Staking Stablecoins — Assessing the Economic Security of Stablecoin Collateral in Restaking

Takeaways

In this research blog, we examine the role that stablecoins, like USDC, can play in supporting cryptoeconomic security within restaking.

• As more (re)staking platforms support staking with arbitrary ERC-20s, stablecoins are becoming a viable means for providing economic security for new networks such as AVSs.
• Our previous work showed how cryptoeconomic security is important to defend against corruption attacks and that the type of collateral asset and portfolio composition can directly improve economic security for the network.
• As each network has a finite budget for reliable economic security, adding stable cryptocurrencies such as USDC in a network’s collateral portfolio directly reduces portfolio volatility and improves network security while potentially improving risk-adjusted profitability for restakers.

Recent developments in ERC-20 Restaking

Restaking was built upon the idea of reusing Ethereum’s economic security and decentralized computing power to support new off-chain networks that share security with an existing Proof of Stake (PoS) network. Recently, major restaking platforms have launched or announced support for ERC-20 staking (listed in alphabetical order):

• Eigenlayer announced Permissionless Token Support and introduced operator sets with the new Eigenlayer Security Model.
• Jito on Solana announced support for restaking any/multiple SPL tokens
• Karak announced multi-asset restaking at launch
• Symbiotic announced Permissionless Restaking with flexible collateral at launch

This evolution means services now have more options — such as stablecoins — to choose from for economic security. Widely used stablecoins like USDC have large user bases, deep market liquidity, and integrations with major DeFi protocols. Additionally, stablecoin security can be sourced from a decentralized group of stakers. These benefits suggest that including stablecoins as restaking collateral can improve network security.

While stablecoin collateral is inherently good for networks and services, it is not obvious that node operators, who are placing capital at risk to earn fees, would prefer this type of collateral. Naturally, stablecoins don’t inherently have outsized returns like most (unstable) restaking collateral. Since operators are inherently holding long collateral positions, the opportunity cost (i.e. lost yield relative to holding unstable assets) they face might be too high to justify using stablecoins in restaking protocols. As such, it is natural to ask what benefits stablecoin collateral can provide to node operators and other network service providers. In this note, we argue that there are a number of circumstances where the addition of stablecoins increases the risk-adjusted yield for node operators while simultaneously lowering the volatility in the amount of security held by networks. A substantial improvement in risk-adjusted return (measured by, e.g., Sharpe ratio) is a tangible benefit for node operators that we believe will incentivize them to prefer dual-staking with a combination of stable and risky assets in the long run.

Economic Security

To understand the value of stablecoins as staking collateral, let’s expand on the collateral health framework built in our previous research blog. We break down the implications for stablecoin collateral in both a single-asset staking environment and a multi-asset staking environment.

Single-asset staking

To recap, in our recent blog on Restaking Collateral Health, we argued that a network’s economic security is tied to the market value and liquidity profile of its collateral asset(s). When a network’s collateral portfolio experiences drawdown, attackers may have a profitable attempt to corrupt the network. Mathematically, we can calculate its boundary as a function of the attacker’s profit \(\pi\) (assumed as a constant), the network’s security threshold \(\alpha\) and USD value of current stake \(\Sigma_V\), and the volatility of the collateral, represented by \(\text{Var}[\gamma]\) where \(\gamma\) is collateral price returns.

\[P(corrupt) \leq 1 -  \frac{(\frac{1-\alpha_s}{\alpha_s} \frac{\pi}{\Sigma_V})^2}{Var[\gamma]+(\frac{1-\alpha_s}{\alpha_s} \frac{\pi}{\Sigma_V})^2}.\]

A high-level interpretation of this result is that the risk of corruption can increase with the volatility of the collateral. This means that low-volatility stablecoins, such as USDC, are the most robust choice against corruption attacks in a single-collateral setting.

Multi-asset staking

Multi-staking is a new primitive in development that would allow a network to be secured by a set of collateral types. These collateral types would not be limited to just ETH and its associated Liquid Staking Tokens (LSTs). To understand the value of stablecoin collateral in a multi-staked portfolio, let’s extend our framework and assume that \({n>1}\) collateral assets are used to secure a network. For example, dual staking \({(n=2)}\) typically adopts a native asset and a secondary asset.

To analyze dual staking, we consider a few different models.

