Institutional staking is often presented as a choice between earning a protocol yield and leaving digital assets idle. That framing is incomplete. Staking exchanges immediate control or liquidity for participation in a blockchain’s security mechanism. The resulting return can combine token issuance, transaction-related income and other block-production revenue, while the risks can sit with the protocol, validator, custodian, staking pool, governance system or asset owner.
The distinction matters because a quoted annual percentage rate is not equivalent to a bond coupon or cash interest rate. It is normally a variable return paid in the network’s native token. Its economic value depends on token supply growth, market price, validator performance, commissions, exit timing, taxes and the institution’s ability to recognise, safeguard and dispose of the rewards.
This is not an argument for or against staking. It is a framework for locating the reward and the corresponding risk before an institution commits assets.
Start by decomposing the headline yield
A proof-of-stake network compensates validators for performing protocol duties. On Ethereum, those duties include timely attestations, block proposals and participation in sync committees. The protocol’s base reward is related to a validator’s effective balance and inversely related to the square root of total active stake. As more ETH is staked, the consensus-layer reward available per validator generally falls. The current formula and duties are documented in Ethereum’s proof-of-stake rewards and penalties documentation.
Rewards may have several components:
- Protocol issuance: newly created tokens distributed under the network’s monetary rules. This increases the recipient’s token balance but may dilute holders who do not stake.
- Transaction income: the portion of transaction fees directed to validators. On Ethereum, the base fee is burned while the priority fee is paid to the block proposer.
- Block-production opportunities: income associated with ordering or selecting transactions, commonly described as maximal extractable value, or MEV. Ethereum defines MEV as value extracted beyond standard block rewards and gas fees through the inclusion, exclusion or ordering of transactions in a block.
- Provider incentives: temporary rebates, promotional rates or tokens supplied by an intermediary. These are not necessarily generated by validation and should not be blended into the protocol return.
The institution should therefore reconcile each quoted yield to an on-chain source. If a provider cannot separate protocol issuance, transaction income, MEV, subsidies and its own fee, the yield cannot be assessed or compared reliably.
Nominal yield is not necessarily economic return
Suppose a network distributes new tokens primarily to stakers. A holder that stakes may preserve more of its percentage ownership than a holder that does not, but part of the apparent yield compensates for monetary dilution. The relevant comparison is not simply staking income versus zero. It is the staker’s net token growth relative to supply growth, operating costs, fees and the return available from retaining liquidity.
A useful internal bridge begins with gross protocol rewards and deducts validator commission, custody and orchestration fees, downtime penalties, expected slashing loss, transaction costs and taxes. It should then distinguish token-denominated return from base-currency return. A positive return in ETH, SOL or another token can coexist with a loss measured in sterling, euros or dollars.
Revenue can also be uneven. Block proposals and MEV are probabilistic, whereas routine consensus rewards are more regular. An institution comparing validators should examine realised results over equivalent periods and stake sizes, not a provider’s best epoch or an annualised snapshot.
Validator economics shape the service
A validator operator pays for infrastructure, monitoring, engineering, security and, on some networks, recurring protocol participation costs. It earns a commission or retains an agreed share of rewards. Low commission is not automatically the lowest-risk choice if it is unsupported by a durable operating model.
On Solana, the official staking documentation states that staking yield depends on the inflation rate, the proportion of SOL staked, validator uptime and validator commission. Rewards are calculated and issued each epoch. Solana also warns that an idealised yield estimate omits the effects of uptime, commission, throttling and potential slashing incidents. This illustrates why economics must be assessed network by network.
Due diligence should ask whether the operator can remain profitable through lower token prices, declining issuance or rising infrastructure costs. It should also identify whether the operator relies on foundation delegations, related-party stake or incentives that can be withdrawn. A validator whose economics depend on a temporary subsidy presents a different continuity risk from one funded by diversified, fee-paying delegators.
Slashing is a tail risk, not the whole risk
Slashing destroys or withholds stake when a validator commits specified protocol offences. The rule is not uniform across networks, and neither is the allocation of loss between operator and delegator.
Ethereum can slash a validator for proposing two blocks for one slot, double voting, or making an attestation that surrounds another attestation. Its current documentation describes an initial penalty, forced exit and a correlation penalty that increases when many validators are slashed in the same period. In an extreme correlated event, the loss can reach the validator’s full effective balance. Ordinary inactivity is different: missed duties reduce rewards and incur penalties, while a prolonged failure of finality activates an inactivity leak that progressively reduces balances of inactive validators.
The important institutional question is not only whether a provider has ever been slashed. It is whether one software bug, cloud outage, signer error or failed migration could affect many of the institution’s validators at once.
Duplicate signing can arise during operational change, not only deliberate misconduct. EIP-3076, created on 27 October 2020 and currently marked Last Call, specifies the exchange of slashing-protection histories between Ethereum validator clients because moving a key without its signing history can cause a new client to sign a conflicting message.
