Author: Steven Shi, Investment Partner of Eco Fund
Original compilation: Leah Yuan, Foresight News
Ethereum may complete the final phase of its move to proof-of-stake within a month. After the upcoming "Shapella" upgrade (a portmanteau of "Shanghai" and "Capella"), more than 16 million ETH staked on the Beacon Chain will be able to be withdrawn. This figure alone has caused anxiety among users, investors, and traders, who are concerned about the possible impact on the supply and demand of ETH and other related tokens.
In this report, we will dissect the details of the Shapella upgrade, provide considerations before the upgrade, and give our best estimate of the impact the upgrade may have on ETH, liquid pledged derivatives, and other applications.
Quick Facts
The Shapella upgrade, which will allow users to partially or fully withdraw ETH staked on the Beacon Chain, is expected to take place from late March to early April.
Over 1 million ETH rewards have been accumulated on the beacon chain, and these rewards will be automatically withdrawn and converted into liquid assets within about 5 to 7 days. Given the competitive dynamics among staking pool providers, we expect about half of these rewards to be re-staked on the Beacon Chain.
More than 16 million ETH principals are pledged on the beacon chain. We expect that about 2.8 million ETH will be withdrawn, but this is only a conservative estimate, and the final result may differ significantly from the estimate. We expect ~50% of the withdrawn ETH to be redirected back onto the Beacon Chain via liquid staking providers. Key beneficiaries include Lido Finance, Frax Finance and Coinbase.
In the long run, as staking ETH will become more secure, Ethereum's staking participation rate should exceed 40%, becoming a true benchmark rate in the Ethereum ecosystem.
Shapella will also have a knock-on effect on DeFi applications, including: liquid staking providers, decentralized exchanges, lending protocols, meta-governance projects, and re-staking infrastructure. Shapella could also have an impact on ETH’s supply elasticity, affecting its volatility as a trading tool.
Basics of the Shanghai/Capella Upgrade
The Beacon Chain is the backbone of the Ethereum PoS (Proof of Stake) consensus, allowing users to participate in block validation and earn rewards by staking their ETH. The chain started accepting deposits in November 2020, andOfficially launched on December 1, 2020. Users need to pledge at least 32 ETH to perform consensus work to obtain issuance rewards. Prior to the merger, consensus work consisted only of reaching consensus on active validators and their balances. After the merger (Bellatrix upgrade), validators are responsible for ensuring network security, such as block proposals, block certification, and synchronization committee participation.
But since launch, staked ETH and rewards have been locked on the beacon chain. As a result, over 16 million ETH staked and 1 million ETH accumulated rewards have been locked on the Beacon Chain for over two years.
Next month, Ethereum will complete an upgrade called Shanghai/Capella. [The upgrades of the consensus layer are named after stars (Altair, Bellatrix, Capella), and the upgrades of the execution layer are named after cities. ]
The most obvious feature of the upcoming Shanghai/Capella upgrade is that validators can choose to withdraw their accumulated rewards or the entire validator balance. Therefore, ETH will be transferred from the beacon chain to the execution layer, where ETH is free flowing and can be used for transactions.
Note that rewards from priority fees and MEV payments accrue directly to the execution layer (i.e. the chain on which Ethereum users are currently transacting). Therefore, until the Shapella upgrade, the only rewards that are locked are the issuance rewards of the consensus layer.
All indications are that the Shapella upgrade will happen around the end of March to early April. Test results for the upgrade have been positive so far. The developers successfully launched Shapella on the Sepolia testnet on February 28th,The results showed only minor problems. They have completed testing of the shadow fork on Zhejiang public testnet and mainnet.
source:
source:@terencechain
The next big milestone is to roll out the Shapella upgrade first on Goerli by mid-March and then on mainnet.
Withdrawal instructions:
After the Shapella upgrade, there will be two types of withdrawals:
1. Full Withdrawal: This allows validators to exit the Beacon Chain completely, withdrawing their entire ETH balance, including their initial balance of 32 ETH and accumulated rewards or penalties.
Ethereum has a churn limit coefficient for the number of verifiers who can withdraw funds for each epoch (1 epoch contains 32 blocks, about 6.4 minutes). This limit increases with the number of validators on the beacon chain. Currently, this churn limit factor is 7. But this limit may change to 8 as the upgrade progresses. So when the upgrade happens, 1,800 validators can withdraw in full every day, which equates to over 57,600 ETH per day.