Model 1: Modular dual staking

In modular dual/multi-staking, two separate quorums are staked with the base collateral (e.g., ETH/LST) and a stable collateral, respectively. The two quorums function independently of each other, and network consensus is reached if and only if both quorums approve the same state transition (e.g., batch of transactions).

In this model, the network’s uptime depends on the least secure quorum, while the network’s security depends on all quorums combined. This is the weakest-link problem explained in Restaking Collateral Health. In the case where the base collateral loses significant value, the stable quorum adds a guarantee that the network cannot be manipulated (even in worst-case scenarios).

\[P(corrupt) = \min_{i = 1,\dots, n} \left\{P\left(\frac{\pi_s(p_i)}{p_i}> \sum_{v\in V_s\backslash B} \sigma_{v,i}\right) \right\}\]

Model 2: Price-weighted dual staking

In a price-weighted multi-staking model, all collateral assets are converted to their dollar value. This allows for fair valuation of operators' stakes, regardless of the asset type they are staking. Note that the staking contract must be connected to an oracle to quote the real-time price of each collateral asset.

Attacker’s conditions:

\[\begin{align*}\text{profitability: }&&\pi_s(\mathbf{p}) &> \Sigma_B \cdot (\mathbf{p})\\ \text{feasibility: }&&\Sigma_B \cdot \mathbf{p} &\geq \alpha_s (\Sigma_V \cdot \mathbf{p} + \Sigma_B \cdot \mathbf{p})\end{align*}\]

where:

• \(\pi_s\) represents the profit from attacking the network (in USD terms)
• \(\Sigma_B=\sum_{v\in B}\mathbf{\sigma}_v\) is the total amount of collateral held by the attacker
• \(\Sigma_V=\sum_{v\in V_s\backslash B}\mathbf{\sigma}_v\) is the total amount of collateral held by good faith operators in the network
• \(\alpha_s\) is the network’s security threshold (e.g. 50% for a longest-chain protocol or 33% for a Byzantine Fault Tolerant protocol)

We derive the security condition for a network with multi-collateral staked:

\[\pi_s(\mathbf{p}) > \frac{\alpha_s}{1-\alpha_s} \Sigma_V \cdot \mathbf{p}\]

corruption risk is therefore measured by

\[\begin{align*} P(corrupt) &= P\left( \pi_s(\mathbf{p}) > \frac{\alpha_s}{1-\alpha_s} \Sigma_V \cdot \mathbf{p} \right)\\ &= P\left( \pi_s(\mathbf{p}) > \frac{\alpha_s}{1-\alpha_s} \sum_{v\in V_s\backslash B} \sum_{i=1}^n \sigma_{v,i} \cdot p_i \right) \end{align*}\]

This result shows that, for a multi-staking network, networks must consider the volatility of the collateral portfolio as a whole instead of individual token volatility. The results above inform us of how multi-staking with a stable collateral (e.g., USDC) can lower portfolio volatility, making the network more economically secure compared with staking with volatile assets such as ETH or native network utility tokens. Even in cases where ETH experiences large drawdowns (like the 22.5% drawdown ETH experienced on August 4th, 2024), stablecoin collateral can still defend economic security by making corruption attacks prohibitively expensive. This is a unique feature of stablecoin collateral that is not provided by any other collateral type.

Yield

The analysis above examines features that make stablecoins premium collateral for networks in both single-asset and multi-asset staking environments. Naturally, though, node operators may be incurring opportunity costs should they provide stablecoin collateral since they are not exposed to market returns they otherwise would accrue had they staked non-stablecoin collateral. There are a number of situations, however, where the benefits in risk-adjusted yield to the node operator adequately compensate for its incurred opportunity costs.

Let’s consider the restaking market as a market for collateral. In this decentralized matching market (see Tarun Chitra’s keynote at Restaking Day for details), both the demand side (networks) and the supply side (staker-operators) have an impact on the market price (yield) for collateral.

On the demand side, balancing the costs and benefits of onboarding different collateral types constitutes a complex optimization problem that is specific to each network. We can assume a network wants to minimize its security cost given a target security level. Let’s denote the market yield for each collateral asset as \({\mathbf{\delta} = [\delta_1, \dots, \delta_n]^T}\), and its security level as a maximum probability of corruption \({P^*}\). Mathematically:

\[\begin{align*}\min_{\mathbf{\Sigma}} \, & \mathbf{\Sigma}^T \mathbf{\delta} \\\text{s.t.} \quad & P(\text{corrupt}) \leq P^*\end{align*}\]

A high-level rationale is that networks should choose their collateral based on their target security level. According to our analysis, networks aiming for greater economic security or lower correlation to ETH would likely be willing to offer higher yields on stablecoin collateral, assuming other costs to the network to accept stablecoins are not overly prohibitive. There are a couple of scenarios that lead to sustainable node operator participation.

