What is Staking?
With the advent of the highly anticipated Merge, Ethereum will transition its consensus mechanism from Proof of Work (PoW) to Proof of Stake (PoS). The Merge to PoS is widely regarded as the most significant change to Ethereum’s technical architecture since Ethereum mainnet launched in 2015. The full technical details and benefits of this upgrade are out of scope for this article, but have been written and spoken about extensively. For readers seeking a deeper understanding, Ben Edgington is in the process of writing a comprehensive, technical handbook on Ethereum’s move to PoS and beyond.
In this article, we speedrun through the basics of Ethereum staking ahead of the Merge. We provide an overview of the various ways users can stake their ETH, and highlight each option’s benefits and drawbacks. Before getting into the details, let’s start by looking at what Ethereum staking really means.
At a high-level, permissionless systems like Ethereum require a mechanism by which all system participants (i.e. nodes on the network) reach agreement on a single source of truth. In Ethereum today, consensus about the state of the blockchain is achieved amongst miners who participate in PoW. Following the Merge, miners will be replaced by validators, and PoW will be replaced by PoS. This change confers many benefits to Ethereum, including stronger settlement finality guarantees, the enablement of sharding, and drastically reduced ETH issuance and energy consumption. From an Ethereum user’s perspective, this change presents a new opportunity for financial gain in the form of staking.
Ethereum holders have been able to stake their ETH since the launch of the Beacon Chain in December 2020. Staking on Ethereum involves depositing, or “locking up”, 32 ETH to activate validator software, which effectively adds another validator (i.e. node) to the Ethereum network. Validators help Ethereum reach consensus by proposing new blocks, or verifying and attesting to the validity of blocks proposed by others. As more honest validators participate in the network, the cost of attacking the network increases, thereby improving Ethereum’s economic security.
Notably, after a user deposits 32 ETH to activate a validator, their deposit can not currently be withdrawn. In fact, withdrawals of staked ETH will not be enabled until the Shanghai upgrade occurs, which is the first fork planned to follow the Merge.
Despite the opportunity cost of non-immediate withdrawals, users remain incentivized to stake their ETH and activate a validator since validators earn financial rewards for carrying out their assigned duties; proposing and validating blocks. These rewards come from new ETH issuance and, after the Merge, priority fees from transactions and maximal extractable value (MEV), both of which currently go to PoW miners.
Rewards are proportional to validator uptime, meaning that a validator that’s always kept online and performing its validator duties will earn maximal rewards. On the other hand, validators that go offline or fail to perform the duties expected of them incur a small penalty. Beyond these minor offenses, there are some validator behaviors which break specific protocol rules and could constitute an attack on the Ethereum network. Validators that exhibit these behaviors are subject to a more severe punishment known as “slashing”. A validator that’s been slashed loses a larger portion of its 32 ETH and is ejected from the network of active validators. This balance of rewards and penalties creates a financial incentive structure that motivates those who operate validators to act in the best interests of the Ethereum network.
In short, users who stake their ETH to activate a validator help the chain reach consensus, contribute to Ethereum’s economic security, and earn financial rewards on the ETH that they stake. At the time of writing, Ethereum users have collectively staked over 14.2M ETH, or ~$23.6B at current prices, and are earning up to ~4.1% in annual returns on their stake.
Types of Ethereum Staking
Those who stake their ETH tend to do so via one of four primary routes:
Solo staking allows you to earn the maximum rewards possible on the ETH that you stake, eliminating any rent-seeking intermediaries and providing you with complete control over your funds and validator setup at all times. However, solo staking requires a minimum of 32 ETH, greater technical proficiency, and generally more effort than other forms of staking.
Solo staking means running an Ethereum node independently, which involves:
- purchasing and maintaining hardware,
- running and periodically upgrading three pieces of software (an execution client, a consensus client, and a validator client),
- generating and securely storing validator signing keys (used to sign on-chain operations) and withdrawal keys (used to withdraw one’s staked ETH once this becomes available),
- maintaining a fast and reliable internet connection, and
- staying aware of any network or critical client upgrades (i.e. you’ll need a Discord and/or Twitter account).