2. Partial withdrawal: Partial withdrawal allows the validator to withdraw the balance beyond the principal of 32 ETH. Validators will continue to be active on the Beacon Chain while withdrawing the revenue they have already earned. Because the accumulated rewards are idle capital (that is, the rewards will not be automatically compounded), partial withdrawals revitalize the idle capital and improve capital efficiency. Without partial withdrawals, validators would have to exit the beacon chain completely to claim their rewards, and then re-enter again. Therefore, separating partial withdrawals from full withdrawals helps improve network stability and prevent queue congestion.
Part of the withdrawal function will be carried out automatically after the upgrade, and it will be cleaned once a week on average. Each slot will automatically make 16 partial withdrawals from the Beacon Chain’s0 indexesInitially, approximately 147,000 validators can make partial withdrawals per day.
Should I stake or quit?
The escalation could create a potentially massive supply shock. Within a week after the upgrade, more than 1 million ETH will automatically enter circulation due to liquidation of partial withdrawals. This accounts for about 0.8% of the total ETH supply and is worth about $1.6 billion at today's prices. In addition, theoretically 16.8 million ETH may also be withdrawn.
These figures have caused some panic. There is a theory brewing that the market may see a massive dump as ETH that has been illiquid for more than two years is allowed to withdraw.
But these concerns may be overblown. In the following, we describe the dynamics of partial and full withdrawals to better illustrate how withdrawals work after the Shapella upgrade, and to illustrate that ETH's supply shock may not be as extreme as the data above suggests.
Note that our estimates are rough numbers and our estimates are subject to many uncertainties. Our analysis involves many assumptions, but the specific numbers estimated in the article are certainly not accurate. Nonetheless, we hope that this report will provide food for thought, and we encourage further participation in the discussion and analysis.
Partial Withdrawals: Re-stakes ease supply shocks
Overall, more than 1 million accumulated ETH rewards will automatically become liquid assets. Since there are 16 partial-withdrawal validator queues per block, partial-withdrawal withdrawers must convert their withdrawal credentials before they can receive these accumulated rewards, so it takes five to seven days to make all partial withdrawals valid. Withdrawals become liquid assets at the execution layer.
But even assuming that most validators are conservative and want to hold liquid ETH on the execution layer, there are some structural dynamics that drive them to re-stake these rewards onto the beacon chain.
53% of cumulative rewards come from staking pools that provide liquid staking tokens (“LSTs”). The main LSTs are provided by relatively decentralized staking providers (such as Lido and Rocket Pool) or centralized custodians (such as Coinbase and Binance). Fair market prices for LSTs vary by provider, with Rocket Pool offering a +1.3% premium to rETH and Ankr offering a -3.0% discount to ankrETH. The price of most LSTs fluctuates in the range of -0.1% to -2.0% discount, and LSTs of centralized custodians are usually discounted even more.
The Current State of Ethereum Liquidity Staking Tokens
Source: Project website, Defillama, CoinGecko, 2023.02.24
We believe that the existence of LST reduces the potential supply/sales that users may bring after withdrawals, as users staking through these providers already have the option to convert their staked ETH at a notional discount.
Additionally, competitive dynamics incentivize restaking. Occupying the leading market share of all staking poolsLido recently confirmed its "Maximize APR" policy, trying to re-stake as much ETH as possible. With a discount of only -0.1% on its stETH token trades, we estimate that the redemption demand for stETH will be minimal. Therefore, Lido may re-stake all 224,000 ETH (approximately $370 million) of its accumulated rewards, which is equivalent to more than 20% of the total accumulated rewards. Other staking providers will also be forced to stake as much ETH as possible to avoid losing market share due to low APR in competition with Lido.
We share our views on the base case assumptions in the table below, i.e. estimating the proportion of partial rewards being re-staked. For staking pools with LST, we hypothesize that higher discounts will result in greater redemption pressure to restore tokens to standard prices.
We acknowledge that our assumptions are relatively loose for a centralized operator without an LST, and that reality may differ from our understanding. For example,In the case of the recent SEC charges and closure of the agreement to provide staking services to US customers, Kraken is unlikely to re-stake. Celsius may also prefer the more liquid ETH.
Nonetheless, our conversations with institutional node operators such as RockX have found that most institutional node operators want to re-stake the majority of their rewards to maximize returns. As such, we expect institutional-focused node operators, such as RockX and Figment, to re-stake the majority of their rewards. For node operators with more retail users, such as Stakefish and most centralized exchanges, we believe that only 20% of the partial rewards will be re-staked, and the remaining tokens will be used to satisfy redemption requests.
