Shapella: ETH’s “Drive to Survive”

Executive Summary

  • Just as John Malone reinvigorated Formula One with a five part plan, including the hit Netflix series “Drive to Survive,” Ethereum developers are successfully executing their own five part road map for the protocol. Last year’s Merge, the first milestone of ETH’s revitalization, which we wrote about here, has had the deflationary impact that we expected and accelerated ETH’s value creation. ETH was up 51% in Q1 2023.
  • The Shapella Upgrade, the next critical milestone, is scheduled for April 12th. It will benefit all ETH holders but will also have many cross-currents, as holders may choose to unstake long-held positions, sell their holdings, change their staking provider, and/or stake for the first time. ETH staking yields will almost certainly decline.
  • We expect a lot of volatility in April as ETH holders take the opportunity to optimize their positions, but we don’t expect a big sell-off.
  • All eyes will be on ETH prices, but the real Shapella winners are likely to be:
    • Liquid staking services like Lido and Rocket Pool
    • Layer 2 protocols like Arbitrum and Optimism



In 2016 most Americans would struggle to name a single driver in Formula One racing, the world’s largest motor-sport. F1 was a distinctly European and male franchise, and one that was losing relevance. Enter new owner Liberty Media and John Malone, who in 2018 unveiled a proposal with five key areas to grow the sport. A deal with Netflix to produce ‘Drive to Survive’ made F1 drivers darlings globally, vaulting F1 back to the top of global sports franchises.


Ethereum has a similarly ambitious five-part roadmap to keep it on top of the protocol standings. 2022’s Merge was a critical milestone, and another is now approaching. Ethereum developers have set a final date for the activation of the Shapella network upgrade for April 12th. Shapella comes from the combination of ‘Shanghai’ and ‘Capella’ which follow the naming convention of using Devcon host cities and the names of stars for network upgrades. The need for using two names in one upgrade is a result of the complexities involved in what will be the final upgrade in the process widely known as The Merge.


Shapella will bring a much needed feature to the Ethereum network in that it will finally allow those who stake their ETH the opportunity to unstake and reclaim possession of their tokens. Up until now this was a one-way deposit requiring full faith in developers to one day deliver this upgrade. While there is much to be excited about for the future of Ethereum including liquidity for ETH stakers brought by Shapella, there may also be volatility ahead due to the 18 million tokens becoming liquid post-upgrade.


The Merge: Heating Up


The Merge, which we wrote about last summer, was the biggest event going into the second half of 2022… or so we thought. FTX and Sam Bankman-Fried quickly overshadowed Ethereum’s last upgrade and sent digital asset markets into a tailspin. One unintended benefit of FTX’s shocking blow up was that many traders went back to decentralized services where they could verify the safety of their assets on-chain. Ethereum was the largest beneficiary due to its dominant position as the largest blockchain in DeFi by almost all metrics.

We anticipated two main benefits from the Merge: deflation in ETH supply and improved profitability of the network. We’ve gotten both. The key variable was no longer the issuance rate but instead how much ETH was ‘burned’ through fees generated by the network. Pre-Merge the supply of ETH was growing at around 3.5% per year resulting in $8 billion worth of dilution annually for ETH holders. Even with subdued network activity during a bear market, ETH’s supply rate decreased to almost exactly 0% from the date of the Merge through the end of the 2022. Then, right as 2023 kicked off and on-chain activity was surging, we began to see the deflation we had called for. In 2023 ETH supply is decreasing at an annual rate of -0.25%, creating structural buy pressure on the token and giving credence to the ‘ultrasound money’ meme.


ETH Supply Since The Merge (September 15, 2022)



The comparison we tried to make in our Merge research was that the Ethereum network uses all profits to perform share (or, in this case, token) buybacks, reducing its outstanding supply as we see above. Since profitability is now directly tied to network activity, it allows investors to use fundamentals to develop an investment view on ETH. In our view this was an underappreciated upgrade from the Merge, as investors have long debated how to underwrite digital assets. Using the same parallels of ETH base fees as revenue and supply issuance as an expense, you can build a simple income statement.


