Ethereum is on the verge of finalizing its transition to proof of stake. The upcoming Shapella fork culminates a series of upgrades bringing the Ethereum consensus and execution layers fully into the proof of stake chain.

These upgrades began in December 2020 when the Beacon Chain was launched, supporting only staking functionality without any applications on top. This allowed the amount of ETH staked to rise, securing the proof of stake network prior to value being settled there. Since then, ETH validators have been depositing into the Beacon Chain’s deposit contract, receiving staking rewards at the expense of locking their capital until proof of stake became fully operational.

The next major step for Ethereum proof of stake was arguably the most complex: migrating applications live on the proof of work chain to the Beacon Chain. This so-called Merge took place on September 14, 2022 after thirteen “shadow forks” tested this process. The merge was successfully implemented, smoothly transitioning applications to operate on top of the proof of stake consensus secured by the deposit contract.

On April 12, 2023 — roughly 850 days after depositors began locking ETH into the deposit contract — validators will be able to begin the process of withdrawing some of the $34B+ of staked funds.

Given the magnitude of capital stake and technical complexity, there is significant uncertainty and lack of understanding of what will happen as these withdrawals begin. Through this piece, we aim to shed light on this process, diving into the dynamics of the Shapella fork, the probable outcomes for the different players in the sector industry and finally on its implications for the future path of the Ethereum network and its native asset.

Overview of Shapella

The Shapella fork consists of conjoined upgrades: Shanghai, a series of Ethereum Improvement Proposals (EIPs) related to the execution layer, and Capella, a major update to Ethereum’s consensus layer.

Out of the EIPs set to be implemented in Shapella, the most notable is EIP-4895, which will enable validators to begin withdrawing their staked ETH. These withdrawals will be split into two categories:

Both of these withdrawals need to be initiated manually by validators prior to entering a queue. These queues are set in place to prevent a mass exodus from validators, which would harm Ethereum’s security.

Partial withdrawals go through a shorter wait, known as the withdrawal queue. Per an IntoTheBlock projection, if all partial withdrawals are attempted just after the Shapella fork (which seems highly improbable), it would take around four and a half days for these ETH profits to enter the market.

More importantly, full withdrawals, which make up for the majority of ETH locked in staking will go through a much longer queue. It would take approximately 100 days for one third of validators to exit if they all attempt to exit simultaneously, translating into ~$80-$100M worth of ETH being withdrawn per day. This would make up about 1% of ETH’s daily trading volume, though it is unlikely that all withdrawals will be sold.

To better understand how ETH withdrawals translate into selling (or buying), it is worth further exploring the key actors in the staking industry and how Shapella affects this space as a whole.

Staking Capital Flows

To better understand how much ETH is likely to be withdrawn from the staking contract, and how much of that will be sold, it is worth taking a closer look at the main Ethereum validators and the forces affecting them.

Liquid Staking Derivatives (LSDs)

These are protocols such as Lido, where a validator position is represented by a synthetic token, like stETH, tied to the price of Ether. This allows holders of LSDs to not be locked in and swap stETH for ETH at any time they wish through a decentralized exchange. stETH preserves a price close to ETH since after withdrawals are enabled on Lido, one stETH can be unstaked for one ETH.

LSDs currently make up over 35% of all ETH staked. Given their liquid nature, one could imagine relatively low demand for withdrawals from this category as they are not subject to the same lock-up as other validators. Moreover, even if people would like to redeem stETH for underlying ETH, Lido stated that they do not expect to be ready to process withdrawals until Mid-May, delaying any potential outflows.

There is a case to be made for LSDs to actually receive *net inflows* after Shapella. Since LSDs are not subject to exit queues, they will be be a faster way to go from staked ETH to regular ETH. Perhaps more importantly, as American exchanges are getting an increasing amount of regulatory scrutiny, some of the ETH staked with them is likely to migrate into other options.

Conclusion: LSDs are poised to increase their deposits and market share of the total amount of ETH staked.

Unidentified Stakers

The next largest group of validators are labelled as “unidentified”. These are mostly individuals and small organizations that stake their ETH independently. This group began as the largest set of validators but has seen its market share decline over time.

Being the most heterogenous set of actors in the staking space, unidentified stakers’ behavior is difficult to predict. Since this group initially made up for approximately half of ETH staked, there is an argument to be made for some unidentified stakers to withdraw at least part of their capital, and potentially realize gains. At the same time, though, since these unidentified stakers took the effort to go through the intricacies of staking independently without knowing when they would even be able to withdraw, it is highly probable that many of them are long-term ETH believers. This suggests that even if unidentified stakers withdraw, they are less likely to sell and may potentially switch to LSDs for simpler and faster staking.

