en
Back to the list

The risks of restaking are extremely overrated

source-logo  blockworks.co 22 April 2024 20:12, UTC

EigenLayer, the biggest proponent and implementer of restaking, is starting to get rid of its training wheels. After briefly removing caps on staking with liquid staking tokens (LST), total value locked soared from $2.1 billion to 11.5 billion in ETH used for “restaking.”

At its core, restaking is about increasing capital efficiency.

ETH is a tremendously widespread and liquid asset, which makes it ideal for bootstrapping new proof-of-stake (PoS) protocols. The deal is simple: New networks get significant security from day one, while ETH stakers get to earn extra on the same assets they already hold.

Restaking is quickly achieving buzzword status, and it’s largely the cautiousness of the EigenLayer team that is, for now, keeping it relatively tame in terms of total value locked figures. As a relatively new staking mechanism, augmenting existing liquid staking protocols like Lido, Rocket Pool, Frax and others, it promises to unlock billions in additional value in the wider staking industry. But as with any new crypto primitive, there are voices of concern about how restaking might threaten the stability of Ethereum and crypto itself.

Some of these concerns are probably unfounded — for example, concerns about financial stability. Other risks, namely technical, are legitimate but vastly overblown. Critics argue that routing a majority of the ether staked into restaking protocols could lead to an unnecessary and risky stacking of these technical risks. But the truth is, there’s likely more risk in not embracing restaking.

How restaking increases the financial stability of Ethereum

Restaking offers a great way to finally unlock the potential of LSTs, which could increase the security of ETH. It’s important to remember that beyond the convenience, LSTs exist for an important structural reason that directly benefits Ethereum security.

In a nutshell, staking yield on Ethereum competes with DeFi yields. Lending protocols and liquidity pools can offer more significant yield than the approximate 4% of ETH staking. If the average yield for ETH is significantly higher than that figure — which can easily happen in particularly active markets — then only a small amount of ETH supply will be dedicated to staking, making the network that much more fragile.

With LSTs, ETH holders don’t need to choose: They can just earn the baseline ETH staking yield at all times, and beef it up with DeFi yields if they’re adventurous.

Unfortunately, holding LSTs seems to be pretty much the only thing you can do in the current environment. LSTs’ usage in DeFi is mostly limited to enabling easy swaps for regular ETH, while DeFi trading usage is extremely limited — ETH has more than 10x the trading volume of Lido’s wstETH, according to Uniswap analytics.

Restaking comes to the rescue here by giving another potential source of income for ETH holders, which should make staking naturally more competitive with DeFi. The end result is that the network wins, as more ETH is staked.

Are there financial risks in restaking?

It’s important to note that restaking is a strictly technical practice — assets deposited to EigenLayer remain in the system, and are never lent to anyone else. Despite sounding similar to “rehypothecation,” restaking is a completely different mechanism that simply does not offer a financial risk surface. It’s worth noting that, for now, EigenLayer is the only significant player in the Ethereum restaking space and future protocols might offer a different risk profile.

Read more from our opinion section: Restaking is a ticking time bomb

EigenLayer is a decentralized protocol, and there is a risk of losing LST value if it were delegated to a faulty EigenLayer operator. The cumulative risks thus also rely on the community of stakeholders doing their own due diligence, similar to the current LST market offering.

There may be confusion coming from Liquid Restaking Tokens (LRT), which are financialized EigenLayer positions — essentially an LST of ETH LSTs deposited to the EigenLayer protocol. The EigenLayer FAQ mentions LRT liquidations, and the explanation might be doing more harm than good in this case.

In practice, it’s important to note that these risks are completely external to the protocol. If a user deposits their LRT into a lending protocol to enter a leveraged position, their liquidation is a fully external event. While there is an incentive for users to deposit their LRT due to the potential of leveraging yields, this risk level is unlikely to reach catastrophic proportions.

Just as nobody worried about Ethereum or Lido security during the stETH de-peg in 2022, nobody should worry about EigenLayer users getting liquidated. In that case, someone else will take control of their assets, and the system will move on. Furthermore, the stETH de-peg occurred before staked ETH was withdrawable, meaning that no significant arbitrage was possible.

Are technical risks of restaking significant?

Technical concerns about restaking are legitimate. After all, a failure in the EigenLayer protocol (or others like it) could result in a significant loss to the community of ETH holders as a whole. Excessive slashing, loss of control over the stake, malicious apps, etc. are all risks that ETH holders are potentially signing up for when restaking.

But it’s important to contextualize the technical risk. New protocol implementations will always have a risk of technical failure, and that is something that the community needs to understand and mitigate. Similar concerns can be voiced about each upgrade to Ethereum — the Merge itself could’ve gone wrong due to subtle bugs in the implementation.

Constant auditing, bug bounties, training wheels and active protocol monitoring are all parts of a defense in-depth security model that works both to prevent losses and minimize them if they were to occur. Plain old redundancy through competitor solutions can also be used to mitigate risks from poor specific implementations of restaking.

Some might argue that restaking is an unnecessary increase in the complexity of Ethereum’s consensus. That makes sense, but this risk has to be viewed in the context of a general risk of not enough ETH being staked.

Ethereum can be progressively crippled by controlling certain percentages of active stake. At over 33%, Ethereum is unable to finalize. When attackers control over 50% of the stake, they can cause minor censorship and reorganization, while controlling over 66% of the staking power grants them full control. Again, the percentage is defined in terms of the currently active stake. If there’s only 1% of ETH supply in staking, attackers would need to add just another 1% to control 50% of staking power.

With only 26% of the supply currently in staking, it is theoretically possible for an attacker to cripple the network by just acquiring 13% of the ETH total supply to reach 33% of the staked ETH share. That equates to just over $56 billion, which is less than three days’ worth of ETH daily volume. Though there are multiple caveats to this figure — for example, $56 billion of net buying pressure would push ETH price significantly higher — these are surprisingly low numbers for a global and neutral settlement layer.

Restaking and DVT could mitigate consensus risks entirely

The great thing about restaking is that it’s an extremely modular and decentralized architecture. The entirety of the stake is siloed into their respective applications, and part of it is always kept dedicated to only validating Ethereum. It is also quite easy to implement decentralized validator technology (DVT) for the operator sets involved in each EigenLayer service (which they call AVS).

DVT enables splitting control over a validator across multiple entities in a verifiable and cryptographically-secure mechanism. Since validators are responsible for producing and accepting new blocks, they are the source of “power” in the Ethereum network. An attack requires controlling validators, and while controlling the ETH stake is an important aspect of it, entities that accept other people’s stake have more power over the network than expected.

If the validator sets tied to EigenLayer and the associated stakes are safely distributed and managed through a DVT protocol, then risks of total system failure are significantly reduced. That would also lead to a much lower risk of mass validator downtime resulting in large slashing events.

If restaking adoption results in the majority of ETH supply being staked — while being distributed to an organic and decentralized set of validators — Ethereum could be significantly more secure than now.

Like any new idea, restaking protocols will need a long break-in period to ensure their technical safety, but otherwise they are extremely promising as the next major primitive in Ethereum infrastructure — and perhaps even necessary.


Adam Efrima is the SSV Core team Co-founder. He has been active in the crypto industry since 2013. Over eight years living in China working in the financial industry and fintech space, Adam has worked in CITIC Bank covering outbound investments for Chinese SOEs, he was also in charge of setting up the eToro Shanghai operation. Since then, Adam has been deeply involved in Ethereum staking, co-founding top performing staking project Bloxstaking as well as co-founding SSV Network, a decentralized validator infrastructure for ETH staking.
blockworks.co