Ethereum Income Statement

Sources: Runa Digital Assets, Dune Analytics, and Blockworks Research.


The full year 2022 data includes the first eight months of the year pre-Merge when ETH’s issuance was substantially higher resulting in the large net loss. With expenses reduced dramatically as Ethereum moved to Proof-of-Stake, the network achieved profitability in Q4 of 2022. The uptick in network activity led to the increase in revenue of 51% to start 2023. As we predicted and hoped for, the market is following the fundamentals closely with the price of ETH also rising 52% in Q1 2023.


Staking Mechanics


Ethereum launched the Beacon Chain and the ability to stake your tokens in December 2020 prior to migrating to a Proof-of-Stake blockchain. They needed to run the Beacon Chain as a live test chain in parallel to the existing chain as they prepared for the Merge. Since December 2020, holders of ETH have had the opportunity to deposit (or ‘stake’) tokens and participate in network validation. Their efforts were and still are compensated with ETH rewards distributed in each block. Risks are involved and, while most of them are technical, we would argue the largest one was accepting an unknown lockup period, since stakers are currently unable to unstake their tokens. The Shapella upgrade on April 12th will enable withdrawals providing a liquidity event to the 18 million tokens currently staked. This is a key upgrade for ETH utility, but one question arises: how many of those who have locked up their ETH for almost three years will take advantage of their new found liquidity to cash out?

Before trying to examine all parties involved, it helps to understand staking mechanics and how withdrawals will be processed. I will do my best to give a brief overview but highly recommend you DYOR.

To stake your ETH you need to either hold 32 ETH or pool your tokens together with others to reach the required 32 ETH minimum. Once you have your 32 ETH deposit and an online validator, you may join the entry queue. This queue restricts the amount of new validators joining the network to 1,800 per day, but this number will increase as more validators come online, eventually reaching 2,200 per day.

There will be two types of withdrawals: full and partial. A full withdrawal is the process of exiting the validator set, no longer participating in the network, and receiving your full 32 ETH deposit back. A partial withdrawal is claiming the staking rewards issued to your validator over its lifetime.

To preserve stability there are two types of queues:  a withdrawal queue and an exit queue. Full withdrawals are subject to the withdrawal queue and an exit queue. Partial withdrawals bypass the exit queue and go straight to the withdrawal queue. The withdrawal queue limits the amount of withdrawals (full or partial) to 16 transactions[1] per block, meaning 115,200 withdrawals can be processed per day. The exit queue limits the amount of validators exiting to 1,800 per day. However, this number is dynamic based on the amount of active validators. As active validators decrease, the amount of validators that can be processed through the queue also decreases and can be as low as 900. Once processed through the exit queue, there is a 27 hour delay before receiving your ETH deposit. All of this is designed to maintain stability in the network.

[1] A validator withdrawing partial rewards counts as 1 transaction, and a validator requesting a full withdrawal also counts as 1 transaction.


Estimating Sell Pressure


A lot of analysis has been put into who might be entering the withdrawal queue once Shapella goes live on April 12th, and, while interesting, we find most analyses miss the key point. As an ETH holder, you should only focus on who might unstake AND liquidate their ETH position. There will be ample amounts of unstaking come April 12th, but we believe most will be the result of early stakers adjusting their operational setup and restaking, thus causing no net change in their ETH position. These could be solo stakers who are running a validator themselves who now want to outsource that to a staking pool or centralized exchanges migrating their customers’ tokens from one staking-as-a-service provider to another. The point being: Shapella’s liquidity event will not result in these unstakers  liquidating their positions. While we believe this to be true for the majority, it is definitely not the case for all.