Stablecoin staking yield exceeds alternative yield

On the supply side, we can consider stablecoin holders choosing between restaking and alternative yield opportunities such as lending. We reference a few major USDC lending pools on ETH mainnet:

We observe that the current market APY for USDC ranges from 8.18% to 12.97%, with some short-term fluctuations. This implies that if networks can provide yield that exceeds lending pool yield, node operators should be willing to stake USDC.

Risky asset yield is not sufficiently larger than stablecoin yield

As an explicit example, suppose a network accepted a risky asset and a stablecoin in a multi-asset staking quorum, where the expected yield paid to the risky asset is higher than the expected yield paid to the stablecoin. However, the yield paid to the risky asset is also more volatile. Then, if the node operator has simple quadratic utility, and the following hold — risk-adjusted yield (sharpe ratio) of the risky asset is less than some multiple of the stable yield.

Then, maximizing utility involves a nontrivial allocation to USDC.

Other major factors impacting the supply curve include:

• Slashing risk: After slashing goes live on mainnet, stakers will need to evaluate whether the risk of slashing justifies the yield offered by networks.
• Stablecoin LRTs: The creation of stablecoin LRTs would unlock capital efficiency on staked stablecoins, making the sector much more appealing to stablecoin holders.
• Risk premium versus ETH staking: Given the potential differing slashing profile for staking stablecoins vs. ETH/LSTs, instead of staking stablecoins directly, stakers may consider using stablecoins to purchase and stake ETH/LSTs instead. We can model this decision by considering a restaker constructing a delta-neutral portfolio hedging their (re)staked ETH vs. directly staking stablecoins. The cost of such a hedge reflects the risk discount or premium of stablecoin staking. If the yield resulting from stablecoin staking exceeds the ETH yield less hedging costs, then delta-neutral stakers naturally gravitate towards stablecoin staking.

Addressing the Opportunity Cost Problem

For most AVSs in the bootstrapping phase, paying a 8.18% yield for the added security that stablecoins provide is likely unsustainable until they begin accruing revenue. To address this, we propose two separate paths, each with its own unique trade-offs.

Yield Before Restaking

Given the expansion of restaking platforms into permissionless collateral deployment, some AVSs may instead prefer to use a yield-bearing stablecoin or vault as collateral. With this approach, instead of using a plain USDC as collateral, the network accepts a tokenized position, wrapping a yield-generating USDC deposit. This wrapper would accrue the underlying yield while gaining additional restaking yield.

For the AVS, this reduces the cost of attracting this collateral but exposes them to potential drawdowns in value stemming from the risks assumed when gaining the wrapped yield. For example, the USDC rates mentioned above come from lending platforms. A wrapped lending position is exposed to risks of socialized losses from unprofitable liquidations or an inability to redeem the position from high borrow utilization. Additionally, wrapping USDC prior to depositing creates a poor staker UX.

Restaking Before Yield

Alternatively, AVSs can choose to accept non-yield-generating stablecoins as collateral on the assumption that an LRT built on top of that staking position will present yield opportunities that justify the opportunity cost. With this approach, AVSs pay a much lower yield for the stablecoin collateral, but yield opportunities for the LRT (via lending, points, incentives, etc) present additional yield. For this approach to work, the AVS risk-adjusted yield + the LRT risk-adjusted onchain yield must be greater than the base rate for the collateral.

This presents AVSs with all the benefits of using a stablecoin as collateral, but exposes the network to the risk of a highly concentrated depositor pool. High TVL LRTs concentrate stake to a few decision-makers, which can increase the probability of corruption and volatility of stake for the AVS.

Conclusion

As more platforms support staked ERC-20 tokens for economic security, networks face a critical decision on which collateral asset(s) to secure their operations with. By formulating the level of security from stablecoins in both single-collateral and multi-collateral settings, we show that stablecoins such as USDC would be premium collateral assets that offer high security. We also show that node operators benefit from providing stablecoin collateral by increasing their risk-adjusted yield. Put together, the addition of stablecoin collateral can lead to more sustainable economic security for networks.