There has been a tremendous effort by Ethereum developers and the broader community to develop resources that make solo staking as accessible as possible, regardless of one’s technical ability. While solo stakers reap the highest possible rewards on their staked ETH and avoid substantial counterparty risk by self-custodying their funds and keys, they do assume the risks of key management, as well as technical and operational risks that could result in penalties or slashing.
Staking as a Service (SaaS)
For users who have enough ETH to fund a full validator (32 ETH), but lack the technical knowhow, time, or desire to operate an Ethereum node independently, staking-as-a-service may be a particularly attractive option. In this arrangement, a staking provider typically supports users through the process of depositing 32 ETH (or more) and generating their validator’s signing and withdrawal keys. Like in solo staking, the user typically custodies their withdrawal keys themselves, meaning that once withdrawals are enabled, they are the only ones who can withdraw their original 32 ETH deposit and the staking rewards earned by their validator (less the fees taken by their provider). However, unlike solo stakers, their validator’s signing keys are custodied by their staking provider.
The provider operates validator hardware and software on the user’s behalf, and uses their signing keys to perform the validator’s on-chain duties like block proposals and attestations. Typically, staking providers charge a recurring monthly fee for this service. This fee is commonly taken as a percent of the rewards earned by each validator, and is variable depending on the amount of ETH a user has to stake. While the fees charged by providers tend to be competitive, this service remains financially inaccessible to some users since, like solo staking, a minimum of 32 ETH is required (~$53K USD at the time of writing). As additional value-adds, some providers offer their clients insurance against losses from slashing, and/or a dashboard to track the value of their stake and other metrics related to validator performance.
Compared to solo stakers, those who stake with a provider personally assume much less technical and operational responsibility for their validator. However, shifting this responsibility to a provider does expose these users to greater third-party technical and operational risks. For example, they must trust their provider to custody their validator’s signing keys, consistently perform validator duties on their behalf, and avoid any technical issues as well as regulatory and cybersecurity threats.
There are a number of providers who offer these services, including ConsenSys’ Codefi Staking, which leverages our leading industry position and deep Ethereum expertise to help institutions achieve optimized staking rewards while eliminating risk and complexity. Each staking provider’s offering is slightly different, and the costs, benefits, and risks of each should be carefully considered by potential stakers.
Staking pools lower both the financial and technical barriers to entry for ETH staking. They provide a simple way for users who hold less than 32 ETH to combine their deposits with others, collectively fund a new validator each time the pool’s value exceeds 32 ETH, and earn staking rewards in proportion to their deposit. From the user’s perspective, this makes staking ETH as simple as signing a single transaction. The simplicity and convenience that staking pools offer has made them an immensely popular option amongst users. At the time of writing, more than one third (~33.5%) of all staked ETH is being staked via pools. Since pooled staking is not supported natively by the Ethereum protocol, a number of distinct pooling solutions – each with their own benefits and drawbacks – have been developed to meet user demand.
While staking pools can be mediated off-chain, most are implemented on-chain via smart contracts that accept, pool, and disburse user deposits into 32 ETH portions to fund new validators. Smart contracts are used to track how much ETH each user deposited and calculate their share of the pool’s earned staking rewards. As of early 2021, it also became possible to set withdrawal credentials to a smart contract address, rather than a BLS digital signature. This change effectively reduced pool depositors’ counterparty risk around withdrawals, since withdrawal keys no longer had to be custodied by a single, trusted key holder (or multiple key shard holders, in the case of a multisig.). Instead, withdrawal credentials could point to a smart contract that is upgradeable by a less centralized entity (e.g. by DAO governance).
