The Merge

Written: September 15, 2022

Ethereum’s transition to proof-of-stake consensus, called the “Merge,” was successfully completed on September 15 at 2:43 a.m. ET. Ethereum is now a proof-of-stake (PoS) blockchain. The below is a consolidated overview of the background and potential impact of this event.

How does the proof-of-stake (PoS) mechanism work?

A blockchain network is a decentralized network of computers called nodes; each node processes and records all the transactions that take place on the network. The transaction record takes the form of a chain of blocks called a blockchain. Consensus mechanisms allow blockchain networks to decide on new blocks to add to the blockchain. In proof-of-work (PoW) consensus, “miners” expend computing power for the chance to propose new blocks and earn associated rewards. In PoS consensus, “validators” deposit cryptocurrency as stake in the network for the chance to propose new blocks and earn associated rewards.

A minimum deposit of 32 ETH is required to become a validator on Ethereum. There are currently more than 400,000 validators active on the network (generally, each instance of 32 ETH is a separate validator).¹ Validators are rewarded with newly minted ETH for participating in network consensus. The amount of validator rewards depends on the proportion of total ether staked in the network. While block production timing in PoW is variable, block production in PoS is scheduled to occur at a fixed rhythm. In Ethereum’s PoS system, time is divided into slots (12 seconds) and epochs (32 slots). One validator is randomly selected to be a block proposer in each slot, and a committee of validators is randomly chosen to attest to the validity of the block being proposed. Rewards for proposing a new block are the largest, while attestation involves smaller (but more frequent) rewards.² A transaction has “finality” when it is part of a block that cannot change without a significant amount of ether getting slashed. Finality is reached after two epochs are confirmed by at least two-thirds of active validators.

Validators who behave dishonestly (e.g., proposing multiple blocks, submitting contradictory votes) can have their deposit slashed, meaning that they lose some or all of their staked ether. The amount of ether slashed depends on the number of validators being slashed around the same time, in what is known as the correlation penalty. Sabotaging the network via something called a “51% attack,” which would allow the saboteur to double spend their tokens and violate the integrity of the blockchain, would require the accumulation of more than half of all staked ether.

Currently, about 14 million ETH have been staked, representing around 11% of total ETH supply.³ It is important to note that staked ether and associated validator rewards cannot be withdrawn until after an upgrade following the Merge, called the Shanghai upgrade (expected to occur six to 12 months following the Merge). Even after the Shanghai upgrade, staked ether may not be very liquid. Withdrawal times will depend on how many other people are actively trying to withdraw. Only six validators may exit per epoch (every 6.4 minutes, so 1,350 validators, or ~43,200 ETH per day), with this limit decreasing to as low as four as more validators withdraw, to prevent large amounts of staked ETH from leaving at once.⁴ Thus, there may be an exit queue that may delay withdrawals for days, weeks or months.

How did the transition process work?

The Merge is so named because it merges the Beacon Chain, Ethereum’s PoS consensus layer introduced in 2020, with the Ethereum mainnet (execution layer). Correspondingly, validators are required to run two software clients, the execution client and the consensus client. The execution client handles transaction processing and state storage, while the consensus client manages the PoS consensus activities.

On September 6, Ethereum activated Bellatrix, a network upgrade to Ethereum’s consensus layer.⁵ This upgrade prepared the consensus layer to accommodate blocks containing transactions initiated on the execution layer, and to start filling up blocks with execution layer transactions once the execution layer reached a specific total terminal difficulty (TTD). The TTD was a measure of the accumulated mining difficulty on Ethereum and was the driver of why the exact timing of the Merge was so hard to predict, as it was dependent on the network hashrate (a measure of mining computing power on the network). Unlike Bitcoin’s two-week adjustment period, Ethereum’s mining difficulty was adjusted each block.

The Merge completed with the Paris upgrade on the execution layer, which officially activated once the TTD threshold was reached. Paris deprecated the PoW mining algorithm and introduced new dependencies on the Beacon Chain. The consensus and execution layers were thus merged as the Beacon Chain started producing blocks containing execution layer transactions. The completion of the Merge came after a long process of development and testing done on test networks, and marked the first transition of this type on such a large scale.

What are the goals and risks of this transition?

