The Merge, Part II
In September, the Ethereum network successfully completed a highly technical major upgrade, known as the Merge, which changed how transactions are added to Ethereum's blockchain ledger. There are many reasons why investors ought to be excited about this transition. We provided an overview at the beginning of August, but these reasons are worth diving more deeply into.
The Merge has been on the Ethereum roadmap from the very beginnings of the project. Much negative press has been generated by Bitcoin's "wasteful" use of energy, and the Merge was designed to, amongst other purposes, alleviate Ethereum’s similarly energy intensive usage. To see how this is accomplished requires some basic understanding of how so-called "Nakamoto Consensus" based blockchains work. Here we provide the bare minimum explanation and please check out the references for more detail.
First, we ask the question, how does Bitcoin manage to maintain a ledger of transactions that the whole world can agree on and yet remain decentralized and censorship-proof? The answer to this question is that the correct ordering of transactions in the blockchain ledger is agreed on by network participants beforehand! How is this possible? It is possible because all network operators have agreed to accept the version of history presented to them (aka a blockchain) which can prove itself to have spent the most amount of electricity in its creation! This is known as the "Proof of Work" consensus mechanism. The pros of this approach are that it is simple and anybody in the world can burn electricity and add the next set of entries into the blockchain ledger. As a result, it is both censorship-proof and decentralized. The downside is that burning of electricity is a core feature of this technology, and in fact it could be argued that the more "wasteful" and non-otherwise productive the energy expenditure is - the better it is for maintaining censorship resistance and decentralization!
Soon after Bitcoin started to consume significant amounts of energy, people began to ask if there was another, less energy intensive, way to achieve the same results (decentralized, censorship-resistant consensus) without using so much energy. Afterall, if Bitcoin were to achieve its ambition of scaling to reach every individual on the planet, then continuing to scale up energy wastage is hardly sustainable. This is where "Proof of Stake" enters the picture. Where Proof of Work creates a lottery system where your chances of being "selected" to add the next block (and hence set of transactions) into the blockchain is probabilistically proportional to the amount of electricity you burn, Proof of Stake changes the mechanism such that the probability is now proportional to the number of coins you "stake" or lock into the network. This seemingly simple change resulted in a 99.95% drop in energy usage by Ethereum in September. Arguments against Proof of Stake include the threat of formation of oligopolies where the rich get richer equally apply to Proof of Work where only a small number of well-capitalized Bitcoin mining corporations can afford to purchase the latest and most powerful mining hardware, operate mining on a large scale, and negotiate with power generators for cheaper electricity rates. A Proof of Stake layer for Ethereum has been running in parallel to the portion of the Ethereum chain dedicated to the execution of smart contracts for the past two years. It is the merging of these two Ethereum layers where September’s network upgrade takes its name.
Back to the issue of what this means for investors. The first and most obvious impact is that Ethereum is now massively more ESG-friendly for institutional investors than Bitcoin, and this discrepancy can only grow as both networks scale in usage. Activist investors have been increasingly critical of Bitcoin’s energy usage and Tesla halted accepting bitcoin for payment citing such concerns. Another major change rolled into the Merge upgrade has been a huge reduction in new ether coming into existence as part of the reward given to stakers (previously miners), post-Merge. This represents an almost 90% reduction in the amount of new ether issued daily. Moreover, previously ether miners would have been required to sell the bulk of their mined ether in order to cover costs, including purchasing and upgrading expensive mining hardware and buying a huge amount of electricity. Those two costs are substantially smaller for stakers so the reduction in sell pressure is significant. This reduction of issuance, combined with some earlier Ethereum upgrades related to the “burning” of part of the ether used to pay transaction fees has resulted in ether effectively becoming a deflationary asset at times of heavy network usage.
Fig 1. Comparison between total ether created since The Merge in September, ~6,000 eth versus 400,000 eth had The Merge never taken place. Actual bitcoin issuance is also shown.
Fig 2. The cumulative amount of ether in existence and the change of ether issuance per block over time. Theoretically post merge the total amount of ether existing may decline over time, depending upon network usage.
