Liquid staked tokens (LSTs) are ERC20 tokens that represent a claim on underlying staked assets and their future yield. These assets can be the native tokens of any blockchain that supports staking as a consensus mechanism, but for the purposes of this article, we will focus on Ethereum.

Ethereum transitioned from a proof of work chain to proof of stake in September 2022 (in a historic event called The Merge). To participate in Ethereum's proof-of-stake consensus protocol, validators must lock up a minimum of 32 ETH (~$72,000 at the time of writing) and operate an ethereum node that validates and orders transactions on the ethereum network. In general, however, these requirements are both too capital intensive and technically complex for the average ETH holder. Only around 24% of the ETH in circulation is currently staked[1].

In order to bridge this usability gap, LST protocols offer ETH holders the ability to earn staking yield simply by depositing ETH to an on-chain smart contract or exchange-run service, and receiving a transferrable ERC20 token in return. This ERC20 token (the LST) both represents the value of the underlying staked ETH and accrues staking rewards on behalf of the user. A unique advantage of LSTs is that these tokens are both liquid — freely swappable for other assets or transferred to other addresses, and also composable — able to be utilized in any DeFi protocol that accepts ERC20 tokens.

This ease-of-use starkly contrasts against operating a validator, in which staked ETH is both locked up and subject to a potentially lengthy withdrawal queue when an exit is required. In practice, the LST protocols themselves manage these underlying deposit/staking/withdrawal mechanics for the ETH entrusted to them in order to offer convenient and liquid exposure to staking yield for their end users[2].

Data taken from DefiLlama

Note that the yield for sfrxETH is slightly higher than average due to its being on the receiving end of the total yield from staked ETH underlying both sfrxETH and its 1:1 ETH pegged counterpart frxETH[3]. Additionally, mETH's APY is currently being boosted through a campaign in which mETH yield is subsidized using yield derived from staking Mantle's own treasury ETH (target 7.2% APY for 2+ months)[4].

Due to the convenience offered by LSTs over self-staking ETH, the overall market share of LSTs has risen to over 40% of all staked ETH[1], and is expected to continue to climb.

The rise of LST dominance over time has also birthed a new sector of DeFi, named LSDFi, which leverages the unique characteristics of LSTs to create new classes of on-chain financial applications. In this article we’ll explore several such application categories, and also speculate on future directions for this nascent category.

Yield Composition of LSTs

To better understand the types of financial products that can be built upon LSTs, it’s first informative to break down the sources of yield for validators. Validators earn yield from ETH staking in the following three ways:

(1) Consensus Layer rewards from proposing and attesting to blocks and participating in the sync committee[5] (2) Execution Layer rewards from priority fee tips paid by users to incentivize validators to include their transactions a block (the base fee paid by a user is set and burnt algorithmically in accordance with EIP-1559)[6] (3) (optional) Payments from block builders via MEV-boost, which passes MEV-related revenue that would otherwise go to searchers back to validators[7]

Though the consensus layer fees (1) are deterministic, fees paid via priority fees (2) and MEV-boost (3) vary based on demand for Ethereum block space. As of the time of writing of this article, the average staking APY for nonMEV-boost participating validators is 3.95%, and for MEV-boost participating validators is 5.69%[8].

Liquid staking protocols such as Mantle LSP ($mETH) leverage staking providers that operate Ethereum nodes using the ETH deposited by their users. These providers take a cut of the generated staking yield as an incentive for node operation. The LST protocol also takes a cut. In the case of Lido, for instance, the aggregate cut taken by the Lido DAO and its staking providers is 10%[9].

Generally, it is up to the underlying staking provider whether or not they want to opt into MEV-Boost[10]. In practice, in the rare case that a staking provider gets slashed, they make best efforts to make LST holders whole to avoid being removed as a service provider[11].

Even though staking providers are not able to steal the ETH delegated to them, it is still up to the LST provider to properly vet their staking providers to make sure they continue to operate high uptime nodes and act honestly. You can view more details about mETHs node operators here and regarding $mETH’s risk management approach here.

Figure 2: Anatomy of Ethereum Staking Yield

The variable nature of Ethereum staking rewards, and ability to categorize yield into multiple buckets, naturally lends itself to the creation of DeFi products that enable speculation or hedging of LST yield in part or whole.

Additionally, since LSTs are able to be used as collateral, many staple DeFi applications (lending, stablecoins, AMMs, treasury management, etc.) can be reimagined to leverage LSTs instead of native ETH. As the yield from such applications would be boosted by the usage of LSTs, they theoretically would be more attractive to DeFi participants than their native ETH-using counterparts.

We explore applications currently being built along these lines as well as others in the rest of this article.

