IIP-141: DG - Launch the Interest Compounding ETH index ($icETH)


IIP: 141

Title: DG - Final - Launch the Interest Compounding ETH Index (icETH)

Status: Proposed

Authors: @allan.g, @afromac, @josephknecht

Created: 11 March 2022


1.0 Simple Summary

The Product Pod would like to propose that the Index Coop launch the Interest Compounding ETH Index (icETH), which will provide enhanced yield on ETH using a leveraged liquid staking strategy.

2.0 Abstract

icETH provides enhanced Ethereum yield by using Lido’s Liquid Staked Ethereum token stETH as collateral to recursively borrow additional wETH through Aave V2 Market and procure more stETH. The token is made possible by the recent acceptance of stETH as collateral on Aave V2 Market.

2.1 Motivation

Investors are continuously seeking better yield opportunities for the largest blue chip assets in crypto, and ETH is no exception. The advent of liquid staking tokens, like stETH from Lido, has enabled investors to earn a competitive yield on ETH compared to market rates across the largest money market protocols like Compound or Aave. There is almost $6B invested in the stETH contract which is a testament to investors’ desire to attain yield.

With the recent listing of stETH on Aave V2 Market, Index Coop can launch an automated product that utilizes productive stETH as collateral and wETH as low-interest debt. This strategy is expected to generate >13% APY, nominated in ETH, subject to wETH borrowing rates. This is more than x2.5 the current yield of stETH and of Yearn’s $527M stETH vault.

2.2 Rationale

ETH is the most prominent and most liquid asset in DeFi. It functions as a utility token, a medium of exchange, an interest bearing asset, and store of value (amongst other things). For this reason, it is arguably the most investable digital asset in crypto today. With a market cap of $310B1 and an average annual return of 194%2 over the last five years, there are many investors eager to grow or compound their ETH holdings.

However, investors today have few options for maximizing the return on their ETH. Lending rates paid on the largest money market protocols are negligible, and more sophisticated strategies like Yearn vaults or leverage protocols can be complicated and cumbersome for the average investor. Staking is arguably the best way to earn yield on ETH, and tokens like stETH simplify the staking process by offering a tokenized, interest-bearing asset that is freely tradeable; however, stETH is only paying 4.4% APR at time of writing, and that rate has fallen gradually over time as more ETH is staked.

With the use of Set’s battle-tested leverage token infrastructure, the Index Coop can create a stETH yield product that is more than x2.5 the return of stETH in a low risk fashion. By tokenizing this simple strategy, we can guarantee outsized returns to token holders and offer the highest scalable yield on ETH in DeFi today.

icETH delivers value to customers by…

  • Offering outsized returns on ETH in a low risk fashion
  • Abstracting the management of collateral and debt balances
  • Socializing gas costs associated with maintenance

icETH delivers value to the Index Coop by…

  • Utilizing existing Set infrastructure and minimizing GTM timeline
  • Diversifying our product offering
  • Expanding the yield-bearing category of index products

3.1 Overview

icETH will utilize Set’s battle-tested leverage token infrastructure on Ethereum main net and integrate with Aave v2 for managing collateral and debt positions. stETH will be deposited to Aave v2 and enabled as collateral, allowing ETH to be borrowed and subsequently swapped for more stETH; this new stETH will be added to the index’s collateral balance in Aave v2, allowing for additional ETH to be borrow and swapped for additional stETH. The index will target a leverage ratio of 3.0x, with minimal rebalancing required because of the high, favorable correlation between the collateral and debt assets. Unlike the FLI tokens that are subject to significant currency risk, the leverage used for icETH will be much lower risk because the fluctuations in debt and collateral value will be symmetrical. Emergency deleveraging will be enabled with incentivized ripcord functionality to prevent liquidation during black swan events.

3.2 Differentiation

DeFi users have a strong preference for yield generating strategies. By providing a broad diversity of yield products, we create an ecosystem for our users to engage in and build a portfolio that is appropriate for their needs.