Partial withdrawals are the first factor to consider when trying to estimate ETH sell pressure. Partial withdrawals will be immediately liquid, as the network will require any balance over 32 ETH to be withdrawn. The validator must specify a withdrawal address for the excess balance to be sent to. This is public data since it is an on-chain transaction, and we can observe how many stakers have specified an address for their withdrawals. As of April 1st, it was 43% of validators. we would expect this number to rise closer to 80%+ by April 12th. Now, of the eventual 80%, we would expect many are withdrawing rewards to restake and compound their staking yield, but some will certainly sell. In our base case, we are expecting 50% to restake and 50% to convert their rewards to cash or use as on-chain ‘working capital.’

To estimate full withdrawals we will need to make projections about the expected behavior for each staking operator. Again, with the benefit of staking being on-chain we can identify the operators of most validators to see their total position size.


Ethereum Staking Depositors

Source: Dune Analytics (@hildobby).


Some names should stick out as likely candidates to be unstaking and liquidating positions:

  • Celsius – Due to their bankruptcy, it should be expected that they will unstake 100% of their position and liquidate the ETH into USD.
  • Kraken – They were fined $30 million and required to shutdown their US staking operations by the SEC. Most of their 1.2 million ETH should end up restaked either through their international entity or another provider. However, the aggressive stance taken by the SEC may result in some ETH being sold off. Conservatively, we will model 10% of Kraken’s ETH being sold due to regulatory concerns.
  • Coinbase – The SEC issued a Wells Notice to Coinbase and identified staking as one area they believe to be operating illegally. Coinbase wrote a fantastic rebuttal to the SEC and maintains all of their services are operating within the ruleset. We don’t expect significant unstaking from Coinbase, but, for the same regulatory concerns, we will model above average sell activity from their customers.
  • Binance – Like Kraken and Coinbase, Binance is dealing with potential regulatory infractions in the United States. The allegations against Binance are primarily surrounding their unwillingness to offboard US users from their offshore products that were not licensed in the US. We believe Binance was reluctant to do so but did eventually ban US-based users from accessing the offshore exchange. We don’t believe their users have the same levels of concerns of those at Kraken and Coinbase. Still, we will take a conservative approach that they may also see slightly above average selling pressure.


For the customers of other centralized exchange providers, solo stakers, and staking pools, there is less of a reason to sell your staked ETH unless you are liquidating all of your holdings. Using the on-chain data, we can observe that most depositors (~75%) are underwater on the ETH they staked since the price is below its level on the date of their deposit. So they are not sitting on large unrealized gains. For this entire cohort, we will estimate a small amount of sell pressure at a base-rate of 5% to account for those who are forced sellers for unspecified reasons.


The last set of stakers to be accounted for are the liquid staking protocols, of which Lido and Rocket Pool are the largest. The entire premise of these protocols is they offer and have offered immediate liquidity through a derivative token (stETH for Lido and rETH for Rocket Pool) that serves as a receipt of their staked ETH deposit. These protocols relied on the growing DeFi utility for the derivative token to gain adoption and acceptance from the market. Aside from a few hiccups, this was mostly successful, and both stETH and rETH trade close to par with ETH. Since stakers on these platforms have the ability to sell their receipt token that is required to access the staked ETH, they already have liquidity for their staked positions. We will take the position that anyone using these will not use Shapella as a liquidity event and maintain their current ETH position.


With all of this put together, we can finally estimate an expected total sell pressure as a result of the Shapella upgrade.


Projected Sell Pressure from Shapella Withdrawals

Sources: Runa Digital Assets, Dune Analytics, Blockworks


If our base case unfolds and 1.3 million ETH ends up changing hands, it will only represent just over 1% of the total supply of ETH. Our worst case only leads to 2% of supply hitting order books. There is not much to worry about from our view. If 1% of supply were to hit the market in one day, it may cause some volatility, as this would represent about 25% of the average daily volume of ETH on exchanges (both centralized and decentralized). However, we can be assured that this won’t happen due to the withdrawal and exit queues. As the analysis shows, the withdrawal and exit queues would create a 14+ day backlog to process all of the withdrawals. We expect the queues to be much longer, as we must also include stakers who are exiting with plans to restake using another provider. To do so, they will have to fully exit the queue before rejoining. This will extend the exit queue beyond 14 days and likely to 30+ days. This should give any ETH owner reassurance that they should not expect material volatility solely due to the Shapella upgrade.