LSTFi Ecosystem

Interest Rate swaps

Ethereum staking APY is variable, depending on factors including total amount of Ethereum staked and overall activity on the Ethereum network, among many others. This variable APY may be undesirable for certain parties who may wish for a more predictable yield. In the meantime, other market participants may wish to speculate on the future staking APY relative to the current.

In traditional finance, such products would take the form of bonds and interest rate swaps, respectively. And one protocol, Pendle, is bringing such financial instruments to LSTFi.

Pendle — Pendle enables both the creation of fixed rate bonds underwritten by Ethereum staking yield and volatile tokens whose price fluctuates with staking APY from an underlying LST. The mechanism works as follows - an LST deposited to Pendle is split into two tokens: a principle PT token, which is redeemable for the current value of the underlying LST at a fixed expiry (in nominal ETH terms), and a variable YT token, which continuously accrues the staking yield from the LST. The user who deposited the LST can withdraw it at any time provided they return the minted PT and YT tokens.

In practice, the PT token trades at a discount to the true value of the LST. This discount can be input to a discounted cash flow model and used to back out a fixed rate yield for buying and holding the PT. Thus, any user wishing to gain fixed rate exposure to staked yield for a specified time period can simply buy the PT token with the relevant expiry.

In contrast, any user wishing to obtain leveraged exposure to ETH staking yield, or speculate that staking APY will increase over time, can buy the relevant YT token. Note that a user can also “short” future staking yield by using a LST to mint PT and YT tokens, and then swapping the YT token for the PT token, thereby locking in the current staking APY as a fixed rate.

Source: Binance Research

Following the launch of $mETH last month, Pendle plans to deploy to the Mantle Network[12].

MEV Supply Chain Products

Theoretically, if an LST controlled enough of the total stake to guarantee they were a regular block proposer, it would be possible to speculate on revenue generated from transaction ordering (i.e., MEV) in these future blocks. Such products could take multiple forms, including:

1. Forced Transaction Inclusion — If a validator is chosen as the proposer for a particular slot, it has the ability to decide what transactions are included in that block. Any LST that controls enough stake to be frequently assigned as proposer could auction off the right to include a transaction in a future block regardless of network congestion. This model was previously implemented in pre-Merge proof-of-work Ethereum by Edenblock through collaborations with various miners [13].
2. Atomic transaction execution — Similar to forced inclusion, this transaction flow is currently offered by MEV-Boost, which is an open protocol that any validator or LST can opt into. Users can send a bundle of transactions to the MEV-Boost protocol along with a bid that incentivizes block builders to include all transactions in the bundle or none. The tip must also exceed the potential revenue from other searchers competing for the same opportunity.
3. Top of block auctions — Though MEV-Boost block builders typically order transaction bundles according to total bid, directly auctioning off the top of a future block could be a more efficient mechanism for pricing this desirable block space.
4. Multi-block atomic execution — In the extreme case, if an LST protocol could guarantee they have proposer rights on two blocks in a row, it would be possible to offer exotic transaction flows such as multi-block bundles. Historically, this has been difficult to achieve, since a new proposer is randomly chosen for every new block. Assuming a sufficient portion of staked ETH opts into this service offering, such workflows could theoretically guarantee atomic execution of transaction bundles that span multiple blocks.
5. L1/L2 atomic execution — Similarly, any protocol that could order and propose transactions on both L1 and L2 could offer atomic execution of transaction bundles spanning both L1 and L2. However, the offering of such a service would require both opt-in from a large proportion of staked ETH and at least one L2 sequencer operator.

However, aside from Flashbots MEV-Boost, no MEV supply chain provider has yet amassed enough buy-in from validators to offer such structured products, even though they are theoretically possible (to our knowledge, please let us know if we have missed one).

As another example, if it is possible to separately stream consensus layer and execution layer rewards from ETH staking to LSTs, it will be possible for a protocol like Pendle to enable fixed rate hedging and interest rate speculation of these two yield components independently.

Usage as Collateral

Since LSTs are composable ERC20s, they can be used as collateral in both traditional DeFi sectors such as lending, AMMs, perpetuals, and stablecoin issuance and in LST-specific variations of these protocols. This section explores some of these applications.

Ethena — Ethena is developing a novel capital-efficient stablecoin that uniquely leverages the yield offered by LSTs and the unique structure of crypto markets. Crypto markets have historically traded contango, paying users to short major crypto assets such as Bitcoin and Ether[14]. Thus, it is possible to create a synthetic delta-neutral, unliquidatable, dollar-pegged position that is yield-bearing by going 1X short against a crypto-asset such as Ether while also using the underlying asset as collateral. Some DeFi protocols such as Lemma[15] have previously experimented with this approach.

