Coinbase: Re-staking — Everything Old Is Reborn

Author: Coinbase Research Translation: Shan Ouba, Golden Finance

Ethereum’s Proof-of-Stake (PoS) consensus mechanism is the largest economic security fund in cryptocurrency, totaling nearly US$112B. But validators who secure the network don’t just earn base rewards through locked ETH. Liquid Staking Tokens (LST) have long been a way for participants to bring their ETH and consensus layer earnings into the DeFi space—to be traded or re-pledged as collateral in other transactions. Now, the advent of re-pledge has introduced another layer in the form of Liquid Re-Pledge Tokens (LRT).

Ethereum’s relatively mature staking infrastructure and excess security budget have enabled EigenLayer to grow into the second-largest DeFi protocol in the ecosystem ($12.4B in total value locked (TVL)). EigenLayer enables validators to earn additional rewards by re-staking ETH to secure Active Validation Services (AVS). As a result, intermediaries in the form of liquid re-staking protocols are also becoming more common, driving the proliferation of LRT.

That said, we believe that re-staking and LRT may pose additional risks compared to existing staking products from a security and financial perspective. These risks may become increasingly opaque as the number of autonomous systems grows and LRT operator strategies diverge. Nonetheless, re-staking (and staking) rewards are laying the foundation for new types of DeFi protocols. If these proposals are implemented, separate discussions around reducing staking issuance to minimum viable issuance (MVI) may also further increase the relative importance of long-term re-staking yields. As a result, the over-focus on re-staking opportunities is becoming one of the biggest crypto themes this year.

Ethereum’s Re-Staking Basis

EigenLayer’s re-staking protocol went live on the Ethereum mainnet in June 2023, with AVS set to roll out in the next phase of its multi-phase deployment (Q2 2024). In effect, EigenLayer’s concept of “re-staking” establishes a way for validators to secure new Ethereum features, such as data availability layers, rollups, bridges, oracles, cross-chain messaging, and more, and potentially earn additional rewards in the process. This represents a new source of revenue for validators in the form of “security as a service.” Why has this become such a hot topic?

As the largest PoS cryptocurrency, ETH currently has a huge economic basis to protect its network from malicious majority attacks. At the same time, however, the continued growth of validators and staked ETH has arguably exceeded what is needed to protect the network. At the time of the merge (September 15, 2022), 13.7 million ETH was staked, roughly enough to secure the network TVL of 22.1 million ETH at the time. As of our upcoming news, there are currently about 31.3 million ETH staked, a threefold increase in ETH terms, but Ethereum's ETH-denominated TVL is actually lower (lower than at the end of 2022) at 14.9 million ETH (see Figure 1).

The excess staked ETH and the security, liquidity, and reliability of the underlying assets make it uniquely positioned to help facilitate the security of other decentralized services. In other words, we believe that re-hypothecation as a concept is largely inevitable as an extension of ETH's inherent value. However, there is no such thing as a free lunch. To ensure the correctness of these services, re-hypothecation is used for behavioral verification and may be subject to withholding or slashing penalties, similar to traditional staking. (That said, slashing will not be enabled when the first set of AVS launches in Q2 2024.) As with staking, re-hypothecation operators receive additional ETH (or AVS tokens) for their services.

Discussion on Re-staking

To date, EigenLayer’s TVL growth has been astounding, second only to Lido (Ethereum’s leading liquid staking protocol). EigenLayer has achieved this while retaining deposit caps for most of its processes, and before launching any live AVS. That said, it is difficult to decouple the ongoing need for re-staking from user interest in short-term credits and airdrop mining. While the amount of ETH re-staking may continue to grow as the protocol matures, we believe that TVL may see a short-term decline when credits mining ends or early AVS rewards are lower than expected.

Other AVS built in the EigenLayer ecosystem include interoperability networks, fast final layers, position proof mechanisms, Cosmos chain security bootstraps, etc. The opportunity space for AVS is extremely broad and growing. Restakers can selectively choose which AVS they want to protect with ETH collateral, although this process becomes increasingly complex for each new AVS.

Potential Issues

This raises the question of how different LRTs will handle (1) AVS selection, (2) potential slashing, and (3) eventual token financialization. In traditional staking, the one-to-one mapping between validator responsibilities and income is clear, making LST a relatively simple matter all things considered. But with re-staking, the many-to-one structure adds some non-trivial complexity (and diversity on the part of LRT issuers) in how gains (and losses) are accumulated and distributed. LRTs pay not only the base ETH staking rewards, but also rewards for earning a set of AVS. This also means that the potential returns paid by different LRT issuers will vary.

Currently, many LRT models are not fully defined. However, since there is only one LRT per project, all token holders in a given protocol are likely to be subject to uniform AVS incentives and slashing conditions. The design of these mechanisms may vary from LRT provider to LRT provider.

