If I secure three AVSs at the same time, does that simply triple my yield?
No. Yield doesn't stack through simple addition — how much you actually earn depends on how much each AVS is willing to pay in rewards, and how much of its "security budget" still needs filling. AVSs don't want unlimited staked ETH protecting them; once they reach what they consider a sufficient security level, additional stake doesn't bring proportionally more yield.
More importantly, as yield stacks, so does risk — and that stacking usually isn't linear. Each of the three AVSs' penalty conditions act on the same principal simultaneously; if any one of them runs into trouble, it affects your entire principal, not a third of it. Treating "securing a few more AVSs" as equivalent to "earning a few more paychecks" is a common misconception — it's actually closer to signing three separate contracts that each carry joint-liability clauses.
How does a Liquid Restaking Token (LRT) differ from an ordinary liquid Staking token (like stETH), and why is the risk tier different?
The similarity is that both wrap an otherwise locked, illiquid position (staked ETH) into a freely tradable token that can be used across other DeFi protocols — letting users participate in staking without sacrificing liquidity.
The difference lies in how many layers of underlying risk each one carries. stETH represents only the risk of "staked to Ethereum itself" — simple slashing rules, battle-tested over a long stretch of time. An LRT (like eETH or rsETH) represents a compound position of "staked to Ethereum plus delegated to one or more AVSs at the same time" — stacking on the AVS's own slashing rules, operator execution risk, and the LRT issuer's own Smart Contract and governance risk. This is also why you can't apply the same evaluation logic and risk expectations you'd use for stETH directly to an LRT — the two share a similar naming structure (both are "tokenized versions of staked X"), but the number of risk layers stacked underneath is completely different.
What role do Operators play in the Restaking architecture, and why does their behavior affect me?
Operators are the technical teams actually running the servers that perform validation work for an AVS. When you stake ETH and delegate it to an Operator, you're effectively handing that Operator full authority over how "that ETH's behavior" is represented. An AVS's slashing rules are triggered based on the Operator's actual behavior — if that Operator's server goes offline, signs an incorrect validation message, or gets compromised and behaves abnormally, the resulting penalty directly hits the principal of every user who delegated to that Operator, not just the Operator's own assets.
This means who you delegate your staked position to has nothing to do with whether you personally did anything wrong — you can make zero bad decisions of your own and still bear the penalty consequences because the Operator you delegated to made a mistake. This is also why some users choose to split their position across multiple Operators, reducing the chance that a single Operator's failure drags down their entire principal.
Where does the extra yield from Restaking actually come from, and is that yield model sustainable long-term?
Most restaking yield in today's market doesn't come from fees an AVS genuinely generates (such as a service fee an Oracle or bridge earns from being widely used) — it comes from Token rewards the restaking protocol issues itself. This means a substantial portion of "extra yield" is essentially the protocol subsidizing early participants with its own token, used to rapidly grow the amount of capital staked.
Whether that subsidy model can last long-term hinges on one thing: whether the AVSs being secured can genuinely find paying users over time (for example, an L2 project willing to pay long-term to use a given data availability layer), shifting the business model from "protocol subsidizes with tokens" to "the service actually has buyers." This is still early-stage and unproven. Adoption at most live AVSs remains far below the scale of capital locked across the overall restaking market — and that gap is exactly the signal worth watching when evaluating whether restaking yield is actually sustainable.
Staking ETH to help validate Ethereum and earning base yield is something most people are already familiar with. But what if that same already-staked ETH could be "lent out again" to help secure other projects and earn additional yield on top? That's what Restaking sets out to do, and the market has already grown past $15 billion — but what it stacks isn't just yield, it's layer after layer of risk that isn't always easy to see clearly.
An Ethereum validator staking ETH is essentially putting up that money as a security deposit — if the validator misbehaves or goes offline, that deposit gets slashed, and this mechanism is what secures the Ethereum network itself. Restaking protocols (the market is currently dominated almost entirely by EigenLayer, with over 90% market share) let you "reuse" that same ETH already staked to Ethereum — or a liquid staking Token representing it, like stETH — to help provide similar security guarantees to other new projects. These new projects are collectively called AVSs (Actively Validated Services), which might be a decentralized Oracle Network, a Cross-Chain Bridge, or a service handling data availability for a Layer 2.
You don't need fresh capital for each AVS — the same staked ETH can, in theory, simultaneously secure multiple AVSs at once, earning extra yield from each one that pays for the service. This is exactly where the name "restaking" comes from: the same principal, staked repeatedly, multiple times over.
This is what beginners overlook most often: when your ETH secures Ethereum itself, it follows Ethereum's consensus-layer slashing rules. The moment you use that same ETH to secure an AVS, you're also simultaneously accepting an additional, independent set of penalty rules that AVS defines for itself. Every AVS's slashing conditions are different, and each takes effect independently — meaning if you're securing three AVSs at once, you're not choosing one of three, you're stacking three separate rule sets on top of the same principal. If any one of those AVS's operating teams (called Operators) makes a mistake or acts maliciously, your principal can be affected.
Most users don't interact directly with EigenLayer's underlying contracts — instead they participate indirectly through a Liquid Restaking Token (LRT, such as ether.fi's eETH or Kelp's rsETH). You deposit ETH and receive a tradable token representing your position, lowering the barrier to participate — but that also adds another layer of trust in the LRT issuer itself, plus its own Smart Contract Risk.
In April 2026, Kelp DAO's rsETH suffered an exploit worth roughly $300 million. In the aftermath, the entire restaking sector saw roughly $5.4 billion in withdrawals within a short window — not just from Kelp's own users, but from users of protocols entirely unrelated to Kelp, driven by panic. The incident also caused some lending positions built on top of LRTs to get liquidated at unfavorable prices amid the liquidity crunch — a direct illustration of the multi-layer contract risk described above: a problem originating in one LRT propagated through a lending protocol into losses for a completely different set of users.
If you're considering participating in restaking, start by asking yourself one question: is the extra yield you're earning primarily coming from real fee revenue an AVS generates, or from token subsidies the protocol issues itself? Most restaking yield in the market today still consists mainly of token rewards — how much genuine, sustainable fee revenue AVSs themselves can generate is still being tested in these early days. Second, if you choose to participate indirectly through an LRT, you're stacking on an additional layer of trust in and contract risk from that LRT's issuer. Evaluating restaking isn't just about base staking yield plus AVS yield — it also means factoring in how mature that specific LRT is, and whether it's been battle-tested through enough real-world stress.