LRT vs LST: What's The Difference (And Why It Matters)

The Question: LRT vs LST

Readers searching "LRT vs LST" want to know whether the higher advertised yield on Liquid Restaking Tokens justifies the compounded risk. The answer requires decomposing what each token represents on-chain, where the yield components actually come from, and which specific conditions cause each mechanism to fail.

Liquid Staking Tokens (LSTs) like stETH and rETH represent staked ETH on Ethereum's consensus layer. You deposit ETH into a protocol, receive a liquid token in return, and earn consensus layer rewards while maintaining liquidity. The yield is singular: base staking from validator rewards. As of mid-2026, LST yields converge around 2.0 to 3.0 percent APY across major protocols.

Liquid Restaking Tokens (LRTs) like rsETH, ezETH, and weETH take an LST (or native ETH) and restake it through EigenLayer to secure additional services called Actively Validated Services (AVSs). You earn base staking rewards plus AVS rewards. The advertised yield is 8 to 12 percent APY. The risk stack is deeper.

The difference is architectural. An LST secures one layer: Ethereum consensus. An LRT secures multiple layers: Ethereum consensus plus however many AVSs the protocol operator has whitelisted. More layers mean more yield sources. They also mean more slashing conditions.

How LSTs Work: Single Layer, Single Yield Source

An LST represents a claim on staked ETH in Ethereum's consensus layer. When you deposit 1 ETH into Lido, you receive stETH. That stETH accrues rewards from validator activity: block proposals, attestations, and sync committee participation. The 3 to 5 percent LST yield breaks down into three components: issuance rewards (new ETH created by the protocol), MEV (miner extractable value from transaction ordering), and priority fees (tips paid by users for faster inclusion).

Lido charges a 10 percent fee on staking rewards. Rocket Pool charges a similar fee structure. After fees, net APY settles around 2.2 percent for stETH and 2.3 percent for rETH as of mid-2026. The yield is predictable because the source is singular. Validator rewards fluctuate based on network activity, but the mechanism is stable. Ethereum's consensus layer has been live since December 2020 with no major protocol-level failures.

The risk profile for LSTs consists of three primary vectors. First, smart contract risk: a bug in the staking protocol contract could lock funds or enable unauthorized withdrawals. Second, validator-side slashing: if a validator double-signs or goes offline, a portion of staked ETH gets burned by the protocol. Third, liquidity risk: during market stress, LSTs can trade below their redemption value on secondary markets. stETH depegged approximately 7 percent below ETH in June 2022 during the Three Arrows Capital collapse, not because of a protocol failure, but because holders sold into thin liquidity to meet margin calls.

LST mechanics are linear. You deposit ETH. You receive a token representing staked ETH. You earn consensus rewards. You can redeem the token for ETH (plus accrued rewards) after the withdrawal queue clears. The failure modes are known and documented.

Major LST Protocols

Lido controls roughly 28 to 30 percent of all staked ETH, approximately 11 to 12 million ETH as of mid-2026. That dominance creates centralization risk (if Lido's operator set colludes, they could theoretically control consensus decisions), but it also creates deep liquidity. stETH is integrated as collateral in nearly every major DeFi protocol: Aave, Maker, Curve, Balancer. The liquidity advantage is structural.

Rocket Pool offers decentralization at the cost of liquidity depth. Anyone can run a Rocket Pool node with 8 ETH plus an RPL bond. The validator set is distributed across thousands of independent operators. rETH uses a reward-bearing model where the token's value relative to ETH increases over time, which makes it tax-efficient in some jurisdictions since daily balance updates do not trigger taxable events.

Coinbase offers cbETH, which carries regulatory clarity (Coinbase is a publicly traded, US-regulated entity) but introduces custodial risk. If Coinbase faces regulatory action or operational failure, cbETH holders face counterparty exposure that decentralized protocols do not carry.

Yield differences across LSTs are minimal. The spread between Lido, Rocket Pool, and Coinbase is 10 to 30 basis points. Protocol choice depends on liquidity needs, decentralization preference, and regulatory posture, not yield maximization.

How LRTs Work: Stacked Layers, Stacked Yield, Stacked Risk

An LRT takes the LST mechanism and adds a restaking layer. The architecture works like this: you deposit ETH or an existing LST into an LRT protocol. That protocol deposits your ETH into Ethereum staking (or wraps your existing LST). Then it restakes the position through EigenLayer to secure additional services: AVSs like data availability networks, decentralized sequencers, oracle networks, and cross-chain bridges.

Each AVS pays rewards to restakers for providing economic security. The payment structure varies. Some AVSs pay in their native token. Some pay in ETH. Some pay in points that convert to future token airdrops. The yield components stack: base Ethereum staking (2 to 3 percent) plus AVS rewards (advertised as 5 to 10 percent additional) equals total advertised yield of 8 to 15 percent.

