Proposal: Add Index Coop's icETH to the DXdao treasury

Author/Co-Authors:

@asira @funkmasterflex @Matthew_Graham

We appreciate the help and insights from @arhat and @sky as well as the entire DXdao community.

Basic Summary

We are requesting feedback from DXdao regarding Index Coop’s Interest Compounding ETH Index (icETH) and propose investing a percentage of the DXdao treasury. icETH gives token holders amplified exposure to the staking yield on Ethereum through a leveraged liquid staking strategy.

Motivation

DXdao could benefit from holding icETH by activating a portion of the ETH based treasury to earn an additional 9.24% interest rate (5.28% increase vs. stETH yield). Live stats can be found here: Dune

Considering DXdao’s alignment building an array of products on top of the Ethereum Ecosystem and the treasury’s significant holdings of both ETH and stETH, icETH is the perfect asset to grow the DXdao Treasury over the long term.

Methodology

Within Aave, icETH deposits Lido’s liquid staked Ethereum token—stETH—as collateral and borrows ETH, which is then swapped for more stETH. That stETH is then deposited as additional collateral in Aave, allowing for more ETH to be borrowed and subsequently swapped for additional stETH. As a result, token holders have spot exposure to ETH and more than twice the yield compared to simply holding stETH.

Risk Assessment

There are a few key factors to note when performing a risk assessment of icETH. While the leverage within icETH may seem risky, the auomtated nature of the token ensures that the leverage ratio is always optimal. In addittion, the tokenized strategy makes the product highly liquid on secondary markets (icETH/ETH Uni V3) as well as plugging into the most liquid markets in DeFi using exchange issuance.

Liquidation Risk

As with any use of leverage, there is liquidation risk if the health of the collateralized debt position were to fall below liquidation thresholds. Though the target leverage ratio is 3.1x, the actual ratio can float within a safe range of 3.0x to 3.3x. An automated keeper system constantly monitors the real leverage ratio of the index and will trigger a hard rebalance if the leverage ratio moves outside of the safe range. For example, if the price of stETH were to suddenly de-peg from ETH and render the real leverage ratio to be 3.4x, the index would recenter to the target leverage ratio of 3.1x. In the case of icETH, the underlying Aave positions would be liquidated if the index’s leverage ratio were to reach 4.0x.

In addition to the primary keeper system, there is a secondary safety mechanism in place called ripcord. If the real leverage ratio were to exceed 3.5x and the primary keeper systems were to fail, the ripcord function could be called to aggressively recenter the index back to the target leverage ratio. This function is publicly callable and incentivized with 1 ETH, adding another layer of permissionless defense against liquidation.

The high correlation between the collateral asset (stETH) and the debt asset (ETH) also significantly lowers liquidation risk. Because their prices move in tandem, there is a more persistent leverage ratio compared to uncorrelated collateral and debt assets like ETH and USDC (the composition of ETH2x-FLI). The inherent price volatility of ETH compared to USDC requires daily rebalancing of ETH2x-FLI to maintain a healthy loan-to-value (LTV) ratio. For icETH, the LTV ratio is more stable and only requires rebalancing every few months. The result is lower liquidation risk and less volatility decay for icETH, enhancing fund safety and preserving NAV.

Interest Rate Risk

A second risk to consider is interest rate risk related to the staking rate for stETH and the borrow rate for ETH. Simply stated, incremental yield can only be generated if the stETH staking rate exceeds the ETH borrow rate. These two factors can change favorably and unfavorably. For stETH, the staking rate can fall as more ETH is deposited into the ETH 2.0 staking contract; this would lead to lower yields on icETH. Post-merge, the staking rate for stETH can also increase, which would increase the effective yield on icETH. For ETH, the variable borrow rate within Aave can increase as ETH utilization increases, compressing the yield on icETH as the borrow rate approaches the stETH staking rate. Conversely, if ETH utilization decreases, the yield on icETH will increase. In the event that the borrow rate were to reach parity with the staking rate, icETH can be deleveraged to 1.0x so that the floor yield is simply the staking rate on stETH. Favorably, Aave recently optimized the rate curve for ETH which will allow for better borrowing conditions over time and a higher potential TVL for this holistic strategy.

Protocol Risk

Set Protocol V2 was audited by OpenZeppelin in September of 2020, launched early October 2020. Additionally, audits from Chainsecurity and Trail of Bits were completed on Set Protocol V1. icETH is built on the same battle-tested infrastructure used to launch over 12 Index Coop products. icETH has been live for nearly three weeks and has over $15M in AUM.

Implementation

The DXdao community can utilize exchange issuance directly through the Index Coop website for large buys. This can be done without slippage, accessing the largest ETH pools in DeFi for no fee. If the treasury wanted to get out of the position, DXdao can opt to redeem the icETH for a fee of 25 bps. This fee is to ensure the deepest liquidity possible. This redeem fee, however, is not the only route DXdao can take if wanting to offload the position. Using the Uni V3 icETH/ETH pool, icETH holders can offload around ~$1M at a time with minimal slippage and no additional fee.

Temperature Check

We would like to gather feedback from the community about this post and invite the DAO to discuss how a treasury diversification could benefit DXdao.