Bixin Ventures: What expectations can the Ethereum Shapella upgrade bring us?

Original: “What can we expect from Shapella?”

Authors: Henry Ang, Mustafa Yilham, Allen Zhao, Jermaine Wong & Jin Hao

Compiled by: Evan Gu, Evie Xie & Vivian Fang

What stage are we at in the Ethereum roadmap?

The Ethereum merge spanned the entirety of 2022. Ethereum passed the Paris upgrade, transitioning from proof of work (PoW) to proof of stake (PoS). On April 12, Ethereum will undergo the Shanghai and Shapella upgrades, allowing staked Ether to be withdrawn, and Ethereum will fully transform into a proof of stake mechanism blockchain.

In the article below, we will share the background of the "Shapella" upgrade, its impact on the $ETH price, who the beneficiaries are, and what the next steps in the Ethereum roadmap are.

Background of the Shapella upgrade

Although Shapella appears to be a single change to Ethereum, it is actually a combination of two simultaneous upgrades—Shanghai and Capella. The execution layer of Ethereum will undergo the Shanghai upgrade, while its consensus layer will undergo the Shapella upgrade.

The Shanghai upgrade consists of five EIPs, but the most important is EIP-4895, which allows all Ether that has been staked and locked by validators in Ethereum's consensus layer (the beacon chain) since its launch in December 2020 to be withdrawn. Shapella is the third major upgrade of the consensus layer, aimed at facilitating this process, not only allowing for the processing of withdrawal requests but also implementing account clearing functionality, which we will detail later.

Once the Shanghai and Shapella upgrades are completed, ETH withdrawals will be available in full or in part.

Partial withdrawals

Partial withdrawals only involve the portion of Ether that exceeds the 32 ETH required to operate a validator. This account clearing (or "scraping") will occur automatically and periodically for any active validators that have been updated to the new withdrawal credentials.

This feature has significant utility for two main reasons:

Since validator rewards do not automatically reinvest, this mechanism can improve the capital efficiency of stakers' funds. Stakers will be allowed to reallocate their excess ETH to earn other yields without incurring any gas fees.

Additionally, partial withdrawals can prevent long exit queues and excessive validator rotation; otherwise, validators would be required to fully exit the beacon chain to access their rewards, which could potentially destabilize the network.

Since only 16 partial withdrawals will occur per epoch, the frequency of these withdrawals heavily depends on the total number of eligible validators. Validators should expect intervals for partial withdrawals to be between 2 to 5 days.

Full withdrawals

If a validator chooses to fully "exit" as an active validator on the network, they will reclaim their 32 ETH along with all rewards accumulated since their creation; this process will occur during a full withdrawal.

Similar to other PoS blockchains, validators will have a release period during which they will wait to receive their full staked and reward balance. The length of this release period is determined by the sum of two variable times: the time required for the validator to exit the Ethereum consensus layer and the time required for the entire withdrawal process. Therefore, validators seeking to make a full withdrawal should expect at least 261 epochs or 28 hours.

Another factor to consider is the circulation cap. This defines the maximum number of validators that can exit the consensus layer per epoch, a number that increases with the total number of active validators. Currently, there are about 525,000 validators, allowing for 8 exits per epoch, which translates to a maximum of 1,800 validators exiting each day. Consensys explains in a post how these estimates are achieved.

What does this mean for the price of $ETH?

So far, over 18 million ETH have been staked on the beacon chain, making it clear why investors have been filled with concern in the months leading up to the Shapella upgrade. On the surface, enabling withdrawals should lead to a supply shock, exerting significant downward pressure on the price of ETH. But what will the real impact be?

The situation of existing ETH stakers

To estimate the potential price impact, we first need to assess the status of existing stakers and the decisions they may make. We believe these decisions are largely driven by the financial situation of the stakers. In other words, are ETH stakers in profit or loss?

Data from hildobby shows that based on the value of staked ETH assets, only 27% of stakers are actually in profit, while the remaining 73% are in a loss position. Based on this data, stakers are aware that selling their current ETH would result in a loss, prompting them to re-stake accumulated rewards in an effort to maximize future gains.

