The pitch for a liquid restaking token is that your staked ETH can do two jobs at once. It keeps earning normal staking yield, and it moonlights as collateral for a set of other services that pay rent for the security. Same asset, bigger number on the dashboard, and the interface never asks you to think about what the second job involves. A friend rotated his whole stETH position into one of these because the displayed APY was a couple of points higher, and when I asked him which services his ETH was now securing and what any of them could slash him for, he sent back a shrug. He is in good company, because the products are built so you never have to look, and the extra yield shows up as a single number with no line items. The line items are worth pulling apart, since that is where the actual trade lives.
What the extra yield actually is
Start with what a plain liquid staking token pays. Hold stETH or rETH and you collect Ethereum's base staking yield, which comes from consensus rewards plus execution layer tips and MEV, minus the protocol's cut. Historically that lands somewhere around three to four percent, drifting with how much ETH is staked and how busy the chain is. The important property is that the risk behind it is well understood. Validators get slashed for a short, public list of offenses, mostly double signing and related faults, and the large staking protocols have operated for years with slashing events staying rare and small.
A liquid restaking token takes that same staked ETH and pledges it again through a restaking protocol, and in practice most of these tokens are built on EigenLayer. The services renting the security, called AVSs, are things like data availability layers, oracle networks, bridges, and keeper systems that want economic backing without bootstrapping their own validator set and token. They pay rewards for it, and those rewards are the first component of the LRT increment.
The second component, and historically often the larger one, is points. Restaking protocols and LRT issuers have leaned hard on points programs, which are IOUs against possible future airdrops. When an LRT advertises a big headline number, it is worth checking how much of it is actual cash flow in ETH or liquid tokens and how much is a points multiplier that someone translated into an implied APY using assumptions about a token that may not exist yet. In my experience the cash portion of the restaking increment has typically been modest, often well under a percent, while the points portion is whatever the market's imagination says it is that month.
The slashing conditions you signed up for
Here is the part the dashboard does not show. Every AVS your stake secures gets to define its own slashing conditions. A data availability service might slash for withholding data. An oracle network might slash for signing a bad price. A bridge might slash for attesting to a state root that turns out to be wrong. When you hold an LRT, the issuer has delegated your stake to a set of node operators, those operators have opted into some basket of AVSs, and your ETH is now exposed to every slashing condition in the basket. You did not read those conditions. Neither have I for most of them, honestly, and I do this for a living.
Two things make this worse than it first sounds. The first is correlated slashing. A fairly small group of professional operators runs a large share of all delegated restaked ETH, and they tend to opt into the same popular AVSs. A bug in one widely adopted AVS, or in its slashing logic itself, penalizes every operator running it at the same time, which means it reaches most of the big LRTs at the same time. Restaking converts a pile of independent small risks into one shared tail risk, and the yield you are paid does not obviously scale with that change.
The second is that slashing can simply be wrong. Ethereum's own slashing conditions were argued over and reviewed for years before anyone's ETH was at stake. A young AVS ships slashing logic that has been reviewed for however long that team has existed. Faulty slashing, where honest operators get penalized because the condition itself has a bug, is a live risk in any new system, and there is no mature insurance market pricing it for you.
The wrapper adds its own failure modes
On top of the restaking exposure sits the token itself. An LRT is a receipt for a position that is slow to unwind by design. Getting your ETH back natively means exiting the AVS delegations, waiting out the restaking protocol's withdrawal delay, which has typically been measured in days, and possibly queueing behind Ethereum's validator exits after that. So in practice people who want out sell the token on the open market, and LRT markets are thinner than the markets for the big LSTs.
That thinness matters because of how these tokens get used. The dominant trade during points seasons has been looping: deposit the LRT as collateral on a lending market, borrow ETH against it, buy more of the LRT, repeat. Looping multiplies your points and also multiplies your exposure to any gap between the token's market price and its underlying value. There has already been at least one sharp depeg of a major LRT during an airdrop related unwind. Leveraged loops liquidated into a thin pool, the liquidations pushed the price further from peg, and that triggered more liquidations. The underlying ETH was never impaired, the losses came from the wrapper, and even unleveraged holders watched their token trade below backing for a stretch.
I would also flag the mundane risk people skip past, which is contract surface. A plain LST is one protocol. An LRT is that protocol plus the restaking protocol plus the issuer's own contracts plus whatever lending market you loop on. Each layer has admin keys, upgrade paths, and oracles, and your real exposure is the union of all of them.
A quick way to judge the premium
When someone asks me whether a given LRT is worth holding over a plain LST, I walk the same short list, and it rarely takes more than half an hour.
- Split the advertised yield into three buckets: base staking, cash AVS rewards, and points. Price the points at zero first and see whether the trade still makes sense. If it only works with the points, you are speculating on an airdrop, which is fine as long as you call it that.
- Find the list of AVSs the issuer's operators have opted into and skim what each one can slash for. If you cannot locate this within fifteen minutes, that difficulty is itself an answer.
- Check operator concentration. If a handful of operators carry most of the delegated stake, one operational mistake propagates straight to you.
- Map the real exit. How deep is the main liquidity pool relative to the token's supply, and what is the native withdrawal timeline once that pool is drained? Assume you will need the exit on the same day everyone else does.
- Set a hurdle. My rough rule is that if the cash portion of the extra yield is under about one percent annualized, the added tail risk and the worse exit are not being paid for, and the plain LST wins.
None of this means LRTs are a bad product. Renting out pooled security is a reasonable idea, and if AVS payments grow into real cash flow, the math changes and I will happily rerun it. What bothers me is the framing of the increment as free yield on an asset you already held, because the extra yield exists precisely to pay you for extra ways to lose. Flows are the tell here, and it is one of the patterns I watch in Blockcircle's whale tracking, since large wallets unwinding LRT loops have tended to move well before any visible depeg. If you hold one of these, know which bucket your yield comes from and know your exit, and you are already ahead of most of the people holding it next to you.