Product Underlying Asset Composition Yield
icETH ETH Single 10-13%
PINT USDC, DAI, USDT Diversified 3-6%
PAY Stables Diversified 10-12%
FIXED Stables Single 7-9%
CCY (aka BASIS) Various Assets Single 15-20%

There are several competing ETH yield offerings on the market today:

Provider Token / Vault Yield (pre-fee) Fees Effective Yield
Lido stETH 4.9% 10% of Staking Reward 4.4%
Yearn yvsteCRV 9.6% 2% AUM + 20% of Yield 5.7%
Beverage ETHMAXY 13% 1.95% of AUM 11.05%
Index Coop icETH 13% 0.90% of AUM 12.10%

icETH will be the lowest cost, highest effective yield on ETH when launched by the Index Coop. It will also draw from the deepest liquidity available for stETH on ethereum main net, minimizing the NAV decay over time brought about by slippage during rebalancing.

3.3 Example composition

At the target leverage ratio of 3.25x, the composition will be ~60% astETH (collateral) and ~40% amETH (debt). Exact composition may vary as the collateral balance gradually grows in value and effectively delevers over time. Automated rebalancing will recenter the index to the target leverage ratio according to predefined parameters.

Assuming the stETH APY = 4.4%, stETH Lend APY = 0%, and ETH Borrow APY = 0.65%, this index will return 12.6% annualized before fees, more than 2.5x the current rate earned by simply holding stETH. The realized yield will likely increase post-merge when MEV begins accruing to stakers / validators.

3.4 Backtest results

icETH backtest results

Backtest calculations were performed for the period 01 Jan 2021 to 15 Feb 2022. The calculations were done for stETH APYs of 0%, 2.0%, 3.0%, and 4.3%. The latter two are shown above alongside ETH. The effective starting LTV was 67.5% and the effective leverage 3x.

For the test period stETH (with APY 4.3%) returned 362% vs 297% for ETH. Note that icETH may underperform ETH over short time periods because of the stETH:ETH volatility. This volatility is substantially lower in the latter part of the test period.

4.0 Size of Opportunity

Yearn’s stETH vault is worth ~$500M and has generated $21.3M in yield for investors. Balancer offers a relatively new stETH:wETH pool with 4.72% APY that has $291M in TVL.

5.0 Market & Customer Research

5.1 Target Customers

icETH is a relevant product for all customer types that the Coop currently serves: retail, DAO treasuries, and institutions. Because ETH is a core component of almost every crypto portfolio, all of these investor types stand to gain from a low-risk, high yield ETH product. During DAO Treasury research, an ETH yield token was the most requested non-stablecoin product, and many treasuries today are seeking to deploy their ETH productively and earn a competitive yield.

Regardless of customer type, this product will deliver an enhanced return on ETH that is unavailable elsewhere in DeFi and as accessible as buying and holding a token!

5.2 User stories

  • As a DAO, I want to earn yield on the ETH in my treasury
  • As a DAO, I want to diversify into a productive form of ETH
  • As a retail user, I want to earn maximum yield on my ETH
  • As an institution, I want to earn maximum yield on my ETH

5.3 Product economics

We predict monthly revenue of $1,583 and negligible rebalancing costs for a gross profit margin of effectively 100%. This is based on a $2M AUM, 0.95% streaming fee, and rebalancing every 4 months to maintain the minimum leverage ratio.

5.4 Product financial forecast

6.0 Methodology

6.1 Initial Composition & Token Inclusion Criteria

icETH will be composed of, and limited to, astETH and aeETH tokens.

6.2 Weightings

icETH will have a dynamic composition of astETH and aeETH tokens with a target weighting of 60% astETH / 40% aeETH. Thresholds are defined for automated rebalancing when collateral or debt positions must be adjusted to maintain a safe leverage ratio.

6.3 On-Chain liquidity analysis of underlying tokens

stETH liquidity is currently distributed amongst the following DEXs / Pools:

6.4 Maintenance / Rebalancing frequency

Because of the high correlation / low currency risk of the underlying assets, this index will only rebalance when the leverage ratio moves outside of the safe range defined below. This should lead to minimal rebalancing on a recurring basis and lower operating costs on Ethereum main net, as well as minimal NAV decay due to rebalancing slippage. Based on prevailing staking and borrow rates, automated rebalancing would occur once every 4-6 months.

Parameters:

  • Target LR: 3.25
  • Max LR: 3.35
  • Min LR: 3.15
  • Rebalance Interval: triggered when current LR moves outside of range above
  • Recentering Speed: 100%
  • Slippage Tolerance: 1%
  • DEXs for Rebalancing: Curve, Balancer
  • Ripcord LR: 3.5
  • Ripcord Slippage: 5%

These parameters may be revised closer to launch. Post-launch, the Product Pod will monitor and manage these parameters and publicly communicate any material changes 24 hours before executing said changes. An example would be updating the LR parameters if borrowing conditions within Aave change drastically, requiring a winding down of the debt position to maintain an acceptable yield.