Ethena takes this approach, and further enhances it by using an LST as the underlying collateral for the short position instead of native ETH. Ethena then mints their protocol's stablecoin, USDe, against this basis traded position. This stablecoin has the benefit of both being highly capital efficient and earning both staking and basis yield.

Ethena is leveraging Bybit liquidity for its basis trade position, and also plans to include $mETH in its basket of LST collateral[16].

Lybra — In addition to Ethena's approach, several DeFi protocols have adapted previous overcollateralized stablecoin designs to yield enhanced LST-backed equivalents. Though V1 of the Lybra protocol only accepted stETH and ETH, V2 has opened up to a basket of additional LSTs. Lybra enables users to mint an interest-earning stablecoin eUSD using their LSTs as collateral. The protocol then uses the staking yield earned by the LST to purchase eUSD, and transfer it to existing eUSD holders in the form of rebasing-based interest. Alternatively, users can convert their eUSD into the more composable peUSD, which can be used in other DeFi protocols and converted back into eUSD, plus accrued interest when the user desired. Since eUSD is overcollateralized by LSTs, the native yield for holding it is around 8% APY. In addition, Lybra has a governance token $LBR that takes a cut of staking yield and protocol fees, and is used to incentivize liquidity provision and other necessary protocol functions.

Source: Lybra

Several other protocols have adopted similar approaches, including Raft and Prisma — which enables minting of their stablecoin $mkUSD using a Liquity-like mechanism against wstETH, rETH, cbETH, or sfrxETH. Unlike Lybra, which passes LST yield onto the stablecoin, Prisma enables stablecoin minters to retain their ETH staking yield. Like Liquity, Prisma further boosts yield incentives using their own native $PRISMA token.

Ion Protocol — Many incumbent lending dApps, such as Aave, have incorporated LSTs such as wstETH into their list of supported assets[17]. However, due to the constraints imposed by their existing protocol mechanics, they are limited to manage liquidations via traditional price-based oracle mechanism. For in-kind leveraged borrow looping of stETH-ETH, the dominant lending utilized lending strategy, this mechanism inefficiently accounts for the primary driver of insolvency — slashing risk.

Ion protocol addresses this inefficiency by offering loans of ETH against LSTs that are not exposed to price-based liquidation risk. Instead, a zero knowledge machine learning framework is used to assess the slashing risk of various LSTs in real time, and adjust loan market parameters (max LTV, interest rate, etc.) according to the inferred risk.

Metastreet — A relatively new entrant into the lending space, Metastreet focuses on offering ETH-backed nonliquidatable loans for NFTs. Recently, the protocol enabled lending of wstETH against NFTs as well, boosting its already high NFT lending APYs with LST staking yield.

Metastreet also offers yield bearing liquid credit tokens (LCTs) that are created when deposting wstETH into the lending pool, and can be redeemed for wstETH plus accrued interest after a redemption period, or swapped for wstETH anytime in an accompanying Curve pool (usually at a slight discount to redemption value).

Treasury and Bridge Asset Management

Due to the perceived safety and high liquidity of LSTs, they offer an attractive approach for DAOs looking to earn yield on their underlying treasuries[18]. Since such treasuries often hold large quantities of assets, LST providers are eager to tap into this potential user base.

In a similar vein, rollups looking to attract TVL can juice yield offered to ETH depositors by converting this ETH into an LST. However, this approach slightly increases risk of fund loss by exposing rollup participants to additional slashing and smart contract risks from the LST protocol.

Blast — Blast exemplifies an extreme version of this strategy. The rollup made headlines and generated contraversy over its points-incentivized deposit program, which rapidly amassed $840M in TVL over the last month[19]. Uniquely, every asset deposited to Blast is converted into a yield bearing equivalent — Ether is converted to wstETH, and stablecoins become treasury-based assets in MakerDAO. Thus, all ETH and stablecoin assets on Blast natively earn yield. Blast's VM exposes several solidity-based endpoints that allow smart contract developers to choose how to handle this native yield on behalf of their users. To further increase yield, Blast has also promised to airdrop its native token on bridge depositors.

Source: Blast

However, as adoption of LSTs continues to trend upward throughout the Ethereum ecosystem, Blast's approach may simply become the status quo within a few years time.

Yield Aggregators

Analogous to other DeFi dApps in this category, LST-focused yield aggregators offer strategies to optimize and amplify the yield obtained from LSTs. In some cases, these aggregators also issue their own token for participating in these strategies to further increase the obtainable yield.

Equilibria — Equilibria operates as a Convex-like layer on top of Pendle, enabling liquid swapping between ePendle, a token representing the protocol's internal vote escrowed Pendle balance, and $PENDLE. Like Convex, assets deposited to Pendle strategies through Equilibria not only earn boosted yield through the "ve" mechanism, but also earn $EQB — Equilibria's native token. $EQB holders earn a cut of protocol fees, and can vote on bribe directed $PENDLE emissions, acting as a meta-governance layer on top of Pendle.

