DeFiMay 30, 2026

What Is Restaking? When the Same Asset Gets Pledged More Than Once

Ethereum's proof-of-stake mechanism works by having validators lock up ETH as a security deposit. They earn a yield for doing this — around 3 to 4 percent annually as of early 2026 — but the ETH itself is frozen. It cannot be moved or used while it is staked.

Liquid staking got around the freezing problem. A protocol like Lido stakes the ETH on your behalf and issues you a receipt token stETH in return. That token represents your staked ETH and moves freely through DeFi while the original ETH stays locked. You get liquidity. The ETH stays put.

Restaking takes the stETH you received and puts it to work a second time. Depositing it into a restaking protocol means that same asset now backs additional services, oracle networks, cross-chain bridges, data availability layers, on top of its original job securing Ethereum. Each service pays its own yield for the security it receives.

There is a word for this in traditional finance: rehypothecation. When a broker takes collateral pledged by one client and uses it as collateral in another transaction, that is what they are doing. Before 2008, the practice ran through the entire global financial system. The same assets moved through chains of institutions, each transaction looking sound in isolation, the true extent of interdependence invisible until it wasn't. When the system buckled, the failures did not stay where they started. They moved along the chains. The regulatory response was largely about making those chains visible; requiring transparency about what was pledged where before they broke. Restaking builds the same kind of chain. Each layer added to the stack depends on the layers beneath it holding.

Participating in restaking means trusting several things at once. The protocol itself has to work as designed, not just the code, but the operators running validators for each service, and the slashing rules those services apply when validators misbehave. Get those rules wrong, or get unlucky with an edge case they did not anticipate, and the penalty flows back to you.

The services being secured have to stay solvent. A failure serious enough to trigger slashing in one of those systems does not stay contained there. It comes back through the restaking layer.

Then there is the receipt token; restaking protocols take stETH or similar tokens, not raw ETH. The hidden risk is that If that token loses its peg under pressure, the security value the protocol is offering changes with it. None of this is guaranteed, every layer of restaking is a bet that all the layers below it hold at the same time.

The yield comes from the protocols paying for security. They need honest validators; the rewards are how they get them. What restaking calls yield, you might also call a fee for taking on obligations you will only fully understand when something goes wrong.

The number shown next to a restaking option is yield. What it does not show is what you have agreed to secure, under what rules, and what happens to your ETH if one of those systems breaks.

Using the same capital to secure multiple systems at once is genuinely useful for users and for the protocols that would otherwise have to attract security from scratch. However, traditional finance tried the same thing before. It did not go well.