Ethereum 2.0 Migration Economics
How to design incentives to maintain consistency during system upgrades
As a follower and supporter of the Ethereum project, I always support healthy conversation around network developments and upgrades. The goal with this article is to surface a thoughtful discussion about the economic merits and challenges of the current Ethereum 2.0 ether migration design with an approach of a one-way bridge.
In my view, Ethereum’s story is equal parts technology, and equal parts finance, and with the migration to Ethereum 2.0 approaching, it is important to consider both technical and economic challenges together, and with equal scrutiny.
From a technical standpoint, the viability of a one-way bridge versus a two-way bridge is well represented and articulated; however, the discussion around the strengths and challenges of the economic design for the migration are currently underrepresented in the discussion, and deliberation over the economic aspects of a one-way migration has been limited so far.
What discussions have occurred so far?
Currently, consensus has formed in the Ethereum developer community around the mechanics for migrating ETH 1.0 ether (on the existing “proof-of-work” chain) to ETH 2.0 ether (on the developing “proof-of-stake” chain) through a one-way bridge.
Additionally, there is support for a two-way bridge, and preliminary discussions around a potential native integration down the road, but given the current spec, the intention with this article is to focus on the economics of the one-way bridge, and to introduce what I believe are a few of the key economic incentives that will be present under this approach.
What does a one-way bridge migration look like?
A one-way bridge will exist whereby current ETH 1.0 ether holders can elect to burn their ETH 1.0 ether in return for an equal amount of ETH 2.0 ether (also being referred to as beacon ether, or bETH). That is, bETH will be minted and locked in a staking deposit contract on the proof-of-stake mainnet (the “beacon chain”). The migration is scheduled to start with the first ETH 1.0 ether deposits at Devcon V in October 2019 and ETH 2.0 staking and Phase 0 mainnet launch in early January 2020.
Why is a one-way bridge the preferred approach?
Below are some of the pros and cons of a one-way and two-way bridge migration highlighted on EthHub. Notice that the listed pros for a one-way bridge are primarily technical and the listed cons primarily pertain to economics. That is, a trade off is present between technical and economic challenges of a one-way bridge vs. a two-way bridge.
What does inflation look like in ETH 1.0 and ETH 2.0?
As of January 2019, ETH 1.0's block reward dropped to 2 ether every ~15 seconds, resultingly about 4.8% new ether is issued annually. At some point in the future, ETH 1.0’s inflation is planned to be reduced to zero as ETH 1.0’s activity is fully migrated over to ETH 2.0.
What is not clear in the above chart is how ETH 2.0’s migration and estimated inflation is factored in. Although it seems there is an upward blip in issuance at Phase 0, which would imply that ETH 1.0 and ETH 2.0 issuance is bundled together above.
Separately, here is the proposed ETH 2.0 issuance schedule. Notice that the issuance is set up in tiers that depend on the total number of bETH staked. The incentives are set to be higher for the early validators, and incentives fall off as more validators and staked bETH come online.
Let’s break down three cryptoeconomic design decisions introduced above:
- “ETH 1.0 ether is burned” —In contrast to Figure 2 above, I think it is helpful to consider ETH 1.0 and ETH 2.0’s supply schedules separately given the proposed one-way bridge approach. A burn-and-mint model implies an inverse relationship between ETH 1.0 and ETH 2.0’s total supply (i.e. for each ETH 1.0 ether that is burned, an equivalent number of bETH is created). This means that the ETH 1.0 ether supply will turn deflationary if more than 4,890,000 ETH 1.0 ether is migrated and burned. That is, ETH 1.0’s projected inflation rate of 4.8% will decrease by the percentage of ETH 1.0 ether burned and migrated to ETH 2.0. As a result, the price of each remaining ETH 1.0 ether should increase in value as a result of this dynamic, ceteris paribus.