Although it could be argued that this sudden reduction of issuance has already been priced in, previous “halving” (approximately every four years the issuance rate of new bitcoin is halved) events in bitcoin have been argued to precede significant price appreciation. This is despite being almost as predictable as the change of the seasons, although the number of such events is too small to be able to draw any firm conclusions. If this holds true for Ethereum, then the issuance reduction post-Merge effectively represents a “triple-halving” event foreshadowing possible price appreciation. A further argument against The Merge being priced in is that the price of ether denominated in bitcoin has barely changed on a yearly basis. Assuming that the relative demand for ether and bitcoin has remained constant, yet Ethereum has seen a massive reduction in supply and overcome a very significant technical uncertainty, then it is hard to believe that the market has given full credit to Ethereum at this point in time.
Fig 3. Ether priced in USD and in bitcoin.
Although the migration to Proof of Stake has, for the time being, seemingly been mostly positive, a few problems have since arisen. The most potentially damaging change to the Ethereum network post-Merge is the breaking of censorship resistance due to stakers running code to optionally accept payments for the reordering of block transactions to the benefit of payer. This reward from block reordering is known as MEV (Maximal Extracted Value).
MEV had previously existed under Proof of Work and was previously known as Miner Extractable Value, as miners controlled the block formation, including transaction ordering. (Interested readers are advised to visit the ethereum.org’s MEV page for fuller details.) This block reordering is by itself not especially problematic; it boosts the stakers’ rewards by allowing interested parties to profitably front run transactions, disadvantaging some network users but not affecting the Ethereum network as a whole.
The problem is that many of these reordering services offer the ability for stakers to censor transactions associated with addresses sanctioned by the US Treasury Dept, such as the anonymity service Tornado Cash. This raises concerns that this is the start of a slippery slope of expecting stakers to police transactions out of fear of being branded as complicit in facilitating illicit activities. Although it's not clear that stakers are required to censor transactions, currently > 50% of stakers are doing so, probably out of an abundance of caution. Although there are some legitimate points raised in this discussion, much of the reporting from crypto news sources and influencers has been biased, hysterical or overblown. Currently, the migration to Proof of Stake has not meaningfully reduced censorship resistance on Ethereum.
Another concern centers around large staking pools. There is still some debate over whether Ethereum is now more decentralized under Proof of Stake rather than Proof of Work, given that major staking pools (individuals effectively pledge their stake to be pooled with other stakers, thereby increasing the probability of being selected to create a block and reducing earnings variance) and exchanges represent the majority of stakers. However, should staking pools or exchanges behave in a manner antithetical to the values of small stakers, these stakers are free to withdraw their stakes from these large pools. In contrast, over the years, as ether mining became a more industrial scale enterprise, the viability of an individual mining at home, whilst ideal for decentralization, rapidly declined. Following migration to Proof of Stake it is currently still viable for an individual to stake at home albeit with a fairly high-spec machine and the requirement of 32 eth to pledge as a stake.
To put all this theory into practice, at Firinne we constructed our own machine specifically to gain a deeper understanding of staking. The machine spec requirements are not cheap and will grow over time with the network. A typical set up would look like the following
• Fast CPU with > 4 cores and cooling
• 32 GB of RAM
• Fast and reliable internet connection
• 2 Tb SSD
The Ethereum blockchain is already at 1Tb in size and growing at 1Gb per day. All of this setup will certainly cost more than $1000 and is expected to depreciate rapidly due to constant usage and the ever-growing size of the Ethereum chain. The processor and hard disk will be in constant use over their entire lifetimes. All of this must be factored into calculations of profitability for stakers.
Fig 4. Custom built staking machine. A reliable power supply, internet connection and cooling device are important.
Finally, this brings us to the question of what yield can be expected for staking on Ethereum. The staking yield can be broken down into several sources. The first and usually largest component is the reward paid for performing validator duties – this includes the reward for successfully proposing a new block and for attesting to the truth of other blocks proposed to the network. The size of this reward depends on the proportion of ether staked on the network; it is currently about 4% per annum. The next component comes from network participants paying to have their transaction prioritized in being added to a block. Finally, MEV as mentioned above, where the specific ordering of transactions within a block may be rearranged to the benefit of the payer, comes into play. When all of these rewards are added together the Ethereum network is currently paying stakers close to 4.7% per annum.
Fig 5. Ethereum staking rewards. Source: Figment
All told, The Merge has been an enormous success for Ethereum, and one which its many critics foretold would never come to pass. True, it has taken far longer than many people initially forecast, however the complexity and risks associated with such a significant network upgrade should not be understated. And in addition to the important energy saving measures brough in by The Merge, the entire cryptocurrency space now has the beginnings of something starting to resemble a “risk-free” rate in the Ethereum staker’s network yield.