Validator Entity Distribution
Partial Withdrawal Assumptions
Source: Beaconcha.in, Rated Network, Dune, project websites, Amber estimates, 2023.02.24
Therefore, our base assumption is that approximately 52% of the total supply of ETH in partial withdrawals will be re-staked back to the beacon chain, greatly narrowing the potential for supply shocks and sell-off frenzy.
For an in-depth analysis of the fractional withdrawal model, we recommend readingData Always of this article. The authors conclude that "as the withdrawal queue is clogged up by validators not intending to sell, the decision to automate partial withdrawals may be influenced, reducing the size of the initial sell-off frenzy."
Full withdrawal: ETH conversion to LST as the basic hypothetical case
We predict that full withdrawals from the top 20 validator entities will account for approximately 75% of total validator withdrawals. Then make generalization assumptions about the remaining 25%:
Staking pool operators who own LST will not withdraw the full amount from any validator. As can be seen above, more than half of the validators are offering LST - we think they can all pull their LST back to parity with ETH with only partial withdrawals. In fact, most staking pool operators will likely deploy new validators due to the rewards for restaking ETH.
Kraken will withdraw most of the staked ETH as it no longer allows staking services to US customers. While it could offer staking in other ways, the attitude of its U.S.-facing customer base and regulatory uncertainty could trigger an outflow of funds from Kraken’s staking offerings. Likewise, we expect Celsius to completely withdraw all of its validators in an attempt to maximize short-term liquidity.
For other centralized exchanges and node operators, we make different assumptions of 10-25% based on our understanding of their target customers (institutional and retail) and operating models (pooled/standalone, custodial/non-custodial). Based on our conversations with various staking participants, we assume that institutional clients have lower withdrawal requirements than retail investors. Those operators with pooling/custodial models face relatively less pressure to withdraw funds in full, as they can balance outflows with inflows.
All in all, we estimate that about 2.8 million ETH will be fully withdrawn from active validators after the Shapella upgrade.
However, similar to partial withdrawals, we expect some ETH to be re-staked. Specifically, many withdrawing stakers will transfer their ETH back to LST's Liquid Staking Provider. This is because suppliers holding LST have several advantages over independent staking:
Potential for instant liquidity through on-chain pools
Use LST as collateral on DeFi, allowing users to lend stable coins, leverage ETH, etc.
Provides a convenient restaking mechanism such as Eigenlayer.
Assumption of full withdrawal
Source: Beaconcha.in, Rated Network, Dune, project websites, Amber estimates, 2023.02.24
Users restaking ETH may reduce the impact of full withdrawals on the total supply shock by about 50%.
A range of on-chain metrics support our forecast for a softening of withdrawal pressure:
After the Terra/3AC crash, the average number of new validators continued to rise within 14 days.
Source: Glassnode
LST tokens trade at a paltry discount for major staking pools, with minimal discounts for non-custodial liquidity staking providers (RPL's LST even has a premium). This supports our view that staked ETH will move from custodial to non-custodial liquid staking.
Source: Company websites, Coingecko
39% of stakers are profitable in USD since they started staking. But a higher profit percentage also means higher withdrawal pressure.
Source: Dune (@hildobby)
Summarize
Source: Glassnode
Summarize
1,136 validators exited the beacon chain, which means that about 39,000 ETH entered the circulation market.
Overall, after subtracting the ETH that may be re-staked, our baseline forecast is that about 1.8 million ETH will be converted into liquid assets through withdrawals. This represents 1.5% of the total ETH supply and 10.5% of the total ETH supply on the beacon chain. We consider this forecast to be conservative and the actual number is likely to be lower than this.
We don't want all of this ETH to be sold. Data suggests that Ethereum’s total supply is increasingly staying on-chain rather than being sent to centralized exchanges. We think this trend is here to stay, and expect a fair amount of ETH unlocked from Shapella to stay on-chain.
Source: Glassnode
Predicting the sell-flow model after the Shapella upgrade is beyond the scope of this report, and interested readers can try to build their own models. However, it’s important to note that any potential sell-off flow from full withdrawals could happen in weeks rather than days due to the Beacon Chain’s exit queue.
the other side of the formula
With the uncertainty of withdrawals finally removed, we expect an influx of net new demand into the staking market. Staking ETH will gradually be seen as the de facto benchmark rate within the Ethereum ecosystem, similar to the fed funds rate in the US economy. All staking yields should be at least on par with DeFi blue chip protocols such as Aave, Compound, and Euler.