7.0 Costs

7.1 Cost to customer

Holders will pay a 0.90% streaming fee, and there will be a mint fee of 0.15% and a redeem fee of 0.35%.

7.2 Cost transparency

There are implicit costs associated with this strategy (though they are excluded from the effective yield communicated in this proposal):

  • Borrow costs from ETH debt balance in Aave v2
  • Slippage incurred during rebalancing

The debt balance the product will carry within Aave v2 is subject to a variable interest rate. While that interest rate is currently and historically very low (0.65% for WETH), a very large increase in utilization or a very large decrease in supply could drive the variable borrow rate higher and thus reduce the effective yield on stETH.

NAV decay due to slippage is expected to be negligible due to the very infrequent rebalancing to maintain the target leverage ratio as well as the deep, concentrated stETH liquidity on Curve.

7.3 Fee split

Streaming fees will go 100% to Index Coop.

8.0 Meta / intrinsic productivity

Token holders stand to benefit from the intrinsic productivity enabled by the underlying, interest-bearing stETH positions held within Aave V2 Market.

There are no metagovernance rights available with this product as it holds stable coins and stablecoin equivalents.

9.0 Liquidity

This product will require approximately $76,000 in seed liquidity at launch in order to provide a target access cost of 1% based on the assumptions below which are subject to market conditions. It is highly likely that this product will naturally attract LPs, given that icETH and ETH will not diverge in price quickly.

Target Access Cost % 1.00%
ETH/USD Price $2,596
Gas Price 40.82 gwei
Exchange Issuance Gas Required 600000 wei
Exchange Issuance Gas Cost $64
Total Baseline Liquidity $3,162,582
Concentration - Lower Bound -5%
Concentration - Upper Bound 5%
ETH Required 14.68
Tokens Required 381.09
Total USD Required $76,218

10.0 Author Background and Commitment

The Product Pod is responsible for designing, developing, and deploying index products for the Index Coop, as well as managing and maintaining products post-launch. Core team members of the Product Pod have also been supporting, operating, and refining Set’s leverage token infrastructure for over a year.

11.0 Marketing support / distribution / partnerships

As an internally-produced product for the Index Coop, there will be a joint effort amongst Product, BD, and Growth to market, distribute, and maximize composability for this product.

Option to Extend

If approved by Index Coop governance, this suite proposal can be extended to other products that adhere to the same methodology (ex. icMATIC, icBTC, other liquid staking tokens).

Revision history

  • Streaming Fee changed to 1.95% from 0.95%; M/R Fee changed to 0%
  • Streaming Fee changed to 0.90% from 1.95%; M/R Fee changed to 0.15% and 0.35%, respectively
  • Streaming Fee split for product pod removed, per IIP-167

Copyright

Copyright and related rights waived via CC0.

20 Likes

@sixtykeys we have completed the community call and finalized the PRD for icETH with engineering.

Could you please assign an IIP number to this proposal and schedule a snapshot vote for Thursday, March 16th?

2 Likes

gm @allan.g -

We’ll get this reviewed today, get an IIP number assigned, and get the status changed to proposed to kick-off the 48-hour discussion-period (barring any open-items). Given there isn’t enough time to hit your requested vote-start time, would you prefer that the vote be run asap (48-hours following review, approximately 18MAR22) or with the weekly cadence that will go up on Monday 21MAR22 at 6pm UTC? Thanks.

@DocHabanero @allan.g is this the first product going to IIP with the new Product pod fee split included? Apollogies if I missed it somewhere else.

hey @mel.eth - please queue up the vote as soon as possible (sounds like Friday, March 18th)

Yes that is correct - this will be the first product to be launched* that factors in the internal fee split. However, other products have already been voted on that factor in the internal split.

*updated in accordance with Joseph’s point below

1 Like

Correction: Both the LAYER1 and FIXED snapshots have the new internal fee split.

2 Likes

gm @allan.g -

This proposal has been reviewed and queued for voting to start on snapshot on 18MAR22 at 6:00PM UTC. Voting will run for 72-hours and is subject to the elevated quorum (10%) and passing-threshold (60% FOR) requirements for product launches. Of note, the PRD that was previously linked within the snapshot proposals is available from Product Nest by request (please reach out to @allan.g to request a copy of the relevant PRD for this proposal).