Controls should include a single authoritative signing path, transferable slashing-protection data, rehearsed migration procedures, separation between validator and withdrawal credentials, and explicit prohibitions on simultaneously activating the same validator key in primary and disaster-recovery environments.
Correlation can dominate individual operator quality
Diversifying across three validator companies does not reduce risk if all three use the same consensus client, cloud region, networking provider or key-management dependency. Ethereum’s penalty design makes this especially important because correlated failures can be more costly than isolated ones.
The Ethereum Foundation’s current distributed validator technology documentation explains how a validator key can be split among multiple nodes so that a threshold can continue signing without any one node holding the complete online key. DVT may reduce single-machine and single-operator exposure, but it introduces middleware, coordination and implementation risks of its own. It should be assessed as another system, not treated as a guarantee.
Institutions should map concentration by legal entity, client software, data centre, cloud provider, geography, signer technology, relay or block-building dependency and incident-response team. Nominal provider diversification is weaker than dependency diversification.
Lock-up is a queue, market and governance problem
Staked assets may be subject to activation, unbonding, exit or withdrawal delays. These are protocol rules rather than contractual maturity dates, and the realised delay may change with network demand.
Ethereum enabled consensus-layer withdrawals through the Shanghai and Capella upgrade on 12 April 2023. Exiting still requires a validator to pass through protocol queues, and Ethereum’s withdrawal documentation states that the timeline depends on network demand. The Pectra upgrade, activated on 7 May 2025, added execution-layer-triggered exits through EIP-7002, allowing the withdrawal address to initiate an exit without the validator signing key. Provider workflows can nevertheless add approval, batching or settlement delays.
Liquidity planning should distinguish four timestamps: the instruction to unstake, protocol exit eligibility, withdrawal to the controlled address, and availability for sale or transfer. Stress testing should assume that many participants seek to exit during market disruption, when protocol queues lengthen and intermediaries may impose their own controls.
Liquid staking relocates rather than removes illiquidity
A liquid staking token, or LST, is issued in connection with staked assets, but the rights it evidences depend on the particular arrangement. It may be transferable or usable while underlying assets remain staked. The arrangement can exchange a direct waiting-period constraint for smart-contract, provider, redemption, secondary-market and governance risks.
Ethereum’s liquid and pooled staking documentation identifies slashing pass-through, smart-contract defects, market discounts, redemption congestion, operator concentration and governance changes among the risks. The Ethereum protocol pays its validators. It does not guarantee the value, convertibility or legal character of an LST issued above it.
An LST should not be assumed to equal the underlying token plus accrued staking rewards. Its contract rights, exchange ratio, redemption mechanics and secondary-market price can differ from the position obtained by staking the native asset directly.
An institution should document who owns the underlying asset, what the token legally evidences, how rewards and losses are allocated, whether redemption is contractual or purely programmatic, who can upgrade the contracts, and what happens during a fork, exploit, sanction event or insolvency. Collateral use adds liquidation and composability risk on top of staking risk.
Custody must be analysed as authority, not branding
“Non-custodial staking” can conceal several distributions of authority. The institution may retain the withdrawal key while an operator controls an online validator key. A custodian may control both. A smart contract or multisignature committee may govern withdrawals. The decisive question is which party can move principal, redirect rewards, initiate exit, change beneficiaries or exercise protocol rights.
The European Union’s Markets in Crypto-Assets Regulation, Regulation (EU) 2023/1114, has applied fully since 30 December 2024. For in-scope custody and administration, Article 75 requires client agreements, position registers, custody policies, segregation, return procedures and liability for attributable losses, subject to its terms. MiCA does not make every staking arrangement safe, nor does its list of crypto-asset services expressly classify staking as a standalone crypto-asset service. A January 2025 joint EBA and ESMA report identified custody risks in staking, including custodian failure, commingling, inadequate records and uncertainty about who may claim the assets.
Contracts should allocate slashing, downtime, erroneous reward distribution, forks, airdrops, insolvency, subcontracting and key compromise. Any indemnity should be tested against caps, exclusions, creditworthiness and the time required to collect. “Slashing protection” is only as valuable as its definition and provider balance sheet.
Governance rights can move while assets are staked
Staking may carry voting power or influence over validator selection and protocol governance. Delegating stake can therefore delegate more than technical work. An intermediary may vote, abstain, follow a policy or pass voting instructions through to the beneficial owner.
The mandate should state who exercises governance rights, how conflicts are handled and whether the institution can opt out. It should also address contentious forks and emergency upgrades. A liquid-staking protocol adds another governance layer whose participants may be able to change fees, operator admission, oracle design, redemption logic or contract implementation.
Concentration is both a portfolio and network concern. A large provider can be operationally efficient while increasing common-mode failure, censorship or governance influence. Provider selection should therefore consider the institution’s contribution to aggregate stake concentration, not only its own bilateral exposure.