Yield excluding token rewards, source: Project websites, Defillama, 2023.02.24
In the long run, we expect more than 40% of ETH to be staked. At a 40% participation rate, the overall APR for a validator is around 2.95%, assuming MEV and preferred fees remain on average over a 30-day period.
Source: Staking Rewards
Implications for Liquidity Staking Providers
Shapella will benefit liquidity staking providers in several ways. As mentioned above, LST is arguably a better collateral than ETH. For example, if stablecoins could be borrowed using stETH instead of ETH, borrowing against ETH would become less capital efficient. Moreover, withdrawal can strengthen LST's anchoring to ETH, thereby increasing its attractiveness as collateral. Therefore, we expect liquidity staking providers to win market share.
Relatedly, it will become cheaper for these projects to maintain their peg between LST and ETH. After the Shapella upgrade, LST's instant on-chain liquidity becomes less important. For example, suppose Lido halved its token issuance to incentivize liquidity on decentralized exchanges like Uniswap. Even if the total value locked shrinks substantially, stETH will remain tightly pegged to ETH. This is because if stETH did trade at a price far below the peg, third-party arbitrageurs would step in. Withdrawals take only 4-5 days in most cases, so we think arbitrageurs will step in even if there is a 10-30 basis point difference from the anchor price. (At 30 basis points, the arbitrageur can expect to achieve 25% annualized return). As a result, DeFi projects like Aave and Compound can still confidently use stETH as a collateral asset, as liquidators and arbitrageurs stand ready to buy any stETH off the peg.
In summary, liquid staking pools are likely to face two outcomes: widespread adoption of LST will increase their revenue, and reduced token issuance will reduce their costs. A counter argument is that these LST protocols still need to incentivize liquidity in order to be well integrated in the DeFi ecosystem and allow instant liquidation for lending protocols. but,Euler's precedent for integrating cbETH as collateral shows, the team realized that instant on-chain liquidity did not fully reflect the quality of token collateral.
We think the biggest beneficiary will be Lido Finance. Because stETH is widely used in different DeFi applications, it provides the greatest utility among LSTs. Lido's portfolio of managed validators allows it to accommodate a large number of staking requirements (More than 150,000 ETH were registered in the last day). Frax Finance and Coinbase would also benefit for similar reasons, although their LSTs are less widespread. Coinbase also offers higher margins with a 25% return.
Rocket Pool may benefit less. While we love its decentralized design and philosophy, it faces scalability challenges due to ETH/RPL collateral requirements. And, in order to attract node operators, its higher profit margin is also an obstacle to growth. Still, it's important to note that its Atlas upgrade is an important catalyst. Atlas allows node operators to halve their collateral requirements from 16 ETH to 8 ETH, which will triple the capacity of each operator - 16 ETH staked in one validator can now be spread across two validators. validators, thus increasing the capacity of other users from 16 ETH to 48 ETH.
second order effect
Shapella doesn’t just affect ETH and staking providers, it has knock-on effects across many industries such as:
A stablecoin with ETH as collateral. Notably, two ETH-collateralized stablecoins, Liquity (LUSD) and Reflexer (RAI), faced resistance after Shapella. With staked ETH de-risking, users may be more willing to use liquid staked ETH as collateral instead of using ETH directly. Also, since Liquity and Reflexer are specifically set up to minimize governance, they cannot switch the collateral base to liquid staked ETH. Maybe that's a good thing, LST is still risky. But these stablecoins already have an extremely small circulating supply (LUSD: $230M, RAI: $7M), and it is likely that they will see lower market share after Shapella.
Curve/Convex and other decentralized exchanges. Lido has beenSpent nearly 86 million of its own governance tokens (about $230 million in today's prices) to incentivize liquidity on decentralized exchanges, mainly for Curve Finance. This is primarily to make stETH a more attractive asset by creating deeper liquidity, allowing transactions between the two assets to have low slippage. The number of LDO issuances usually accounts for at least half of the APR that provides liquidity for the stETH/ETH trading pair. If Lido reduces these rewards, we may see lower stETH/ETH TVL on Curve and other decentralized exchanges, possibly causing their valuations to re-evaluate. For better understanding, on Curve, the stETH/ETH trading pair accounts for about 32% of the total value locked.