Pooled staking is similar to staking-as-a-service in that the technical complexity of operating validator hardware and software is outsourced to someone besides the staker. This responsibility falls to the pool’s ‘validator set’. Validator sets are typically composed of multiple, independent validator operators (often the same providers that offer staking-as-a-service) who handle the technicalities of staking on behalf of the pool and its depositors.
Validator sets can be either: permissioned, meaning that only those with permission from some entity like the pool’s DAO can operate the pool’s validators, or permissionless, meaning that anyone who meets some minimum standard can operate the pool’s validators. Since validator rewards are proportional to validator uptime, the pool is incentivized to exclusively onboard high-quality validator operators (in the case of a permissioned validator set), or create mechanisms that incentivize validator operators to perform to the best of their abilities (in the case of a permissionless validator set).
A notable differentiator between pooled staking and other staking models is that no single user deposit is tied directly to any single validator in the pool’s validator set. In solo staking and when staking via a provider, the relationship between the staker and validator can be thought of as a one-to-one relationship (a single user with 32 ETH funds a single validator). In pooled staking, the relationship between stakers and validators can be thought of as a many-to-many relationship (multiple users’ funds are distributed to the multiple validator operators that make up the validator set).
This many-to-many relationship means there is no traceability between a user’s deposit and the validator(s) which their deposit is directed towards. Due to this lack of traceability, all of the staking rewards earned, and penalties incurred, by the pool’s validators must be mutualized across all pool depositors. The mutualization of rewards and penalties means that no single user will benefit more than others if one of the pool’s validators earns outsized rewards, and similarly, no single user will suffer disproportionate losses if one of the pool’s validators is penalized. From the validator operators’ perspective, since the rewards they earn are also mutualized across the validator set, they don’t receive outsized rewards for above par performance. Therefore, a pool’s validators are not necessarily incentivized to outperform their validator peers, but rather, to perform as well as the median validator such that they avoid scrutiny by the pool and continue receiving user deposits.
From a cost standpoint, staking pools tend to charge users slightly higher fees than those who stake 32 ETH directly with a staking provider. This higher fee is justified by the additional technical and operational overhead required to pool user deposits, manage or incentivize the pool’s validator set, and handle the potentially messy process of withdrawing pooled funds once that becomes available. Since multiple, high-quality pooling solutions exist and are competing for user deposits, staking pool fees still tend to be competitive and worthwhile for users seeking simplicity and convenience.
Staking pool depositors personally assume virtually none of the technical or operational risk associated with operating a validator, and they aren’t required to custody either signing or withdrawal keys. However, like those who stake with a provider, they assume third-party and counterparty risk since they must trust the pool’s validator set to manage signing keys, perform validator duties, and avoid technical, regulatory, and cybersecurity issues. Further, depositors to smart contract-based pools assume additional risks around the security of the smart contracts that enable the solution, and must trust whatever entity and process are employed to upgrade any upgradeable smart contracts (e.g. DAO governance vote).
Liquid Staking Pools
A popular subset of staking pools offer what’s known as “liquid staking”, which provides pool depositors with an ERC-20 token that represents their staking position. The ERC-20 tokens that these liquid staking pools issue are called “liquid” staking derivatives (LSDs) since they provide the staker with a limited form of liquidity on their staked position that they otherwise wouldn’t have (until withdrawals are enabled via the Shanghai upgrade). Beyond providing exit liquidity from staking positions, LSDs can be used across DeFi as collateral assets, so holders can earn additional yield on top of staking rewards. The advantages of tokenizing a user’s stake are undeniable, and this is reflected by the fact that over 97% of all user deposits to staking pools are to pools that offer LSDs.
These tokens are implemented quite differently by different liquid staking pools. Broadly, the user receives a token that represents their initial deposit and, over time, either the token supply (under the aToken model) or its value relative to ETH (under the cToken model) changes dynamically to reflect the user’s earned staking rewards. These implementation differences mean that staked ETH tokens from different pools are not fungible with each other.
