The numbers are staggering. As of this week, liquid restaking protocols have collectively amassed over $15 billion in total value locked. That is a doubling in just two months. But here is the detail that should make every prudent analyst pause: over 70% of all restaked ETH is controlled by just eight operator entities. This is not a critique of the technology itself—it is a reality check on a narrative that has been sold with more enthusiasm than scrutiny.
Let me step back. Restaking, pioneered by EigenLayer, allows ETH stakers to opt into securing other networks—called Actively Validated Services (AVSs)—in exchange for additional yield. Liquid restaking tokens (LRTs) were born to make this accessible: deposit your ETH, get a liquid token that can be deployed elsewhere in DeFi, while the underlying ETH is restaked through a protocol like ether.fi, Renzo, or Kelp. The pitch is elegant: earn multiple yields from a single capital base, all while contributing to the security of a burgeoning ecosystem. The promise is that this structure ‘democratizes’ access to off-chain security markets.
But during my years auditing ICO whitepapers back in 2017, I learned that the most dangerous flaws are not in the code—they are in the assumptions the code makes about human behavior. And the assumption that liquid restaking is inherently decentralized because it uses smart contracts is exactly that kind of flaw.
The core insight is this: LRT protocols are not permissionless security providers. They are delegated trust networks wearing a decentralized mask.
Here is how it works in practice. When you deposit ETH into an LRT protocol, that protocol does not run its own validator. Instead, it selects a set of operators—typically 10 to 30 entities—to validate on its behalf. The selection is often made by the protocol’s founding team, a small DAO, or a set of predefined criteria. The user has little to no say in who these operators are. And the same few large staking providers—Coinbase Cloud, Figment, Kiln, and a handful of others—appear on the operator lists of multiple LRT protocols.
I pulled the latest data from Dune dashboards last night. Across the top three LRTs, the overlap is striking: six of the top ten operators serve at least two of these protocols. That means a single slashing event at one operator could cascade across multiple LRTs, punishing ETH that was restaked for completely different AVSs. The contagion risk is real, yet it is almost never discussed in the marketing material.
Furthermore, the incentives for operators need scrutiny. Operators earn fees from the LRT protocol and potentially from the AVSs they help secure. But the transparency around how these fees are distributed is minimal. In many cases, the operator is not directly exposed to the risk of the AVS they are validating for—the risk is borne by the restaker. This is a classic principals–agent misalignment. As I wrote in a piece during the 2020 DeFi summer, ‘When the party securing capital is not the party bearing the loss, the system is fragile.’ Noise filtered. Signal preserved.
Now let me offer the contrarian view. Many proponents argue that liquid restaking actually enhances decentralization by allowing smaller holders to participate. They point out that before restaking, only large stakers could run their own validators; now anyone with 0.1 ETH can contribute. That is true in one sense, but it conflates capital participation with actual operational decentralization. The capital is distributed, but the operational control remains concentrated. This is the same fallacy that undergirded the ‘decentralized’ centralized exchanges of 2022—plenty of users, but a single point of failure.
Trust is the only currency that matters. And right now, the trust we are placing in a handful of operator entities is a fragile foundation for a system that aspires to secure the entire crypto economy.
There is also a deeper, more philosophical issue. The complexity of restacking—multiple AVSs, multiple slashing conditions, multiple reward streams—creates an information asymmetry that favors insiders. The average user cannot evaluate the risk profile of an AVS. They see the yield number and assume it is safe because ‘it’s on EigenLayer.’ During the 2021 NFT boom, I interviewed collectors who bought Bored Apes because they believed the community narrative, not because they understood the art. The same pattern is repeating here: people are buying into the narrative of ‘restaking = passive income’ without understanding the structural risks.
What does this mean for the market? The bull market euphoria is currently masking these flaws. As long as yields keep flowing and no major slashing event occurs, few will question the operator concentration. But I have seen this movie before. In 2017, I warned about token distribution centralization in ICOs long before anyone cared. By the time the flaws became obvious, the damage was done.
Truth over hype. Always.
So where do we go from here? The next narrative shift will likely be toward ‘sovereign restaking’—protocols that allow users to choose their own operators, or even run their own, while still benefiting from liquid restaking tokens. Some projects are already experimenting with this model. But the economic incentives for the average user to run a validator are weak; the overhead of monitoring AVS performance is high. It is more likely that we will see a hybrid model where users delegate to a diversified set of operators, but with more granular control and transparency.
For now, my advice to readers is simple: look at the operator list of the LRT you are using. Check if those same operators appear in other LRTs. Ask how they are selected and what happens if they are slashed. The answers might surprise you.
In a bull market, the safest bets are often the ones that challenge the prevailing narrative. The noise tells you to chase yield. The signal tells you to understand where that yield comes from, and who controls the keys. Based on my experience auditing the structural integrity of protocols, I believe the biggest opportunity in restaking is not the next AVS—it is the infrastructure for transparent, user-controlled delegation. That will be the narrative that survives the next cycle.