Sommelier Finance — Sommelier Finance offers several yield optimization strategies centered around LSTs, such as the aforementioned looped LST-ETH lending strategy, as well as concentrated Uni-V3 liquidity provision. For instance, its Turbo stETH strategy boasts a 10% APY — a significant boost over simply holding stETH - albeit at increased exposure to smart contract risks. Sommelier also issues their own native $SOMM token for participation in certain strategies.

Several other LSTfi focused yield aggregator applications offer similar or competing products to Sommelier, including Cian, Range Protocol, and many others.

Indexes and Rebalancers

Recently, the growing dominance of Lido reached a breakpoint as it approached control of one-third of all staked ETH[20]. This critical threshold, if exceeded, would mean that Lido has the power to cause liveness failures on Ethereum. If Lido’s dominance continued to grow past this point, it would eventually have the (theoretical) ability to perform censoring attacks, form MEV cartels, and even potentially double spends[21, 22].

Even the possibility of a single protocol capable of harming the liveness, security, or censorship resistance guarantees of Ethereum severely undermines its value as the world computer. As such, several protocols are using DeFi incentives to encourage a more equitable distribution of staked ETH across a basket of LST providers.

In tandem, these protocols also seek to optimize the safety and yield profile of their underlying basket of LSTs.

unshETH — unshETH is developing an LST index product along these lines. ETH deposited to the protocol is subsequently deposited into a basket of underlying LSTs. In addition, the basket operates as a virtual AMM, allowing users to swap between different LSTs and ETH (until that LST reaches some maximum percentage of the index) for a fee. This mechanism both encourages validator decentralization by increasing liquidity in LST-LST swaps, and increases the yield for holding unshETH through swap and mint/redeem fees (beyond the staking yield from the underlying LST basket).

Source: unshETH

Several other protocols, such as Asymmetry Finance and Index Coop, offer similar LST index products. However, the specific underlying basket of LSTs and specific strategies surrounding yield optimization differ.

Future Directions

The advent of Ethereum restaking through EigenLayer has both introduced an additional source of yield for Ethereum staking and a new liquid derivative — LRTs (Liquid Restaking Tokens).

Though Eigenlayer is not yet live, the possibilities offered by restaking have captured a considerable amount of mind share and capital commitment, with the total amount of staked ETH pre-committed to EigenLayer approaching 500,000 already[23].

Many LRT providers are already vying for the opportunity to grab restaking market share. As such providers will need to leverage underlying LSTs as collateral, we expect them to both act as a source of demand for $mETH and other LSTs, while also themselves serving as collateral for a suite of LRTFi applications, which can take inspiration from the categories described above.

Source: mETH

1. https://www.stakingrewards.com/asset/ethereum-2-0
2. https://docs.mantle.xyz/meth/concepts/risk-management/exchange-rate
3. https://docs.frax.finance/frax-ether/frxeth-and-sfrxeth
4. https://www.mantle.xyz/blog/announcements/meth-double-dose-drive
5. https://ethereum.org/en/developers/docs/consensus-mechanisms/pos/rewards-and-penalties/
6. https://ethereum.org/en/developers/docs/gas/
7. https://github.com/flashbots/mev-boost
8. https://www.blocknative.com/ethereum-staking-calculator
9. https://stake.lido.fi/
10. https://research.lido.fi/t/research-mev-boost-for-lido-validators/2333
11. https://blog.lido.fi/post-mortem-launchnodes-slashing-incident/
12. https://www.mantle.xyz/blog/ecosystem/pendle-first-lst-protocol-mantle-network
13. https://medium.com/edennetwork/how-to-leverage-eden-network-to-trade-with-priority-1171851978c2
14. https://www.coinglass.com/funding/ETH
15. https://medium.com/coinmonks/earning-defi-yield-via-basis-trading-379d1d5e7207
16. https://www.mantle.xyz/blog/ecosystem/ethena-to-deploy-usde-on-mantle-network-special-pre-launch-mantle-showcase
17. https://app.aave.com/governance/proposal/210/
18. https://research.lido.fi/t/should-lidodao-stake-treasury-eth/3869/6
19. https://blast.io/en
20. https://www.coindesk.com/consensus-magazine/2023/12/04/lido-dao-democratized-eth-staking-then-dominated-it/
21. https://github.com/djrtwo/writing/blob/main/docs/2022-05-30_the-risks-of-lsd.md
22. https://projects.iq.harvard.edu/files/applied-cryptography-society/files/gtib20_paper_8.pdf
23. https://app.eigenlayer.xyz/