The mechanism is modular. An LRT holder earning base staking yield of 3.5 percent might also earn 2 percent from a data availability AVS, 1 percent from a decentralized sequencer AVS, and token incentives from the LRT protocol itself. The yield is higher because the same capital secures multiple services simultaneously. The capital is doing more work.

This is where the risk compounds. Each AVS has its own slashing conditions. If the data availability AVS detects that a validator submitted incorrect data, it can slash a portion of the restaked ETH. If the sequencer AVS detects that a validator censored transactions, it can slash again. The slashing events are independent. A single piece of capital can be slashed multiple times by different services.

As of March 2026, incremental restaking yield from EigenLayer AVSs ranges from 0.3 to 1.5 percent additional APY in real cash flow. The rest of the advertised yield comes from points farming and token incentives. Many protocols distribute governance tokens or airdrop points to early restakers. This is not sustainable yield. It is protocol bootstrapping. When the incentives end, the yield compresses.

Major LRT Protocols

EigenLayer holds roughly 90 percent of the Ethereum restaking market with over 4.6 million ETH committed and total value locked that peaked above 15 billion dollars in early 2026. EigenLayer is the infrastructure layer. It does not issue an LRT directly. Instead, protocols like Kelp DAO, Renzo, and Puffer Finance build on top of EigenLayer and issue their own LRTs.

rsETH (Kelp DAO) aggregates restaking across multiple operators and AVSs. In April 2026, Kelp suffered a security incident where approximately 280 to 293 million dollars of rsETH was drained. The vulnerability was not in the Kelp contract itself. It was in the LayerZero cross-chain adapter used to bridge rsETH to other chains. This is a new risk category: LRTs bridged across multiple chains introduce cross-chain bridge risk that base LSTs do not carry.

ezETH (Renzo) offers a similar restaking wrapper with its own approach to operator selection and AVS allocation. pufETH (Puffer Finance) differentiates through anti-slashing technology designed to reduce validator penalties. ether.fi issues both rebasing eETH and non-rebasing weETH. The non-rebasing version is more compatible with DeFi protocols that do not handle rebasing tokens well.

The choice of AVSs matters. Each LRT protocol whitelists a different set of AVSs. An overly aggressive AVS selection increases yield but also increases slashing risk. An overly conservative selection reduces slashing risk but makes the LRT less competitive on yield. The risk profiles are not interchangeable. rsETH, ezETH, and pufETH are not equivalent products even though they all advertise similar yields.

The Risk Stack: What Breaks Where

LST risks are contained. Smart contract risk, validator slashing, and liquidity risk are the three failure modes. All three have historical precedents, and all three are understood. A severe depeg event (like stETH in June 2022) creates a buying opportunity for holders who do not need immediate liquidity. The worst-case scenario is a multi-day withdrawal queue and a temporary discount to redemption value.

LRT risks cascade. Start with the LST risk base: smart contract risk, validator slashing, liquidity risk. Add AVS-level slashing: each AVS has its own conditions under which it can burn a portion of your restaked ETH. Add cross-chain bridge risk if the LRT is bridged to other chains. Add protocol-level governance risk: the LRT protocol operator decides which AVSs to whitelist, and a poor choice exposes all holders to that AVS's failure modes.

The fee stack also cascades. LST protocols charge around 10 percent of rewards. LRT protocols charge an additional fee on top of that. AVS operators may charge their own fees. The gross yield might be 12 percent, but after three layers of fees, the net yield to the holder could be 8 percent or lower.

By 2026, slashing on EigenLayer is live. The protocol has added the ability for slashed funds to be redistributed rather than simply destroyed. This raises the stakes. A slashing event is not just a penalty. It is a redistribution to other participants. The incentive to report slashable behavior increases. The probability of slashing increases.

The architectural difference matters. LSTs are single-point-of-failure systems. If Ethereum consensus fails, the LST fails. LRTs are multi-point-of-failure systems. If Ethereum consensus fails, the LRT fails. If any whitelisted AVS fails, the LRT can be partially slashed. If the cross-chain bridge is exploited, the LRT can be drained. The number of failure conditions is higher.

Real Yield vs Points Farming

The advertised 8 to 15 percent LRT yield is not all cash flow. A significant portion comes from points and token incentives. Protocols distribute governance tokens or airdrop points that convert to tokens at a future date. This is not real yield. It is speculative yield. The value depends on future token price, future airdrop conversion rates, and future protocol adoption.

As of early 2026, most AVS rewards are still bootstrapping. Actual cash flow yields from AVSs range from 0.3 to 1.5 percent additional APY. Add that to base staking yield of 2 to 3 percent, and the sustainable yield on LRTs is 3 to 5 percent in real cash flow. The rest is protocol incentives.

When incentives end, the yield compresses. This is not theoretical. It happened with early DeFi yield farming in 2020 and 2021. Protocols that advertised 100 percent APY saw yields collapse to 5 percent when token emissions stopped. LRTs will follow the same pattern. The question is when the incentives end and how much cash flow yield remains.

