[ARFC] Add fUSDC to Ethereum v3

[MarcZeller]

Proposal updated to include Risk teams feedback

---

title: [ARFC] Add fUSDC to Ethereum v3
author: @marczeller - Aave-Chan Initiative
created: 2023-05-23

Summary

This ARFC presents the community with the opportunity to add fUSDC to the Ethereum v3 Liquidity Pool.

Abstract

fUSDC is a yield-bearing token representing USDC deposits in Flux, a fork of Compound V2. Adding fUSDC to Ethereum v3 would allow fUSDC holders to borrow stablecoins on Aave and leverage their fUSDC position, boosting the stablecoin utilization rate on Aave, while attracting new stablecoin deposits thanks to boosted supply rates.

Motivation

fUSDC is a new financial primitive with arguably the best risk-adjusted yield available in DeFi. Adding support for fUSDC on Ethereum V3 in isolated mode would allow fUSDC holders to borrow stablecoins on Aave and leverage their fUSDC position, boosting the stablecoin utilization rate on Aave, while attracting new stablecoin deposits thanks to boosted supply rates.

Specification

Ticker: fUSDC

Contract Address: [0x465a5a630482f3abd6d3b84b39b29b07214d19e5](https://etherscan.io/address/0x465a5a630482f3abd6d3b84b39b29b07214d19e5)

Parameter	Value
Isolation Mode	Yes
Borrowable	No
Collateral Enabled	Yes
Supply Cap	220M (~$4.45M)
Borrow Cap	N/A
Debt Ceiling	$1M
LTV	77%
LT	79%
Liquidation Bonus	4.5%
Liquidation Protocol Fee	20%
Variable Base	0.00%
Variable Slope1	4.00%
Variable Slope2	60.00%
Uoptimal	90.00%
Reserve Factor	10.00%
Stable Borrowing	Disabled

Next Steps

• If the ARFC snapshot vote passes, move the proposal to the Aave Improvement Proposal (AIP) stage for final approval and implementation.

Disclaimer

@marczeller and the Aave-Chan Initiative are not associated with or compensated by Flux Finance for publishing this AFRC.

Copyright

Copyright and related rights waived via CC0.

[Gauntlet]

Gauntlet recommendation on fUSDC

fUSDC represents claims to USDC deposits on Flux Finance, a Compound v2 fork where the only acceptable collateral is OUSG, a tokenzied implementation of short term US Treasuries that directly invests in the ETF iShares SHV. fUSDC is analagous to cUSDC on Compound v2. That being said, Gauntlet cannot quantify the above RWA risk; the below parameters are determined from the on chain characteristics of fUSDC only.

We do not recommend setting fUSDC in isolation mode, since this could prevent the natural behavior of fUSDC collateral from evolving without significantly reducing risk.

Risk Parameter	Gauntlet Rec
Isolation Mode	NO
Enable Borrow	NO
Enable Collateral	YES
Borrowable in Isolation	NO
Loan To Value	74%
Liquidation Threshold	76%
Liquidation Bonus	4.5%
Reserve Factor	10%
Liquidation Protocol Fee	20%
Borrow Cap*	N/A
Supply Cap*	270M ($5.4M)
Debt Ceiling	N/A
Base	0%
Slope1	4%
Uoptimal	90%
Slope2	60%

*1 USDC = 49.4 fUSDC

Isolation Mode
We do not recommend setting fUSDC in isolation mode, since this could prevent the natural behavior of fUSDC collateral from evolving without significantly reducing risk.

LTV, LT, LB, RF, LPF, IR
As stated above, fUSDC represents claims to USDC deposits on Flux Finance. Utilization for USDC (as well as USDT and DAI) on Flux all hover at the kink of 90%. fUSDC is directly convertible to USDC via Flux Finance in normal situations (when USDC utilization on Flux Finance < 1). We recommend using the same LTV, LT, LB, RF, LPF, and IR curve parameters (Slope 1, Uopt, Slope 2) as USDC.

Supply Cap
fUSDC is unique in that it has no external liquidity and its only exit is to be redeemed for USDC on Flux Finance. We recommend initializing supply cap at 25% of the circulating supply (~$22M has been minted on Flux Finance), and will monitor usage and growth before recommending subsequent increases.

Next steps

Should community approve the snapshot vote to list fUSDC, Gauntlet will follow up with all the parameters for next steps.