Regulatory status depends on facts and jurisdiction
In the United States, the SEC Division of Corporation Finance issued a staff statement on certain protocol staking activities on 29 May 2025. It expressed the staff’s view that the protocol staking activities described in the statement do not involve the offer and sale of securities. A second staff statement on certain liquid staking activities, dated 5 August 2025, reached a similar view for a defined fact pattern involving staking receipt tokens.
Both documents expressly state that they are staff views, not rules or Commission statements, have no legal force, and depend on the specified facts. The liquid-staking statement excludes restaking and indicates that arrangements involving additional managerial decisions or return-generating services may fall outside its scope. Commissioner Caroline Crenshaw published dissents on both dates, underscoring that the legal interpretation was not institutionally unanimous.
An institution should not export those US staff views to the UK, EU, UAE or another jurisdiction, or assume that protocol staking, a managed programme, an LST and a leveraged restaking strategy share one legal classification. Product design, marketing, custody, discretion, client type and cross-border distribution can change the analysis.
Tax follows receipt, control and transaction structure
Tax treatment is not established by calling a return “yield”. In the United States, IRS Revenue Ruling 2023-14, published in Internal Revenue Bulletin 2023-33 on 14 August 2023, states that a cash-method taxpayer includes the fair market value of native-token staking rewards in gross income when it gains dominion and control over them. The ruling also applies to its stated exchange-staking fact pattern. It does not answer every question involving blocked rewards, LSTs, non-US entities or more complex arrangements.
For UK individuals, HMRC’s Cryptoassets Manual CRYPTO21200, updated on 28 November 2025, says that whether staking is a taxable trade depends on activity, organisation, risk and commerciality. If it is not a trade, the pound sterling value of awarded tokens at receipt is taxable as miscellaneous income, subject to appropriate expenses, and a later disposal may attract Capital Gains Tax. This individual guidance does not itself determine the treatment of a company, fund or other institutional vehicle.
Operationally, this requires timestamped records of reward entitlement, receipt, control, quantity, token identity, wallet, valuation source, fees and subsequent disposal. Auto-compounding does not eliminate the need to determine whether a taxable recognition event occurred.
Accounting contains important boundaries, not a universal staking rule
Under US GAAP, FASB Accounting Standards Update 2023-08 applies fair-value measurement, with changes recognised in net income, to crypto assets meeting its scope criteria. It became effective for fiscal years beginning after 15 December 2024, including interim periods. The update addresses qualifying crypto-asset holdings, but it does not provide a complete, staking-specific answer for reward recognition, validator services, LSTs or arrangements involving contracts and intermediaries.
Under IFRS, the IFRS Interpretations Committee’s June 2019 agenda decision addressed a defined subset of cryptocurrencies: digital or virtual currencies recorded on a distributed ledger, secured by cryptography, not issued by a jurisdictional authority or other party, and not giving rise to a contract with another party. It concluded that IAS 2 applies when such cryptocurrencies are held for sale in the ordinary course of business and IAS 38 applies otherwise. That decision concerned holdings of cryptocurrencies, not every staking transaction. Questions about when rewards are recognised, whether an LST is the same unit of account as the deposited token, and whether control has been transferred require analysis of the actual rights and applicable standards.
Before staking, finance teams should approve the unit of account, recognition point, valuation hierarchy, principal-versus-agent conclusion, presentation of rewards and fees, impairment or fair-value treatment, and disclosures. For a material programme, those judgements ordinarily need to be resolved with the reporting entity’s accounting advisers and auditor before the first affected reporting date.
An institutional control framework
A defensible staking programme connects each risk to an owner, limit and evidence trail.
- Mandate: define permitted networks, staking forms, counterparties, LSTs, governance activity and prohibited leverage or restaking.
- Reward verification: reconcile gross on-chain rewards to commissions, penalties, MEV policy, provider statements and controlled wallets.
- Key separation: segregate withdrawal, validator and governance authority; require quorum controls and tested recovery.
- Validator resilience: set limits by operator, software client, cloud, region and signer stack; review correlated dependencies.
- Slashing controls: require signing-history protection, migration runbooks, incident reporting and quantified contractual loss allocation.
- Liquidity: model activation and exit queues, intermediary processing, LST discounts and stressed disposal capacity.
- Custody and legal: verify asset ownership, segregation, insolvency treatment, subcontractors, liability and governing law.
- Governance: pre-authorise voting, fork and upgrade responses, conflicts and escalation thresholds.
- Tax and accounting: preserve event-level data and obtain jurisdiction-specific conclusions before launch.
- Monitoring: track realised net yield, uptime, penalties, concentration, contract upgrades, regulatory changes and exceptions against limits.
For an investment committee, the relevant comparison is net, risk-adjusted economics and operational fit, rather than the largest displayed rate. Staking can produce native protocol income, but every additional layer promising convenience, liquidity or enhancement also changes where control and loss sit. The institutional task is to make that transfer visible, priced and governable.
This article is for editorial and informational purposes only. It is not investment, legal, tax or accounting advice.