tldr: Request for papers

Understanding Risk in Restaking

Apply Here

Understanding Risk in Restaking

Problem Statement

There has been a huge explosion in interest in restaking techniques. Restaking allows new blockchain networks to reuse existing validator networks and capital to bootstrap new ones. While such networks have been proposed in the past, most notably within the Polkadot ecosystem, they have found much more capital-market fit than their predecessors in the last 6 months. For example, the largest restaking ecosystem, Eigenlayer, went from $0 to $15B in capital aggregated between January and April 2024. But restaking is not risk-free — validators have to contend with slashing, leverage, and portfolio risks that resemble risk in DeFi.

A successful proposal will answer a plurality of following questions:

  • Can we formalize the risk that restaking protocols impose on their host protocols?
  • What are worse-case and average-case outcomes for restaking protocols?
  • How does leverage within restaking protocols impact their security?
  • What are the centralization risks of restaking protocols? Do those risks impact the base protocol?
  • What do sustainable economics look like for restaking protocols? How much do they need to subsidize usage versus generating real economic fees?
  • How do they pass some of these earnings back to the base protocol? Should they have to do this? What are the benefits/detriments from doing this?

Background

Bootstrapping decentralized networks is generally an expensive and time-consuming procedure. Network participants have to acquire resources, such as energy (for PoW), capital (for PoS), or storage (for PoST), to participate. This initial capital cost for an uncertain payoff that depends on the success of future network usage can be quite high for non-specialized parties with sufficient resources to wait out network profitability. This naturally leads to concentration in node operators who participate in these networks and increases the barriers to entry. In decentralized networks, low barriers to entry are a crucial requirement for ensuring censorship resistance and mechanisms that enhance participation are crucial to ensure censorship resistance and long-term network growth.

Restaking is a set of techniques and mechanisms to allow node operators of existing networks, such as Ethereum or Solana, to reuse their stake to validate new networks. These techniques aim to reduce the barriers to entry and base cost of participation. This allows existing participants to avoid having to acquire a significant new resources to participate in a new network. In a tweet, the core ideas of restaking can be described as, “node operators opt into new slashing rules for new networks and in exchange earn fees from the networks.” Formally, a service is a smart contract that node operators can deposit resources into (e.g. stake for PoS networks) that enforces additional payment and slashing rules. For instance, an Ethereum validator can subscribe to a service that provides oracle price updates from Coinbase. The validator has to do the additional work, beyond participating as a proposer and voter in Ethereum consensus, of submitting prices to the contract. In exchange, the validator earns fees if they provide prices within acceptable SLAs and are slashed if they don’t provide prices on time. This latter slashing rule is enforced in addition to Ethereum slashing rules via the service’s contract.

There are several risks immediately posed by such a construction. For instance, what happens if over 50% of the network opts into a service which subsequently slashes all of its restaked stake. In such a scenario, the base protocol implicitly loses 50% of its staked network supply, as the node operators slashed by the service have no incentive to continue validating Ethereum (as their stake is burned and cannot be withdrawn). This example demonstrates the close coupling between the correlation of slashing events on the base protocol and within a service. A natural question is if there any levers that the base protocol can utilize to limit the overloading of consensus (as per Vitalik Buterin).

On the other hand, numerous DeFi-related risks occurring from users securitizing portfolios of services. The most popular means of using Eigenlayer, much like Ethereum staking, is via liquid (re)staking tokens. These tokens allow users to deposit Ethereum that is delegated to a series of node operators who validate many services and pass through the rewards (less a fee). Liquid restaking, which useful for liquidity and improving ease-of-use for users, provides a series of risk vectors to restaking networks that are more complex than those of standard PoS networks. Qualitative descriptions that enumerate the risks can be found in, e.g., the Eigenlayer liquid restaking risk post and Tarun Chitra and Mike Neuder’s The Risks of LRTs. Are there ways of formalizing these risks and comparing them to the existing DeFi literature?

Finally, we note that the design of services themselves is an open problem. While there are 9 live services on Eigenlayer (as of 5/5/2024), the largest and oldest restaking network, it is still unclear how the economics of restaking networks will work. Unlike standard PoS networks or even Polkadot Parachains, there is not intrinsic yield for service tokens (such as Omni or AltLayer). This is because the design of most restaking protocols passes fees directly to the Ethereum restaked. There have been discussions of dual-token models, which allow for new services to subsidize new operators while redistributing how slashes are distributed amongst restakers and token holders. A natural question here is: what are formal models for reasoning about dual-token models? Can we show there exist (or not) equilibria? How do these models become financially sustainable.

Numerous questions within restaking are lacking formal analysis and guarantees and we hope that this short background has illustrated some of the outstanding questions. As the restaking world is new, this RFP is much more open-ended to allow successful candidates to answer questions about risk and incentives via whatever techniques they believe to be most impactful

Places to Start 

Sponsored By
The Uniswap Foundation has committed up to $3,000,000 over the next 3 years to DeFi Research.

Get Involved

sign up