When It Matters, When It Doesn't

If you hold a position for passive staking income and want predictable yield with minimal complexity, LSTs are the correct choice. The yield is lower, but the risk is contained and well-understood. If you need deep liquidity for collateral or trading, stETH is the dominant option. If you prioritize decentralization, rETH is the better choice. If you need regulatory clarity, cbETH offers that.

If you are willing to accept compounded risk for higher yield and believe AVS adoption will create sustainable cash flow over the next two to three years, LRTs are the higher-risk, higher-reward option. The additional 2 to 4 percent yield (in real cash flow, excluding points) comes from taking on additional slashing exposure and protocol risk.

The yield difference matters when the capital is large. On 100 ETH, the difference between 2.5 percent LST yield and 5 percent LRT yield (excluding points) is 2.5 ETH per year, roughly 6,250 dollars at 2,500 dollar ETH. On 10 ETH, the difference is 0.25 ETH per year, 625 dollars. The absolute return scales with position size. The risk scales identically.

If you cannot monitor AVS health, operator selection, and cross-chain bridge security, do not hold LRTs. The mechanism requires active oversight. The compounded risk profile demands compounded diligence. LSTs are set-and-forget. LRTs are not.

The Takeaway

LRTs sit on top of LSTs. The architecture stacks an additional yield layer and an additional risk layer. Base staking yield of 2 to 3 percent plus AVS yield of 0.3 to 1.5 percent (in real cash flow, as of mid-2026) equals 3 to 5 percent sustainable yield before points. The rest is protocol incentives that will compress over time.

The failure mode for LSTs is contained: smart contract exploit, validator slashing, or temporary liquidity depeg. The failure mode for LRTs is compounded: any of the LST failure modes plus AVS-level slashing plus cross-chain bridge exploit plus poor AVS selection by the protocol operator. A reserve sized to a single slashing event is inadequate for restaking, where the same capital secures multiple services.

Historical precedent: stETH depegged 7 percent in June 2022 but recovered. Kelp DAO lost 280 to 293 million dollars in April 2026 due to a cross-chain bridge exploit. The LST risk materialized as a temporary price discount. The LRT risk materialized as a permanent capital loss. That difference is structural.

If you want the highest sustainable yield and can accept multi-vector risk, LRTs offer 1 to 2 percent additional real cash flow over LSTs. If you want predictable yield with minimal oversight, LSTs remain the better choice. The mechanism is simpler, the risk is contained, and the liquidity is deeper.

Frequently Asked Questions

What is the main difference between LSTs and LRTs?

LSTs represent staked ETH on Ethereum's consensus layer, earning base staking rewards of 2 to 3 percent APY. LRTs take an LST (or native ETH) and restake it through EigenLayer to secure additional services called AVSs, earning base staking plus AVS rewards. The architecture differs: LSTs secure one layer (Ethereum consensus), while LRTs secure multiple layers (Ethereum plus multiple AVSs). More layers mean more yield sources and more slashing conditions.

How much real yield do LRTs actually generate?

As of mid-2026, sustainable cash flow yield on LRTs is 3 to 5 percent APY: base Ethereum staking (2 to 3 percent) plus incremental AVS rewards (0.3 to 1.5 percent). Advertised yields of 8 to 15 percent include points farming and token incentives, which are not sustainable cash flow. When protocol incentives end, yield compresses to the cash flow base. The additional 1 to 2 percent over base LST yield comes from accepting multi-vector slashing risk.

What are the specific risks of holding LRTs vs LSTs?

LST risks are contained: smart contract exploit, validator slashing, or temporary liquidity depeg. LRT risks cascade: all LST risks plus AVS-level slashing (each AVS can slash independently), cross-chain bridge risk (if bridged to other chains), and protocol governance risk (poor AVS selection exposes all holders). Kelp DAO lost 280 to 293 million dollars in April 2026 from a bridge exploit. A reserve sized for single slashing is inadequate for restaking where capital secures multiple services simultaneously.

Which LST protocol should I choose?

Yield differences are minimal (10 to 30 basis points across Lido, Rocket Pool, and Coinbase). Choose based on liquidity needs, decentralization preference, and regulatory posture. Lido (stETH) offers deepest liquidity and widest DeFi integration. Rocket Pool (rETH) offers strongest decentralization with permissionless node operation. Coinbase (cbETH) offers regulatory clarity but introduces custodial risk. Protocol choice depends on your priorities, not yield maximization. All converge around 2 to 3 percent net APY.

When should I hold LRTs instead of LSTs?

Hold LRTs if you accept compounded risk for 1 to 2 percent additional real cash flow and can actively monitor AVS health, operator selection, and bridge security. Hold LSTs if you want predictable yield with minimal oversight. The mechanism difference is structural: LSTs are set-and-forget with contained risk. LRTs require active diligence because failure modes compound across multiple layers. On large positions (100-plus ETH), the additional yield is material. On small positions, the risk-adjusted return may not justify the complexity.