PoW mining consumes large amounts of energy. The Bitcoin network, which uses PoW, is often criticized for its environmental impact. The transition to PoS consensus is expected to decrease Ethereum’s energy consumption by 99.9%.⁶ In addition, the transition to PoS helps prepare the Ethereum network for future upgrades related to scalability (making transactions faster and cheaper) and decentralization, both important areas of development for Ethereum. Ethereum’s co-founder Vitalik Buterin has said that Ethereum’s development is only 55% complete after the Merge.⁷

A common misconception is that transaction fees will go down and transaction speeds will go up significantly after the Merge. Ethereum transaction fees are not expected to change as a result of the Merge. Transaction speed will go up on average, but only by a small amount. Ethereum blocks were issued once every 13 to 14 seconds, on average, under the PoW system.⁸ Since the Merge, PoS blocks are issued at regular 12-second intervals.⁹ “Layer 2” solutions can significantly increase effective transaction throughput on Ethereum and will continue to play an important role in Ethereum’s development going forward.

PoS is less battle-tested in terms of security compared with PoW. PoS design is also more complex to implement and may introduce new vulnerabilities that could undermine the security of the network. Since in PoS, ether (as opposed to computing power) is the resource that determines influence over network consensus, another potential concern is centralization risks related to the concentration of ether ownership, as well as the concentration of staking activity among a few staking service providers. As noted, 32 ETH (~$50,000 at the current ETH price) is required to become a validator, which may be prohibitively expensive for some people. In addition, running a validator node requires sufficient hardware and connectivity, as well as a certain degree of technical expertise. Thus, many people may choose to stake via third-party staking solutions that do not require users to run their own node (e.g., staking-as-a-service) or stake the full 32 ETH (e.g., pooled staking). Many pooled staking solutions also offer “liquid staking”, which involves the issuance of tokens representing staked ETH that users can then use elsewhere. Major liquid staking solution Lido currently accounts for about a third of total ether deposited as stake.¹⁰ Using third-party solutions can introduce risks that operators act in a malicious way or become subject to strict regulation.

What would be the impact of a chain split (a.k.a. a “fork”)?

Unlike Bitcoin’s mining algorithm, SHA-256, the mining algorithm that was used by Ethereum prior to the Merge, ethash, was designed to be more ASIC-resistant and amenable to GPU mining. ASICs are integrated circuits that are created to perform a specific computing task. After the Merge, Ethereum’s miners have the option of joining other existing (but less lucrative) PoW networks such as Ethereum Classic, selling their hardware or joining a potential new PoW fork of Ethereum.

Upon a chain split, the forked chain would proceed from a specific state of the original chain, and each address would have its holdings “duplicated”, in the sense of having tokens on both the forked chain and the main chain.

The long-term success of such a new PoW fork would depend on the extent of community support. Tether and Circle, the two issuers behind the top two stablecoins by market cap, USDT and USDC, have indicated that they will not support a PoW version of Ethereum post-Merge,¹¹ which may significantly hinder the DeFi ecosystem on the alternative chain, given the importance of stablecoins in DeFi. More broadly, the PoS transition has widespread support from the Ethereum community due to its importance to Ethereum’s road map. Therefore, although an alternative chain may emerge, its adoption may be limited compared with the main Ethereum network, and the tokens on the alternative chain, including the alternative version of ether, may accordingly trade at a significant discount to their counterparts on the main chain.

Ethereum Classic is an existing example of an alternative version of Ethereum that never gained as much adoption as Ethereum. Ethereum Classic was created in 2016 due to disagreement among the community following the hack of a major application on Ethereum at the time, called The DAO. To date, the market cap of ETC, the native token of Ethereum Classic, has only ever been a small fraction of the market cap of ether. There’s also very limited DeFi activity on Ethereum Classic.

What will be the impact on the supply and price of ether?

Since the end of the second quarter, ether’s price has increased by over 50%, largely driven by positive sentiment regarding the Merge. After the Merge, the amount of new ether issued per day is expected to decrease by ~90%.¹² This is because rewards for PoS validators are much lower than the rewards that were issued to PoW miners, since operating a validator node is not as economically intense as being a miner. The decrease in the issuance rate is expected to bring the net issuance rate to zero or less. The exact monetary inflation rate in the future is uncertain, because there is variability both in terms of the amount of new ether created through validator rewards and in the amount of ether burned (destroyed) as transaction fees per EIP-1559.

Ether is expected to go from being an inflationary coin to a disinflationary or deflationary coin, which may have a positive impact on ether’s price. The fact that people can stake ether and earn yield on their stake may also have a positive impact on the price of ether. More broadly, the Merge marks Ethereum’s transition to a more ESG-friendly blockchain network and helps the network along its development road map, which may increase adoption of the Ethereum network and, accordingly, the price of ether in the long run.