One ​​suggestion is to take a tiered approach, whereby LRT issuers could adopt a range of “high” and “low” risk AVS, although this would require the establishment of yet-to-be-defined risk criteria. Furthermore, depending on the architectural design, the final reward for token holders may still be paid out as the sum of all AVS, which we believe defeats the purpose of a risk-tiered framework. Alternatively, a decentralized autonomous organization (DAO) could decide which AVS are chosen, but this raises questions about who the key decision-makers in these DAOs are. Otherwise, LRT providers could act as an interface to EigenLayer and allow users to retain decision-making power over which AVS are adopted.

New Risks

However, at launch, the re-hypothecation process should be relatively simple for operators, as EigenDA will be the only AVS that needs to be secured. However, a feature of EigenLayer is that ETH staked to one AVS can be further re-hypothecated to other AVS. While this can increase returns, it also exacerbates risk. Submitting the same re-staked ETH to multiple AVSes presents challenges when it comes to untangling the hierarchy of slashing and claiming conditions between services. Each service will create its own custom slashing conditions, so a situation could arise where one AVS slashes re-staked ETH for misconduct, while another AVS wants to reclaim the same re-staked ETH as compensation to a damaged participant. This could lead to eventual slashing conflicts, although as mentioned, EigenDA will not have slashing conditions when it first launches. Further complicating this setup is that EigenLayer’s “pool security” model (where AVSs leverage a common pool of staked ETH to secure their services) can be further customized through “vested security”. That is, individual AVSs can acquire (additional) re-staked ETH that is used solely to secure their specific service - a form of insurance or safety net that the AVS pays a premium for. As a result, the role of the operator becomes more technically complex as more AVSs are launched, and the slashing rules become more difficult to follow. In addition to this re-hypothecation complexity, the expansion of LRT also abstracts away many potential strategies and risks from token holders. This is a problem because we believe that people will ultimately go to where these LRT providers offer the highest returns. As a result, LRTs may be incentivized to maximize their yields in order to gain market share, but this may come at the expense of a higher (albeit hidden) risk profile. In other words, we believe that risk-adjusted returns are important, not absolute returns, but it may be difficult to be transparent in this regard. This may lead to additional risk because the LRT DAO would be incentivized to maximize multiple re-hypothecations to remain competitive. In addition, LRT could also create downward selling pressure on non-ETH AVS rewards if LRT spending is entirely in ETH. That is, if LRT needs to convert native AVS tokens to ETH (or ETH equivalent) in order to redistribute rewards to LRT token holders, the value of rehypothecation may be limited by recurring selling pressure.  

Additionally, LRTs carry valuation risks that cannot be ignored. For example, if the queue for staking withdrawals lengthens (validator churn limit was reduced from 14 to 8 after the Ethereum Dencun fork), LRT could temporarily deviate from its underlying value. If LRT becomes a widely accepted form of collateral in DeFi (such as LST in lending protocols), this could inadvertently exacerbate liquidations, especially in low-liquidity markets.

This is assuming that these DeFi protocols are able to properly assess the collateral value of LRT in the first place. In reality, LRT represents different portfolio holdings, and the risk profile of these holdings may change over time. New constituents may be added or removed, or the yield or solvency risk of the AVS itself may change. Hypothetically, we may see a situation where a market downturn could affect multiple AVSs simultaneously, destabilizing LRT and amplifying the risk of forced liquidations and market volatility. Recursive lending would only amplify these losses. On the other hand, protocols that are able to decompose LRT into its principle and yield components can help mitigate this risk to some extent.

Finally, as Ethereum co-founder Vitalik Buterin has emphasized, in some cases, major failures in the re-pledge mechanism could threaten Ethereum's underlying consensus protocol. If the amount of ETH re-staked is large enough relative to all staked ETH, there may be economic incentives to enforce bad decisions that could lead to network instability.

Conclusion

EigenLayer's re-pledge protocol is expected to become a cornerstone for a variety of new services and middleware on Ethereum, which in turn can generate a meaningful source of ETH rewards for validators in the future. AVS from EigenDA to Lagrange can also greatly enrich the Ethereum ecosystem itself.  

That said, the adoption of LRT wrappers around underlying protocols may result in hidden risks due to opaque redistribution strategies or temporary misalignment of the underlying protocol. How different issuers choose which AVS to secure and allocate risks and rewards to LRT holders remains an open question. Additionally, initial yields on AVS may not meet the extremely high expectations set by the market, but we expect this to change over time as AVS adoption grows. Nonetheless, we believe re-hypothecation supports open innovation on Ethereum and will become a core part of the ecosystem’s infrastructure.