Staked amount in liquid staking protocols

Another factor to consider is how much ETH has already been staked through liquid staking protocols (LSPs). As of this writing, according to DefiLlama, this number exceeds 30% (and approaches 45% if including staking on Coinbase), with most ETH staked in Lido.

In the case of Lido, stETH automatically readjusts to account for the cumulative rewards earned through liquid staking. This means that users who choose to stake in Lido can "realize" ETH rewards by selling on the secondary market. Therefore, for the validators behind Lido, direct withdrawal pressure is relatively low, and they can flexibly manage excess rewards.

We believe these validators will better utilize these excess rewards by expanding their number of validators rather than selling them. This further alleviates the overall downward pressure on ETH withdrawals and inflows affecting its price.

The price of $ETH

Now that we have considered the key factors determining the potential sale of ETH rewards, what does this mean for the price of $ETH?

"If we assume that validators will sell 50% of their staked ETH rewards (rather than their primary staked ETH), we expect approximately 553,650 ETH to be sold. Amortized over 7 days, this equates to about 1% of daily ETH trading volume (including spot and perpetual futures trading volume).

Given the overall risk environment and the anticipated liquidity situation for ETH during the Shanghai upgrade in early April, we believe the impact of this amount will gradually translate into a slight decline in ETHUSD. Another perspective is that the smooth execution of the Shanghai upgrade is an overall bullish factor for Ethereum as a technology, thus ETHUSD may rise."— Galaxy

While we cannot provide price predictions, we would like to share Galaxy's perspective, who wrote an article exploring the potential outcomes based on partial and full withdrawal execution volumes in a post, which we highly recommend reading.

Regardless of how the price ultimately plays out, we believe Shapella will bring new opportunities for Ethereum. We will share specific details in the following sections.

Beneficiaries of the Shapella upgrade

Liquid Staking Protocols

The Shapella upgrade could serve as a bull market catalyst for liquid staking protocols (LSPs) from multiple perspectives.

With ETH now being withdrawable, LSPs should find it easier and more economical to maintain the price peg between their respective liquid staking derivatives (LSDs) and ETH. This allows LSDs to better borrow collateral, as tighter peg rates help limit price volatility, thereby reducing liquidation risks. Coupled with the growing adoption of LSDs within the Ethereum DeFi ecosystem, this will create opportunities for investors seeking higher capital efficiency yields.

Tighter pegs also promote liquid staking of ETH, making it feel more like "risk-free" or benchmark interest rates for investors. They have previously been cautious about the long-term viability of the peg, as it largely depends on the supply-demand balance of various liquidity pools. Recent failures (due to massive sell-offs on the curve, leading to the devaluation of stETH) have left a strong impression.

For these reasons, we expect that most of the ETH withdrawn and idle ETH will flow into LSPs, and their adoption rates will further increase. However, which LSPs have made more adequate preparations than others? Below, we will introduce some different LSPs, detailing what they do, their flywheel effects and feedback loops, and how these LSPs allow users more control over their withdrawal keys, NFT utilization, and MEV strategies for additional yields.

Lido Finance (LDO, stETH)

After the upgrade, Lido's TVL is expected to see significant growth, solidifying its position as a leader in the LSP space. With a strong reputation, deeper liquidity, and extensive DeFi integrations, users looking for liquid staking projects have ample reason to choose Lido. Since November, the number of Lido token holders and its TVL have been on the rise, accounting for nearly 75% of the total ETH staked so far.

With Lido's increasingly strong position in the LSP space leading to centralization concerns, coupled with heightened governance scrutiny following the Arbitrum incident, the governance system will be more important than ever. While they have taken steps to incorporate stETH holders into the LDO holder governance system, which is commendable, users looking to help decentralize ETH tokens may also consider other alternatives listed below. Lido recently announced that its upcoming v2 will take steps toward further decentralization, focusing on two major upgrades around withdrawals and staking routing.