[MathisGD]

Hello @Gauntlet,

The current liquidity of fUSDC (instantly redeemable USDC) is about $2M USDC, and the proposed supply cap is $5.4M. The numbers look ok for the moment, but this could quickly change if the fUSDC market happens to see big withdrawals (NB: the 3 biggest suppliers account for 70% of the total supply). So listing fUSDC with the proposed parameters would mean that now the fUSDC market needs to be closely monitored, and that the Aave governance must be ready to react quickly on any big Flux market/governance moves. On the other hand listing fUSDC in isolation mode would reduce by a lot the risk associated with this collateral, as fUSDC is supposed to be pegged to other stablecoins.

[ChaosLabs]

Chaos Risk Analysis

Alongside the benefits of listing fUSDC discussed in the initial TEMP-CHECK and above, it is important to highlight the risks associated with listing fUSDC as collateral on Aave.

Given the fact that fUSDC is not tradeable, liquidations rely on the availability of USDC liquidity on the Flux protocol. The availability of liquidity could be compromised due to any of the following risks materializing:

1. Exposure to OUSG - Since USDC on Flux is borrowed against OUSG collateral, Aave is exposed to the risks associated with this “permissioned” asset, including Smart Contract Risk, Centralization Risk, and any unexpected issues regarding the underlying bond, which could influence the stability and reliability of OUSG as collateral.
2. Exposure to Default Risk of Flux Finance Protocol:
   By listing fUSDC as collateral on Aave, there is an inherent risk of default by the Flux Finance protocol, as this token is a receipt for the underlying assets locked in the Flux protocol.

Lack of Historical Data:
Flux Finance and OUSG are relatively nascent (the first transaction of this token on Ethereum was on 01/26/23 - https://etherscan.io/token/0x1b19c19393e2d034d8ff31ff34c81252fcbbee92), and the absence of historical data should be noted. The protocol and its governance are relatively untested, making it challenging to predict the community’s response to unforeseen events or significant market shifts. The lack of a track record increases the overall risk and ability to extract quality signals on the overall robustness of the platform. According to the Aave V3 Risk documentation, this would yield a D- rating on the time axis and low grades across other metrics.

---

Parameter recommendations

Given the above, if the community decides to onboard fUSDC as a collateral asset, we recommend initially listing the asset in isolation mode to bound protocol risk exposure.

As fUSDC can be redeemed for USDC, we support the recommendations to match the LTV, LT, LB, RF, LPF, and IR curve parameters to those of USDC.

We do note that there is a pending Chaos proposal to increase the LT and LTV of USDC on V3 Ethereum that, if passed, would require updating this proposal.

fUSDC volatility vs. USDC volatility showing strong correlation between the asset prices

Supply Cap
For the supply cap, as fUSDC is redeemable only through the Flux Finance USDC market, meaning that to support liquidations on Aave, there needs to be enough liquid USDC in Flux Finance to redeem the fUSDC that is being used as collateral on Aave. As the current USDC liquidity on Flux Finance is ~$2.22M, we suggest setting the initial supply cap at 2X the amount of liquidity available - ~$4.5M = ~220M fUSDC. This is a prudent approach that allows for half of the total supply to be liquidated at once. Chaos Labs will continue to monitor the available liquidity on Flux Finance and the positions on AAVE V3 on Ethereum to update these recommendations when needed.

Parameter Recommendations:

Parameter	Value
Isolation Mode	Yes
Borrowable	No
Collateral Enabled	Yes
Supply Cap	220M (~$4.45M)
Borrow Cap	N/A
Debt Ceiling	$1M
LTV	74% (77%*)
LT	76% (79%*)
Liquidation Bonus	4.5%
Liquidation Protocol Fee	20%
Variable Base	0.00%
Variable Slope1	4.00%
Variable Slope2	60.00%
Uoptimal	90.00%
Reserve Factor	10.00%
Stable Borrowing	Disabled

*LT and LTV will be updated according to the outcome of the following proposal

[Gauntlet]

Hey @MathisGD - thanks for the followup and observations regarding the concentration of fUSDC. Here are some of the tradeoffs regarding putting fUSDC in isolation mode vs not.

Pros

• Reduces risk of unhealthy liquidations by only allowing stablecoin borrowing.