Meta-governance projects like Convex and Aura will pool voting power and receive rewards from projects like Lido, which will also see lower rewards in this case.
On the other hand, the Shapella upgrade may also lead to a new batch of liquid pledge tokens. Frax Finance launched its LST just a few months ago and is using its control of CVX/CRV to gain market share. Year FinanceA basket of LST will be launched soon, will follow a similar strategy to obtain liquidity. After the Shapella upgrade, we may see a sudden LST war.
Leveraged pledge. Since the peg between LST and ETH is easier to maintain, users may have more confidence in leveraged ETH via on-chain lending protocols. Primitives like Aave v3's e-Mode allow users to borrow and borrow ETH via LST at 90% LTV, whileGearbox allows trades with up to 10x leverage on stETH, providing interesting places for users to experiment (and fail). Therefore, even if no new users want to stake ETH, existing users may drive the growth of ETH on the Beacon Chain.
Restaking as a primitive. AlthoughEigenLayer's white paper was released this monthYes, but the idea of restaking ETH as a fundamental operation for bootstrapping the new network has already created a buzz in the Ethereum community. The idea isn't necessarily new, both Avalanche and Cosmos have variants of the same concept. But this will further strengthen the position of staking ETH as the required collateral. Furthermore, if the staking participation rate does reach above 40%, the base rate of return will compress to 3% or lower. In this case, platforms that offer higher returns on the same asset will be very welcome.
in conclusion
in conclusion
We don't pay for panic. While anything can happen in the short term, the Shapella upgrade is largely a positive event. Among other things, it induces higher staking rate participation, strengthens the position of staking ETH as an ideal collateral, improves the security of Ethereum, and reduces the risk of Ethereum as an asset.
If we have to call Ethereum "done", when can we call it "done"? end of 2023. The merge is complete and the withdrawal is complete. Ability to withdraw funds from proof-of-stake systems. The basic expansion is completed. Many basic improvements have been made. Enough cryptography has been added to Ethereum so that we can support privacy solutions on top of it. We can stop here if we want. —Vitalik Buterin on the Network State Podcast said
References
References
Beaconcha.in - Beacon Chain blockchain explorer and statistics
Beaconscan - Beacon chain blockchain explorer for Etherscan
Rated Network - Ratings and metrics for Ethereum validators
Lido Ethereum validator and node metrics
Staking Calculator- Beacon Chain
Hildobby's Ethereum Staking Dashboard (Dune)
Appendix: Extraction Design for Lido Finance
Lido's withdrawal is designed to manage withdrawals from the Beacon Chain, which clarifies the overall structure of most LSTs. stETH holders need to go through three steps to withdraw:
1. Request: To withdraw stETH, users can send a withdrawal request to Lido to lock the amount of stETH to be withdrawn.
2. Fulfillment: After that, the user enters the queue, and the Lido protocol processes the requests in a first-in, first-out order. If there is ETH in the buffer from new deposits or partial withdrawals, it can be used to satisfy withdrawals. If there is not enough ETH in the buffer, Lido will ask the necessary number of validators to exit. Note that this queue is internal to the Lido protocol and separate from Ethereum's withdrawal queue.
3. Claims: Users can claim their ETH shares.
Source: Lido
An interesting choice in the design is that fetch requests cannot be cancelled, but can be transferred. This will potentially create a secondary market, where users who have previously queued to withdraw Lido can sell their positions to those who are more in need of liquidity.
Withdrawal of Lido will adopt two modes. Under normal circumstances, Lido withdrawals will run in Express mode. Users can usually convert stETH to ETH in about 3-4 days, which may be shortened to as little as 1-24 hours if there is enough ETH in the buffer.
However, in the event of a "mass punishment event," Lido will go into bunker mode. This can happen when more than 600 Lido validators are penalized at the same time. So far, none of the Lido validators have been penalized. In this mode, all pull requests will be postponed until the loss is socialized among all users. This is to prevent savvy users from gaining an advantage, as both penalties and fines happen asynchronously. Without bunker mode, savvy players predicting massive punishment will occur, so they will withdraw stETH to ETH to avoid losses, and then convert ETH back to stETH.
Lido's architecture prioritizes stakers, with the principle of maximizing the annualized rate of return. Instead of keeping more ETH in the buffer to facilitate faster withdrawals, Lido will ensure that ETH stays in the buffer for as short a time as possible.
Original link
IOS
Android