Swell Network

Swell Network is another upcoming LSP aimed at providing users with higher staking yields. Currently, their latest version, Seawolf, is running on the Goerli testnet, and Swell is preparing to launch on the Ethereum mainnet in late April following the Shapella upgrade. While initially still needing to work with a group of authorized professional node operators to provide users with scalable and reliable yields in a competitive environment, subsequent iterations will see an expansion of the operator pool, ultimately transitioning to a permissionless model with sufficient liquidity, stability, and risk mitigation technologies like DVT, which align with the protocol's value orientation.

Swell's liquid staking token, swETH, will be a yield-bearing token whose value increases with the accumulation of on-chain rewards. Users staking in the Swell Network can earn rewards from consensus layer rewards (e.g., staking yields) and execution layer rewards (e.g., priority fees and MEV). Users staking in the Swell Network will also have access to vaults running yield-enhancing strategies, all completed within their dApp. Swell will position itself as the lowest-cost staking project in the market, with no protocol staking fees.

Rocket Pool (RPL, rETH)

Rocket Pool appears to have advantages over more dominant players like LDO in certain aspects, such as a more decentralized node operator network and lower capital requirements for node operation (only 16 ETH required). However, it also has its drawbacks. These include higher performance fees compared to competitors and the requirement for node operators to lock a certain proportion of the value of their staked ETH in RPL as collateral for protocol insurance.

Despite this higher friction for users wishing to operate nodes from Rocket Pool, the advantages of the upcoming Atlas upgrade may be enough to offset it. Scheduled for release on April 18, the Atlas upgrade brings several improvements, including enhanced protocol efficiency, increased rewards for node operators, and significantly increased rETH capacity while maintaining the protocol's fully permissionless nature. One of the most important features introduced is the 8 ETH mini-pools, which will further lower the minimum capital requirements. Additionally, this feature can provide higher returns for node operators and rETH stakers, offering up to 18% additional returns when users run two 8-ETH mini-pools instead of one 16-ETH mini-pool. While this is beneficial for Rocket Pool's future development, we still need to observe whether it effectively attracts TVL and market share from its competitors.

Frax Finance (FXS, frxETH/sfrxETH)

This year, Frax's rapid market share growth has made it a strong competitor with promising prospects ahead. This growth is easily linked to Frax's offering of approximately 5.6% maximum annual staking yield and its dual-token model. In this context, frxETH acts as a loosely pegged stablecoin to ETH, while sfrxETH is the staked version of frxETH that can earn staking rewards. Through this design, Frax allows frxETH holders to earn in various ways, either by staking to receive sfrxETH and validator rewards or by providing frxETH-ETH liquidity on Curve. These two options together not only create deeper liquidity for frxETH but also increase the annual yield in sfrxETH and Curve liquidity pools.

Like Rocket Pool, Frax also plans to upgrade after Shapella, which could serve as a powerful catalyst for its adoption. The frxETH v2 upgrade aims to increase decentralization by enabling validators to operate without restrictions (run by the protocol itself). Additionally, a recent proposal was passed to utilize FXS bribery and incentives to kickstart liquidity for future frxETH trading pairs, creating a strong flywheel that theoretically will lead to increased value for FXS and frxETH, as well as generate higher yields and attract more liquidity for LSDs. ZhouYeMen from DWF Labs has provided an in-depth explanation of this topic.

Ether.fi (eETH)

Ether.fi is another decentralized LSP entering the market, but it has some important distinguishing features. Ether.fi uses a non-custodial delegated staking protocol, allowing users to generate and hold their own ETH keys. Another feature is that it utilizes NFTs, minting one for each validator launched through the protocol. These NFTs control the 32 ETH stake and store metadata related to the validator, such as client, geographic location, node operator, and any node services. The eETH from Ether.fi is minted from the liquidity pool of these NFTs.