Naturally, users are incentivized to adopt the token with the most liquidity, creating strong network effects and “winner-take-most” dynamics. This has played out in the liquid staking market as a single liquid staking protocol, Lido, has captured over 90% of the liquid staking market, and staggeringly, nearly one third (~31%) of the entire ETH staking market as of the time of writing. Lido’s dominance has become a controversial topic in recent months as a potential vector for centralization that could reduce Ethereum’s censorship-resistance and increase the viability of various attacks on the network. These risks have raised concerns from prominent members of the Ethereum community and prompted in-depth DAO governance discussions about Lido’s position in the market. By contrast, Ethereum’s second most popular liquid staking pool by TVL, Rocket Pool, has made different implementation decisions that mitigate or eliminate many of these risks. Similarly to Lido, Rocket Pool offers a liquid staking token of their own, rETH.
Centralized Exchange Staking
Lastly, many centralized exchanges offer Ethereum staking to their users. These offerings tend to resemble a centralized and custodial version of the staking pools described above. Since centralized exchanges custody assets on behalf of their users, they can easily pool the assets of users who opt into staking, and run large numbers of validators on behalf of their users.
From the user’s perspective, staking via a centralized exchange is as simple and convenient as depositing to a staking pool. Some exchanges offer users a limited form of liquidity on their staking position, similar to an LSD from a liquid staking pool. However, rather than providing an ERC-20 token that can be swapped on multiple exchanges and used across a variety of DeFi protocols, most centralized exchanges offer a token which must stay within the exchange, and can only be swapped for ETH in a liquidity pool provided by the exchange and its users. Coinbase recently bucked this trend with the launch of cbETH, a wrapped version of their on-exchange liquid staking derivative, ETH2. The cbETH token provides users who stake via Coinbase with all the utility of an LSD from a staking pool, albeit at a higher fee on the underlying staking position compared to the commission taken by the most popular staking pools.
Fees taken by centralized exchanges vary significantly, ranging from no fee (presumably as a user acquisition strategy, or to increase liquidity for their on-exchange LSD), up to 25% of all earned staking rewards (roughly double the fees charged by the most popular staking pools).
Like staking pool depositors, users who stake through a centralized exchange personally assume virtually none of the technical and operational risks of operating a validator. However, these users are exposed to a greater degree of counterparty risk since they entrust the exchange with custody of validator signing and withdrawal keys, the operation of the validator(s) they’ve funded, avoidance of technical, regulatory, and cybersecurity risks, and critically, custody of their funds. Beyond risks to individual users, centralized providers consolidating and staking large amounts of ETH creates large, centralized targets and points of failure. In a worst case scenario, this could make the Ethereum network itself more susceptible to attacks, bugs, or regulatory capture.
In this article, we’ve provided a brief overview of key Ethereum staking concepts and taken a closer look at the four primary ways that Ethereum users stake their ETH: solo staking, staking via a service provider, pooled staking, and centralized exchange staking.
For users with more than 32 ETH, solo-staking is considered to be the gold standard of Ethereum staking and confers the most benefits to both the staker and the Ethereum network. Staking-as-a-service offers those with more than 32 ETH a simplified user experience in exchange for a cut of earned staking rewards. For users with less than 32 ETH, staking pools take relatively low fees, often provide users with liquidity on their stake, and can be architected in such a way that they improve Ethereum’s decentralization. Regardless of how a user stakes, their participation in Ethereum’s consensus layer provides a personal, financial benefit as well as a public good; supporting the world’s open and permissionless financial infrastructure.
For regular Ethereum roadmap updates, Ben Edgington’s bi-weekly update is a great place to learn. For a more focused update on the state of Ethereum staking, Kuhan Tharmananthar writes regular blog posts about the Ethereum staking ecosystem and the exciting work being done by Codefi Staking. Lastly, ethereum.org is a reference cited throughout this article, which contains succinct information and links related to all things Ethereum.