The ETH staking yield is also worth mentioning here, as it is directly linked to the amount of stakers on the network. Currently the 15% of ETH holders who stake are earning a ~5.5% yield in return for validating transactions. However, the yield moves lower with each additional staker, as the rewards are split among all of them. In our original research, we wrote that we expected the staking participation rate to begin to rise once withdrawals were enabled and reach a steady state around 50%. Our analysis shows that this would send the yield to between 2.5% – 3.5%. This may be a deterrent for a small number of holders who are primarily attracted by the existing above-market yield compared to other rates currently available in DeFi. We don’t expect this to cause any immediate sell pressure because the higher staking rate will take months to achieve. It is likely the staking rate will remain unchanged at 15% for the first 1-2 months, as both the entry and exit queues will be at full capacity, making the validator queue 1 in 1 out in the weeks after Shapella.

ETH Yield Curve[1]

Source: Runa Digital Assets.
[1] Based on Runa’s internal projections. Inputs include total ETH supply, amount of ETH staked, transactions fees, priority fees, MEV.

Final Takeaways

Shapella is unquestionably a positive upgrade and achievement in Etheruem’s long term roadmap, but we found it worthwhile to prepare ourselves for a number of outcomes including potential ETH sell pressure. Our ultimate view is it will not be consequential on its own, and macro factors are more likely to contribute to price action over the next 1-3 months. Perhaps an unexciting conclusion depending on your exposure to ETH.

We do believe Shapella has the potential to be a bigger catalyst to other tokens though. Liquid staking derivatives could be the biggest beneficiary. While they offer liquidity today through their respective derivative tokens, it can be rather spotty liquidity and disappear in times of serious market stress. In June for example, Lido’s stETH was trading at a 10% discount to ETH, as markets went through a violent deleveraging process. Shapella will bring native liquidity in allowing for derivative tokens to be unstaked and redeemed at par regardless of market conditions. We believe these protocols will continue to amass market share among staking providers, as centralized parties offer limited DeFi utility and confront potential regulatory headwinds.

Another sub-sector that is already trending and we think will soon benefit from Shapella is Layer 2s like Arbitrum and Optimism. While not directly tied to Shapella in any way, they will be gaining developer mindshare in the weeks ahead. Shapella is the final upgrade in the Proof-of-Stake migration and completes an amazing feat for Ethereum developers in making what is the equivalent of a mid-flight engine swap. Their attention will now turn to the scaling roadmap, the first stop being the EIP-4844 upgrade. This upgrade delivers ‘proto-danksharding’, a new transaction type that will drastically reduce fees for Layer 2 networks who use Ethereum to store their data. Activity on Layer 2 networks has been rising for months, and a further reduction in transaction fees should provide a catalyst for more activity. Other Ethereum-based upgrades like account abstraction will improve the user experience, and Layer 2 networks will be uniquely positioned to offer a better user experience at a lower cost.


Daily Active Addresses for Ethereum and Layer 2s (10/1/2022 – 4/1/2023)


The momentum in the Ethereum ecosystem is hard to ignore, and Shapella provides another demonstration of the ability to set a roadmap and execute on the plan. Sure, it may have taken longer than expected, but the upgrade process has been a smooth one so far. As they say in racing, ‘slow is smooth, smooth is fast.’ Ethereum sits comfortably in the driver seat as the dominant Layer 1 blockchain but knows scaling upgrades are needed to stay on track.


This article (i) is only for informational and educational purposes, (ii) is not intended to provide, and should not be relied upon, for investment, accounting, legal or tax advice, and (iii) is not a recommendation as to any portfolio, allocation, strategy or investment. This article is not an offer to sell or the solicitation of an offer to buy any securities or other instruments.