The combination of these two mechanisms allows users to submit withdrawal requests themselves, rather than relying on the usual node operators. The reclaimed staked ETH will be deposited into a withdrawal safety vault, where users will be able to net recover their ETH by burning their NFTs. This effectively reduces the significant and opaque counterparty risks users face in other LSPs while ensuring that there will always be sufficient ETH liquidity for eETH holders to redeem. Therefore, we expect Ether.fi to be one of the LSPs experiencing a surge in users and TVL after Shapella.

In the future, Ether.fi plans to leverage EigenLayer to create a node service marketplace where users and node operators can register their minted NFTs, provide node infrastructure services, and share the revenue from these services with users and node operators. Currently, Ether.fi is already offering institutional staking delegation services, customized investment structures, and developing early user programs for retail liquidity staking. With the enablement of ETH withdrawals, institutions may also view ETH staking delegation as a viable investment, positioning Ether.fi as one of many beneficiaries.

Manifold Finance (FOLD, mevETH)

Manifold is also a new entrant in the LSP space, having launched its MEV-optimized, multi-chain LSP just under a week ago. mevETH is implemented as a cross-chain universal token (OFT), allowing it to bridge directly between chains without wrappers, making it the most composable LSD to date. At the same time, mevETH is designed around MEV capture, employing Manifold LSP's unique and novel MEV approach to provide users with additional yields beyond staking rewards. One method is through arbitraging the peg between ETH and mevETH, which not only provides a means for additional yields but also helps strengthen the peg relationship.

Manifold has stated that their goal is to accumulate 100K ETH by the end of this year, which is about one-sixth of Lido's currently staked ETH. As LSDs gradually integrate into existing DeFi protocols, the high composability of mevETH may propel it to become a leader in adoption, carving out a strong market share for itself.

Views on LSDs

Regardless of which protocol ultimately benefits the most from the Shapella upgrade, the increased appeal of liquid staking will almost certainly lead to higher participation in ETH staking, whether due to natural increases or protocol-driven developments. Based on the staking ratios of other PoS chains, such as Polygon and Solana, the ETH staking ratio seems likely to increase to a range of 40-70%, at least double the current 15%. While this may raise concerns about the eventual decline in staking yields and greater supply inflation, it also brings many other opportunities and possibilities, which we will explore below.

DeFi

Pendle Finance (PENDLE)

Pendle is a yield trading protocol that allows users to execute various yield management strategies by wrapping yield tokens and splitting them into two tokens—PT representing the principal and YT representing the yield, both of which can be traded through their custom AMM. 1 PT gives users the right to redeem the underlying asset at maturity, with the maturity date determined by the initial token holder, while 1 YT gives users the right to receive yield on the underlying asset before the maturity date. This enables Pendle to effectively create a yield market, allowing users to adopt various strategies, including holding assets long-term at a discount, targeting yield annualized rates, or even obtaining low-risk and stable fixed income.

Similarly, locked ETH users can utilize Pendle to split their stETH into 1 PT stETH and 1 YT stETH, then trade either token through Pendle's custom AMM and direct bet on yields. With the upcoming Shapella upgrade and the anticipated decline in staking yields over time, one possible trade that stETH holders can take is to sell YT stETH at the current price to realize differential yields based on their existing earnings.

Since December, Pendle's TVL has experienced a sevenfold surge as more users discover its potential. According to DefiLlama's token TVL breakdown, over one-third can be attributed to LSDs, including stETH and frxETH. With Pendle's latest integration of sfrxETH, we may see an increase in the proportion of LSDs in Pendle's TVL.

Gearbox (GEAR)

On the topic of leverage, the tighter connection between LSDs and ETH following the Shapella upgrade reduces liquidity risks and unlocks the feasibility of leveraged ETH staking. Gearbox is a bi-directional lending market that connects users seeking to provide their assets with passive and secure APYs to leverage enthusiasts seeking additional capital. The protocol introduces leveraged liquid staking derivatives, or LLSDs, as a one-click strategy that allows users to stake up to 10 times the value of their ETH collateral to achieve higher staking yields, exceeding 12%, compared to Lido's current 4.3% APR.