Cons

• Prevents a potentially major use case of fUSDC collateral, which is to borrow WETH.
• WETH borrowing is the primary use case of USDC collateral on Ethereum v3. The chart shows the summed borrow positions of the top 100 suppliers of USDC (supplying in total of $150mm), excluding those that are in ETH-correlated Emode.

As discussed previously, fUSDC effectively represents USDC deposit on Flux Finance. Unless there is severe smart contract failure or other types of unquantifiable underlying devaluation risk, fUSDC can always be available to be redeemed to USDC, pending liquidity. If available liquidity is depleted, then borrow rates for fUSDC on Flux should increase quickly via Flux Slope 2 parameters to prompt USDC borrowers on Flux to repay, thus bringing back liquidity. fUSDC thus should be very similar to USDC in collateral quality.

If the community has a more conservative view towards these types of nascent assets, it can vote for listing in isolation mode and/or reducing the supply cap to the available liquidity ($2.5m).

[MathisGD]

@Gauntlet

Thanks for the detailed answer.

@Gauntlet

If available liquidity is depleted, then borrow rates for fUSDC on Flux should increase quickly via Flux Slope 2 parameters to prompt USDC borrowers on Flux to repay, thus bringing back liquidity.

The absolute amount can still decrease if the total supply decreases. This is the main reason why this market would require a careful monitoring.

---

@Gauntlet

• Prevents a potentially major use case of fUSDC collateral, which is to borrow WETH.
• WETH borrowing is the primary use case of USDC collateral on Ethereum v3. The chart shows the summed borrow positions of the top 100 suppliers of USDC (supplying in total of $150mm), excluding those that are in ETH-correlated Emode.

This is interesting data ! Although adding fUSDC as collateral for borrowing stablecoins is useful for people who want to take leverage on fUSDC’s returns (the same way it is done with LSDs and ETH), as mentioned by @MarcZeller. This is something that can’t be done today (and why it can’t show up on the data).

---

In the end I think that both options (isolation mode or not) are possible, but listing it as normal collateral requires a closer monitoring, and if the main use-case turns out to be borrowing stablecoins, it is not useful to take this “risk”.

[MarcZeller]

Hello and thanks for everyone involved in this discussion for their feedback.

following governance usage, this proposal is almost ready to be escalated to Snapshot stage, following the most conservative recommendation. Risk parameters will have the opportunity to evolve after onboarding as the asset gain maturity.

The use-case of fUSDC in Aave is mainly to borrow USDC and potentially mid-term in a specific emode designed for this. We do not believe in a significant potential use-case to use fUSDC to borrow wETH, Therefore, there’s no issue with onboarding it in isolation mode.

However, we think the debt ceiling of 1M$ is way too conservative and hurts the use-case of this asset onboarding without noticeable safety gains. We would like @ChaosLabs to reconsider if they would be comfortable with a 4M$ ceiling before escalating to a snapshot vote.

The risk parameters candidate to be escalated to the snapshot stage are:

Parameter	Value
Isolation Mode	Yes
Borrowable	No
Collateral Enabled	Yes
Supply Cap	220M (~$4.45M)
Borrow Cap	N/A
Debt Ceiling	$4M
LTV	77%
LT	79%
Liquidation Bonus	4.5%
Liquidation Protocol Fee	20%
Variable Base	0.00%
Variable Slope1	4.00%
Variable Slope2	60.00%
Uoptimal	90.00%
Reserve Factor	10.00%
Stable Borrowing	Disabled

[ChaosLabs]

We recommended the initial conservative $1M debt ceiling based on the available USDC liquidity on Flux and the limited historical data on protocol usage. Our general approach in these cases is setting a lower debt ceiling and raising it later after analyzing the usage of fUSDC on Aave and the availability of liquidity on Flux. The current liquidity snapshot for fUSDC is even lower than when the original recommendation was made, which demonstrates the dynamic relationship between usage and set parameters.

Ultimately, the debt ceiling is a community-controlled parameter to cap the risk for a specific asset.
We think adding an “aggressive” option on the coming Snapshot with a higher debt ceiling is appropriate in this case, to gauge the community preference for fUSDC.

[MarcZeller]

understood,

proposal escalated to Snapshot with a conservative Ceiling, governance will have opportunity later if cap is reached to raise ceiling via a later governance process.