Conclusion: Likely to withdraw some ETH, but not so likely to sell it.

American Exchanges

Coinbase and Kraken were early adopters to offer staking services for their users to earn yield on their ETH. While this has allowed them to gain nearly 20% market share of ETH staked, government intervention is likely to lead this percentage to decline.

Just two weeks ago, Coinbase received a Wells Notice from the SEC, alleging that its staking products are unregistered securities. Prior to that, Kraken was fined $30M to settle penalties with the SEC for similar complaints and later announced it would be shutting down its staking operations. Due to these reasons, American CEXs are likely to experience the largest withdrawals in the days following the Shapella fork.

Since American exchanges charge higher fees on staking than LSDs, one can presumably expect that those staking with them do it because of convenience. Once this option is removed, or at least strongly opposed by the government, it is viable to believe many of their users could sell the assets they were staking. Another portion of them may choose to simply withdraw the ETH and hold it, or potentially move it into LSDs if they are slightly more advanced users.

Conclusion: Large withdrawals to be expected, some selling pressure as lazy users may not want to go stake elsewhere.

Other Staking Services

These are institutions like Stakefish, Figment or Blockdaemon. Companies like these manage validators in a centralized way and target mostly institutions in exchange for a fee.

Since the majority of the profits obtained from staking ETH for other entities is locked, one can expect staking services to conduct partial withdrawals to cover their costs of operation. Full withdrawals are less likely, though, as they manage ETH holdings on behalf of other clients. Here one could interpret that, similar to unidentified stakers, clients of staking services are to likely to have a long-term interest on ETH since they were willing to lock their capital without knowing when they would be able to withdraw it.

Conclusion: Partial withdrawals and profit taking to be expected, but not en masse.

Global Exchanges

International crypto exchanges are likely to benefit from the troubles of their American counterparts. Though many theoretically don’t accept American users, they are typically able to support users with a second nationality that simply use a VPN routing their IP address outside of the US. This, in addition to shorter lock-up periods, is likely to drive inflows into global CEXs, which at the moment have just under 10% market share.

Out of this 10%, some users are expected to withdraw after being locked in for as long as years. The bear market may have also caused some to lose faith in crypto and lead them to sell, or potentially just forget about these funds.

Conclusion: Net inflows projected, though some selling may take place.

The dynamics explained here are likely to lead market share to reshuffle between the players in the staking industry. But perhaps more importantly, we should discuss how we should expect the size of the industry to change following the Shapella upgrade.

Growing the Pie

Even though withdrawals will decrease the amount of ETH staked in the days after Shapella, there are reasons to believe why it would increase in the following weeks.

In a Blockworks podcast, Coinbase’s head of staking predicts that the amount of ETH staked will be shaped as a J curve, meaning that they would first decline prior to climbing. A key reason to support this is that following the Shapella fork, staking on Ethereum should become effectively de-risked.

For one, technical risk should be eliminated assuming the upgrade works as intended. Then economic risk is also diminished as validators would no longer be locked for an uncertain amount of time. These risks are likely among the reasons why ETH currently has a significantly smaller percentage of supply staked in comparison to other proof of stake chains.

The top PoS chains average 45% of the supply of their native stoken staked. Apart from ETH, BNB stands out as the only other chain with less than a third of its supply staked. A potential reason for this is that Binance users who hold BNB receive a 25% discount on trading fees, giving the tokens extra utility not available across other proof of stake networks.

Since this feature does not apply to Ethereum, it is highly probable that the amount of ETH staked will increase in the weeks after the Shanghai fork. Retail holders who were unsure about depositing their assets for an uncertain amount of time may begin to slowly stake more. To a larger extent, institutional investors that were previously unable to lock their limited partners’ capital into staking will now be able to earn yield on their ETH as this risk is eliminated. This should lead the amount of ETH staked to increase over time, potentially reaching as high as 25%-30% within a year. Furthermore, this would decrease the amount of ETH supply available to be sold, while making the Ethereum network more secure.

Overall, the Shapella fork concludes a milestone years in the making. EIP-4895’s enabling of withdrawals is expected to lead to short-term volatility as some locked ETH is withdrawn, sold or re-staked elsewhere. Through this process, market share within staking is likely to further strengthen LSDs at the expense of American exchanges undergoing regulatory scrutiny. Finally, both ETH holders and Ethereum users will benefit from the upgrade longer-term as it leads to a more scarce native asset and a more secure blockchain.

Read more: https://medium.com/intotheblock/estimating-the-impact-of-ethereums-shapella-upgrade-f3b698b24a24