So far, this strategy has been enabled for Lido's stETH and Coinbase's cbETH (pending DAO vote), with plans to integrate frxETH and rETH once they have Chainlink oracles on the mainnet. As a foundational layer for leverage, Gearbox's positioning allows it to indirectly benefit from the anticipated increase in ETH staking volumes and the growth and adoption of LSPs and LSDs, as more ETH stakers seek relatively low-risk methods to increase their staking yields. In recent months, stETH has accounted for over one-third of Gearbox's total credit account TVL, approximately $22 million.

With the upcoming launch of Gearbox v3, new features such as automated portfolio management and health factor maintenance can be expected, further incentivizing LSD token holders to leverage Gearbox's capabilities, pushing it to the forefront of DeFi.

Flashstake (FLASH)

Flashstake is another novel protocol that introduces a unique mechanism to enhance the capital efficiency of user assets. Users deposit and stake their assets, locking in the current yield annual rate, and mint time-based derivatives (TBDs) corresponding to their holdings. Users can then burn or exchange these TBDs to receive their assets' prepaid yields, reclaiming their principal deposit after the selected lock-up period. This is particularly interesting considering that ETH staking yields are expected to decrease over time. Flashstake allows early users to not only secure and fix the current (reportedly highest) ETH staking yields for up to a year but also continue to generate yields elsewhere.

Understandably, this may raise some concerns. Where do these prepaid yields come from? To provide users with this yield, Flashstake employs a three-pool approach consisting of a yield pool, liquidity pool, and acceleration pool.

Unlike loans, Flashstake's pool-based approach also ensures that the assets deposited by users are always 100% collateralized and do not face any liquidation risks. The isolation of each vault strategy also provides an additional layer of security for users, as any potential attack impacts are limited to the affected vault. With its strong value proposition, innovative mechanisms, and robust security, Flashstake's total locked value seems likely to benefit from the influx of liquid ETH and the increase in liquid staking.

Re-staking

EigenLayer

With the Shapella upgrade enabling Ethereum withdrawals, ETH locked in EigenLayer contracts can be withdrawn and reduced if malicious intent is detected. This paves the way for EigenLayer to become a decentralized trust layer built on applications, where existing ETH holders can choose to use their stakes to secure new applications on EigenLayer. EigenLayer's re-staking mechanism provides users with an incremental yield opportunity beyond what they would earn through simply staking ETH and receiving MEV-related rewards, while achieving its primary goal of modularizing Ethereum's trust layer and enabling a broad design space for new middleware and adjacent chains.

Finality Capital Partners expects that by 2028, ETH re-stakers will earn up to 11.5% APY, with an expected base ETH staking yield of around 2.0% in 2028. This 2.0% staking yield is based on their expectation that the ETH staking ratio will reach 60%, rather than the current 15%. Therefore, we see re-staking through EigenLayer becoming a common practice for ETH holders, especially those seeking relatively low-risk methods to increase yields.

New Innovations

Ion Protocol

Despite some success in increasing the adoption of LSDs in DeFi, significant friction still exists today. Due to the presence of various LSDs and liquid staking designs, DeFi protocols wishing to integrate each one will require multiple custom and novel solutions. For example, Uniswap cannot accept variable-base LSDs like stETH from Lido, while blue-chip protocols like Aave and Maker need to create separate vaults and pools for different LSD tokens. This leads to fragmented and shallow liquidity for LSDs, resulting in increased slippage and inefficient price discovery, undermining their original purpose of representing staked ETH and distributing staking yields.

Among its many goals, Ion Protocol aims to utilize a dual-token model to aggregate deposits of various existing LSD tokens into a universal accounting unit, providing a solution to the aforementioned issues. This universal accounting unit will allow for better frictionless integration into DeFi, fully leveraging the potential of LSDs as low-risk, yield-generating collateral, and even improving capital efficiency through leverage.