- “bETH is minted and locked” — bETH will be locked (~1.5 years) on the beacon chain. As such, only limited bETH liquidity will be available through futures or off-market transfers negotiated direct between individual validators. Given bETH’s limited liquidity, I’d expect that ETH 1.0 ether will hold a significant liquidity premium over bETH. Said otherwise, the price of ETH 1.0 ether will be greater than bETH for the estimate 1.5-year lockup. For ETH 2.0 validators that choose to accept the liquidity and execution risk, rewards are provided in the form of bETH yield (see Figure 3). However, the success of early validator’s decision to burn ETH 1.0 ether, and stake bETH, depends on the bETH to ETH 1.0 exchange rate (closer to 1-to-1 the better) and bETH interest earned as a validator. Validators risk the possibility that bETH may trade at a discount to ETH 1.0 for a significant amount of time (e.g. if ETH 2.0 development is delayed or if the two chains actually never fully merge). Given its illiquidity, I wonder about the possibility of a portion of bETH to be wrapped on ETH 1.0 to unlock greater liquidity and utility (i.e. wrapped-bETH, or wbETH).
- “bETH locked in the staking deposit contract will earn interest” — It is important to note that ETH 2.0 validator rewards will be denominated in bETH. When savvy crypto holders are evaluating the best way to deploy and earn with their assets, not all interest is created equal. Let’s quickly compare a decision a crypto holder may face: Option 1: Supply DAI on Compound and earn 14.80% annualized floating interest paid out in DAI, or Option 2: Supply bETH as a validator and earn between 18.10% and 1.56% annualized floating interest paid out in bETH. These interest rates under Option 1 and Option 2 cannot be directly compared because of the differing denomination and risk sets, e.g. while DAI is exposed to smart contract risk, bETH is exposed to illiquidity and execution risk. The point here is that the bETH interest earned by early ETH 2.0 validators may not provide a strong enough financial incentive for the savvy crypto investor to accept the risk associated with ETH 2.0 staking, and to forego the many other ways that they can deploy their ETH 1.0 ether.
Bringing all this back, the primary issue in my view under the proposed ETH 1.0 to ETH 2.0 one-way bridge migration, is that 1 ETH 1.0 ≠ 1 bETH, and this may have a ripple-like effect across both ETH 1.0 and ETH 2.0’s developing cryptoeconomic ecosystems, with unintended consequences for the community.
Ultimately, I believe we would want to create a healthy economic situation for both ETH 1.0 and ETH 2.0, such that the right incentives are in place for both to grow in tandem, but with economically isolated ecosystems, unintended competition may emerge between ETH 1.0 ether and bETH holders. Ideally, financial incentives would exist to align the Ethereum community on both chains.
Future settlement as a price stability mechanism
I want to introduce one final point here, and that is how the game theory may play out considering there is a probability that ETH 1.0 will fully merge over to ETH 2.0 at some undefined date in the future (as is the current plan).
I’d suggest that this probability is neither 100%, or 0%, but somewhere in between. And even a small 10% probability that the migration is fully complete at some point in the future, should have a significant impact to limit price volatility between the ETH 1.0 ether and bETH in the interim, but I’m not sure how this will play out in the markets.
Example 1: In equities, I think proposed mergers are a good example of the above: if Company A offers $1.00 per share for all of Company B’s outstanding shares, and there is a 99.0% chance the merger closes, the market will reflect this likelihood by pricing Company B shares near $1.00.
Example 2: In decentralized finance, I think one of the best examples of the above dynamic is MakerDAO’s Global Settlement as a stabilization mechanism for DAI: For those that own DAI, part of the reason why DAI remains close to $1 is because at any point in the future, all of the DAI may be returned for $1 via a global settlement of assets. If you buy (sell) below $1 you make (lose) money if a settlement occurs. Similarly, if the merging of ETH 1.0 to ETH 2.0 is reasonably likely, bETH should exchanged hands for a price similar to that of ETH 1.0. Any delta should only then reflect ETH 2.0 implementation risk, or again, a liquidity premium / illiquidity discount.
Reserve’s Analysis of the MakerDAO Protocol
Thanks to the MakerDAO team for their feedback on this post, and for their openness to constructive criticism!
The analogies to equity mergers and MakerDAO are not perfect here (i.e. ETH 2.0 is not a company nor a collateral-backed stable token system), but it is possible that a similar dynamic may form for bETH / ETH 1.0 ether.
Question for the community:
Under a one-way bridge migration, what are economic incentives that can be added to better align ETH 1.0 and ETH 2.0 stakeholders?
I hope this article helps to initiate discussion around upcoming migration and look forward to hearing the perspective of others in the community. Thanks!
Feel free to reach out on Twitter @aaronahay for discussion or corrections.Thanks!