2 Likes

Hi @allan.g

What are your thoughts here if people want to vote FOR the product but AGAINST the internal fee split?

I think it would be more prudent to have separate IIPs and then retrospective add internal fee split pending it passing vote.

@mel.eth what are your thoughts on combining what could be a contentious topic into a product people want. It might put a number of voters between a rock and a hard place.

5 Likes

Great feedback @Matthew_Graham

Given that fee-split agreements are by standard included in the Product Launch vote and typically an active discussion point through the product life-cycle, my recommendation given this vote is now live would be to vote on the holistic package. Some insight that might be helpful: product launch votes have a high barrier to passage (10% quorum, 60% FOR) but fee agreements have a lower bar (5% quorum, 50% FOR) to adjust, currently. Might be worth consideration if looking to launch and then fine-tune at a later time, subject all other governance considerations. This is all just information and I hope it’s helpful. Thanks for highlighting this as a specific consideration in your decision-process as I’d love a future product launch process update to address this point directly if possible.

1 Like

Love the product innovation but would love to understand better. Could you elaborate on the potential risks associated with the levered return while providing liquidation and decay context ? I ask because of the wrapping or looping effect.

In addition, why take this approach vs the following example or something similar…
-buy eth
-borrow stables against it (ex. 20-30% LTV)
-earn return on stables

3 Likes

Hi @Eric_Tomaszewski

Thanks for the question.

As long as stETH maintains its peg to ETH there is zero risk of liquidation, and negligible risk of decay. Over time, the product will gradually de-lever as the deposit asset (stETH) gains value against the borrow asset (ETH), this will lead to a slight tapering of leveraged yield over time - but the strategy will re-lever at set intervals.

The main difference to the strategy you are proposing is the denomination of the yield. In the case of icETH, the return will be denominated in stETH.

Another difference (as mentioned above) is the lack of price divergence between the deposit asset and the borrow asset. Depositing a volatile asset like ETH and borrowing stables against it introduces risk that is not inherent in a strategy where one asset price is pegged to the other.

A final difference, is that icETH will automatically rebalance at specific thresholds and gas costs will be absorbed by the IC. This saves the customer some expensive and any need to manually reposition the strategy.

One potential risk with the strategy is the borrow asset interest rate (ETH in this case) could rise and decrease the realised yield of the strategy. This has already happened to some degree, as ETH borrowed on Aave has increased likely due to investors manually recreating the strategy of the product.

6 Likes

Hi @Matthew_Graham

Thanks for your thoughtful suggestion. This will be the 4th product to have been voted on with the fee split clearly defined as including a share to the product pod. Historically, all product IIPs have included the fee split with no need for additional governance interventions.

Also of note, this is not the first product vote to include a split to IC contributors.

Checking the voting record, on IIP-102 the delegate account you control voted to award a proportion of the fee split to an IC contributor. What has changed in your mind since then?

Moreso, unincentivized rewards are deviation from the norm/best practices. Product has kick-started the process, but now is the time to begin discussing the mechanism to extend incentivized rewards to other Nests in IC. This is how we can attract and retain the best talent, and ensure each Nest is laser focused on achieving the results that will take us to $2B TVL.

7 Likes

Let me mention that this topic (“Roll out performance-based compensation”) is on the Index Council Agenda Queue. I encourage the contributor community to upvote (or downvote) this topic if they’d like the Council to take it up (or not).

3 Likes

Hi Matt, I appreciate the feedback. Fee splits have always been include in the product proposal and approved at the same time as the product. We decided against doing an IIP for the whole framework because the fee split can change based on where most of the effort was completed. For example, some products may rely heavily on a quantitative analyst, so they would receive a larger fee split compared to the product designer.

If you have recommendations to improve the process I’m happy to work through those. Keep in mind we used to pay 40% to external methodologists while this fee split allows us to build the same products at 1/4 the cost (in fee splits).

3 Likes

Gm @allan.g , @JosephKnecht , @afromac

It is with great pleasure that I inform you that your proposal has PASSED with 395k ( 94.61%) INDEX voting FOR! Thank you for trusting Index Coop with your ideas; we’ll now endeavor to expeditiously launch and execute your product vision.

Well done! Please update this space when you have a launch date. :owl:

image

10 Likes