UnshETH (USH)

UnshETH promotes the decentralization of validators through incentives, directing more yields to less dominant LSPs. This is accomplished through a new class of primitives called LSDfi, which includes validator decentralization mining (vdMining) and validator dominance options (VDO).

vdMining is a token distribution mechanism that rewards users as they invest more in LSPs that meet predefined optimal decentralization ratios, while VDO is a mechanism that allows holders of dominant LSDs to sell "put options" on their dominant percentage, with the strike price set below their current dominance. If at expiration their dominance is below the strike price, VDO holders will lose a portion of their yield, which will be distributed to holders of other LSDs. These two mechanisms together incentivize new stakeholders to engage with less dominant LSPs and encourage more existing dominant LSD holders to diversify their capital across various LSPs.

New features planned for after the Shapella upgrade also include routers for staking ETH liquidity and other enhancements to improve the utility of UnshETH's native token USH. Next, we will see how much TVL UnshETH ultimately captures and whether it can become the de facto yield solution for ETH holders.

Zero Liquid (ZERO)

Zero Liquid is another new protocol that may belong to the new LSDfi track, offering self-repaying loans against users' LSDs. Similar to more traditional money market protocols like Aave and Euler, users first deposit assets as collateral and then can obtain loans based on a portion of their collateral value.

In the case of Zero Liquid, users can deposit LSDs and native chain tokens (ETH, MATIC, etc.) as collateral to receive synthetic versions of their assets as loans. However, these loans are characterized by their 0% interest and self-repaying nature, combined with the use of synthetic assets, allowing the protocol to offer a zero-liquidation model. Zero Liquid achieves this by automatically repaying debts with the yields generated from the collateral deposited by users.

Although the protocol has not yet officially launched its product, many unique use cases beyond maximizing user assets can already be anticipated, making it very convenient in these aspects. With zero liquidity, users can essentially invest and purchase without any upfront capital. There are some similarities to the real-world "buy now, pay later" model, allowing users to make payments through Zero Liquid's products, enabling them to purchase or "consume" immediately and only pay with the future yield of their deposited collateral (or time, depending on how you view it).

Like UnshETH, Zero Liquid introduces entirely new models and mechanisms to address existing problems faced by users while unlocking potential and increasing the use cases for LSDs. Still in its infancy, both protocols have much to prove regarding PMF and their actual ability to attract users to achieve their set goals, but there is no doubt that these new developments provide more optimistic reasons for LSDs.

What will happen after Shapella?

After Shapella, the next anticipated Ethereum upgrade will be the Cancun upgrade, which is part of the scaling roadmap. While second-layer scaling solutions like Optimism and Arbitrum have already reduced costs by over 8 times, the costs of data storage remain high, accounting for over 90% of transaction costs.

EIP-4844 aims to reduce L2 rollup costs by 10-100 times by introducing proto-danksharding to Ethereum, ushering in a new era of low-cost on-chain activity. It supports the realization of Danksharding schemes (e.g., transaction formats, validation rules). EIP-4844 introduces temporary "blob" storage that can be removed from Ethereum when not needed. Ethereum is expected to provide over 100 times the throughput and reduce transaction costs to below $0.001.

Other potential improvements in the Cancun upgrade may include six EIPs from the EVM Object Format (EOF) group that have accumulated over the years, aimed at better constructing bytecode or computer object code, enabling the interpreter to convert it into binary machine code for processing by computer hardware, making the system faster, more efficient, cheaper, and safer. One of these has already been included in the previous London upgrade, some may come in the upcoming Shanghai upgrade, and the rest may arrive with the Cancun upgrade. EOF2 is a new EOF extension capable of fundamentally changing control flow, and these EIPs will support the implementation of EOF2.

The direct beneficiaries of the Cancun upgrade will be those rollups and users that can directly supply Dapp deployments. Thanks to the significant reduction in transaction costs, new applications will also benefit from this upgrade, such as lower costs for on-chain order books, and decentralized physical infrastructure networks will also be able to utilize faster L2 for transactions while settling to the trusted Ethereum base layer. As the Cancun upgrade approaches, we will delve deeper into these opportunities.

Disclaimer: Bixin Ventures is an investor in the three projects mentioned in the text: EigenLayer, Pendle Finance, and Swell Network.