CIP: Add USDT Omnipool

mudshrk · March 11, 2023, 4:53pm

Summary:

Conic currently supports Omnipools for three underlying assets: DAI, FRAX, and USDC. This proposal recommends adding an Omnipool for USDT.

If this proposal is successful, a separate proposal would follow for selecting whitelisted Curve pools for the USDT Omnipool.

Background:

Conic Omnipools went live on March 1, 2023, and users are currently able to deposit into three genesis Omnipools: DAI, FRAX, and USDC. The choice of these underlying assets was determined by a governance vote that concluded on December 11, 2022: Snapshot

The underlying assets for the genesis Omnipools were required to be USD-pegged and have at least two high TVL Curve pools. DAI, FRAX, USDC, and USDT were presented as four possible choices, where an asset needed at least 50% voting support to receive an Omnipool. USDT received only 13.6% support and so was not included. This proposal seeks to revisit the question of adding a USDT Omnipool.

Motivation:

Omnipool TVL has grown quickly since launch, and the Conic community acknowledges the value of Omnipool TVL for the success of the protocol and the utility of vlCNC. A USDT Omnipool has the potential to significantly grow Omnipool TVL and thereby benefit the Conic protocol.

USDT is issued by Tether Limited and has the largest supply of any dollar-pegged stablecoin at approximately 70billion in circulation—nearly twice the size of USDC. However, USDT-holders are currently disincentivized from depositing into Conic since there is no USDT Omnipool. A USDT-holder who wishes to deposit into an Omnipool must first swap into DAI, FRAX, or USDC, thereby incurring gas fees and slippage. The need to swap into another stablecoin also degrades the user experience, as the user cannot swap directly on the Conic website. Finally, the need to deposit into a DAI, FRAX, or USDC Omnipool forecloses the possibility of allowing USDT-holders to choose their own distinct Curve pool allocation via vlCNC governance. A USDT Omnipool resolves these problems and has the potential to grow TVL by reducing friction on USDT deposits.

Risks:

Community members have raised several objections to a USDT Omnipool. We review these objections and present replies in the hope of stimulating further discussion in the comment section below.

Objection 1: A USDT Omnipool would dilute CNC yields for current Omnipool depositors.

Reply: CNC rewards are divided among the total TVL of all Omnipools combined, so deposits to a new Omnipool would not dilute rewards any more than deposits to an existing Omnipool.

Objection 2: USDT is a risky asset since Tether is less than perfectly transparent about the assets backing USDT.

Reply: Assessing risk is complex, as the recent stablecoin depegs illustrate. But depositors into a USDT Omnipool are users who already hold USDT, and so have already tacitly accepted the risk of holding USDT. Note also that deposits into a USDT Omnipool are denominated in USDT: the user receives cncUSDT with the expectation of being able to redeem cncUSDT for USDT upon withdrawal. In the event of a USDT depeg, it is even easier to acquire USDT to satisfy cncUSDT liabilities. So it is the depositor, not the Conic protocol, who bears the risk of a USDT depeg.

Objection 3: A USDT Omnipool is either redundant or exposes depositors to impermanent loss. That is, a USDT Omnipool either (i) deposits into the same Curve pools as the DAI, FRAX, and USDC Omnipools and so is redundant, or (ii) a USDT Omnipool deposits into the Tricrypto pool, which contains nonUSD-pegged assets and thus exposes the depositor to impermanent loss.

Reply: It is up to governance to decide which Curve pools to whitelist for a USDT Omnipool. But there are options that are neither redundant nor expose depositors to nonUSD-pegged assets. Notice that there are several high-TVL stable pools that have not yet been whitelisted for any of the existing Omnipools. A non-redundant set of stable pools might thus look as follows:

MIM-3crv
TUSD-3crv
LUSD-3crv

A USDT Omnipool could also include a mix of the above and the following Curve pools, which have already been whitelisted for the existing Omnipools:

3crv
sUSD-3crv
FRAX-3crv

We welcome additional suggestions in the comments below.

daesu · March 14, 2023, 3:13am

I support this proposal. USDC has lost some user confidence due to recent events, this should lead to more and more users choosing USDT as their go-to stablecoin. The need to hold USDC or DAI to use Omnipools can now represent a higher risk and an extra step that can discourage future users in their first approach to Conic. So we need to offer another option to existing and new USDT holders, making a USDT Omnipool now obvious.

bb8 · March 14, 2023, 4:18pm

Thanks for providing an overview on this topic. There’s been quite a bit of discussion regarding supporting USDT on Discord over the past few weeks and this summarized it nicely. Beyond the fact that USDT is the largest stablecoin by market cap, there’s several reasons why a USDT Omnipool would be beneficial to LPs (as outlined in the proposal). Furthermore, the recent USDC depeg highlights the value in supporting a variety of different stablecoins.

I’ll push this proposal to Snapshot tomorrow. If this proposal is to pass the next steps would be:

Deciding which Curve pools to whitelist (CAP)
Deciding the initial liquidity allocation weights (LAV)

ConicIntern · March 14, 2023, 4:51pm

Agree with @bb8 was a great overview.

I think including USDT is a sensible step forward. Looking forward to the snapshot, and the upcoming discussions on which pools to whitelist / their respective weights.