to hold large amounts of Ether, and are worried about the value of their holdings (and about ETH’s inflation as a consequence). They currently add no value to the ETH network’s security, but can potentially move the market in potentially unpredictable directions;
• ETH security officers: these are the ones who have the ETH network’s security in their top-of-mind.
Also, as per previously mentioned arguments, the main courses of action are the following:
1. Increase ETH rewards;
2. Decrease ETH rewards;
3. Implement an ASIC-resistant mining algorithm;
4. 3 and 2, simultaneously;
The interaction between the aforementioned actors and courses of action can be outlined in the following table:
Table 1. Benefit table (an “x” means that the corresponding group of interest will be benefited by the corresponding course of action).
• Increasing ETH rewards generates more inflation and puts more money on the table, which are detrimental to investors (holdings value dilution) and retail miners (more incentives to professionalized miners to ramp up their capabilities), respectively. It also increases the likelihood of a 51% attack, as lesss retail miners are kicked out by breakeven energy prices, thus decreasing security;
• Decreasing ETH rewards benefits both investors (decreased holdings value dilution) and professionalized miners (less retail miners that can withstand tighter breakeven energy costs), thus increasing the likelihood of a 51% attack and decreasing security;
• Implementing an ASIC-resistant mining algorithm will definitely phase out a significant portion of professionalized mining, benefiting retail miners and thus decreasing the likelihood of a 51% attack and thus increasing security. This alternative is neither benefitial nor detrimental to investors;
• 3 and 2 simultaneously implemented will benefit retail miners, ETH security officers and investors, as retail miniers will have more future participation in the mining rewards, investors will have a lower rate of holding value dilution and ETH security officers will have a more segregated mining power distribution;
• 3 and 1 will benefit retail miners and ETH security officers, as professionalized miners will be significantly phased out (increased security is obtained as a consequence) and investors will see a higher rate of holding value dilution;
• 1 and 2 are mutually contradictory, so it is not considered here.
Conclusion: ETH’s security issues won’t get (significantly) solved by increasing or decreasing inflation, i.e. tweaking monetary policies. Such tweaking must me accompanied of a dilution (or elimination) of professionalized miners’ power by a) Finding a way to isolatedly incentivate small miners to enter into the game, b) Changing the security algorithm to PoW (or other security algorithm).
Special thanks to Sergio Tapia for his comments and help, as well as to the OMNI Analytics team. Without them, this would not be possible.