RFP Idea: Adding Collateral Support for MAI

Feedback
1

Abstract

MAI is the first and largest stablecoin protocol on Polygon, now available on 12 chains. It allows users to lock their tokens and use them as collateral to mint MAI at 0% interest.

MAI is one of the most crosschain assets in DeFi, currently on 22 chains through Multichain, Celer, and Stargate. Accepted collateral includes interest-bearing assets like Beefy strategies. Users are not charged interest for borrowing MAI - instead, they are charged a 0.5% fee upon repayment.

All accepted collaterals are priced using Chainlink price feeds. MAI has stayed within its peg through several market downturns.

MAI currently has a TVL of around $56M, and has around 42M MAI outstanding.

Key Terms (N/A)

Motivation

MAI is a decentralized stablecoin, not backed by centralized assets like USDC. This reduces risks associated with the most recent stablecoin depegging event in our industry.

MAI is also extremely ubiquitous, being used by most major applications and on all major chains. As such, adding MAI is a great way to onboard new users into Radiant.

MAI’s risk profile is low. It is 4 times more stable than its average peer (see below).

Volatility analysis (normalized average daily volatility):

  • MAI: 0.17%
  • MIM: 0.28%
  • sUSD: 0.73%
  • alUSD: 0.89%
  • LUSD: 1.34%

MAI is also much more used than its peers, seeing a very high volume to market cap ratio. Healthy usage means MAI will see more activity than its peers on Radiant.

Usage (daily volume / market cap)

  • MAI: 12.40%
  • sUSD: 7.75%
  • LUSD: 4.78%
  • MIM: 4.37%
  • alUSD: 1.43%

Market data

  • Market Cap: $42M
  • 24h Volume: $8M
  • Volatility: Very low
  • Exchanges: Velodrome, Curve, Balancer, QuickSwap, KyberSwap, Equilibre, Ramses, Solidlizzard, Spookyswap, and others
  • Maturity: Launched 5/4/21
  • Social channels data (Size of communities, activity on Github) - posted separately due to restrictions on amount of links in a post.

Further information on token

  • Date of deployment: May 4, 2021
  • Number of transactions: 9M+
  • Number of crosschain token holders: 45k+

Security Considerations

MAI has undergone 2 thorough audits (see here). Constant internal reviews are performed to ensure MAI’s economic and technical safety.

The protocol currently has around $56M TVL.

Specifications

Only specification is that a Chainlink oracle should be used to price MAI. MAI is an ERC20 token.

Steps to Implement

Adding a new asset should be fairly simple, as the code for the existing markets should be reusable.

Timeline

Within two weeks of the vote concluding.

Overall Cost

Cost of adding a new market should be negligible as it requires no new operating costs.

2

Looks like a good proposal subject to risk analysis.

I’d like to highlight/ask on a couple of points which I hope you can clarify for me.

This market.xyz pool has circa $400k of MAI-related ‘bad debt’. If the MAI was minted into the pool via an AMO, then I believe this would become protocol bad debt and reduce MAI’s backing to somewhere around $0.99 (not a dealbreaker by any means, but important to consider). If, however, this was supplied to the pool by third party users, then MAI’s backing is retained fully.

Would you be able to clarify which of these scenarios is the case, and any other disclosures of similar situations with undercollateralized lending markets involving MAI?

Also, what parameters are you proposing for Max LTV, Liquidation threshold, and Liquidation penalty for MAI? Somewhere around 70, 75, 15 seems sensible?

Finally, would you consider incentivizing MAI usage on Radiant using some QI tokens? As with all collateral additions, this would have at least some associated risk for both liquidity providers and dLP lockers (with total loss a possibility - again, as with all collateral additions), so some tangible and measurable benefit to taking this risk would likely help dLP lockers in making their decision.

Thanks!

3

Adding any form of stablecoin as collateral that does not have multi billions in marketcap is nothing but unnecessary risk.

4

Totally agreed. Less liquidity, higher chance of market manipulation, we need bluechips tokens and they are probably enough for a cross-chain lending and borrowing protocol like Rdnt to grow. We don’t need to assume the risk.

5

MAI would be a great addition. This would offer a lot of composability options for RDNT users as well as raising our visibility with the MAI userbase. Also QIDAO are always busy promoting their product in novel ways that would benefit the Radiant protocol via partnership. I think concerns regarding quality of collateral could be addressed via a slightly lower LTV ratio like 70-75%.

1 Like
6

Hey, thanks for your feedback.

None of the MAI bad debt on Market.xyz is related to our AMOs. Those are third party user funds that were lost by users of Market.xyz.

Currently, there are no AMOs run by QiDao.

For risk parameters, we had the following in mind. It’s what we’ve maintained on Aave.

  • Max LTV: 75%
  • Liquidation threshold: 80%
  • Liquidation bonus: 5%

We should set a borrow cap of 600k MAI and a supply cap of 1M MAI. As we continue to scale liquidity on Arbitrum, we can increases these figures.

As for incentivization, we should look at how MAI markets have developed on other protocol. For our Aave V3 deployment on Polygon, we’ve seen a great natural market for MAI. The deposit APR has remained between 2% and 3% with no intervention.

1 Like
7

Personally, I dont think that market cap size is a good predictor of risk. UST had a very large market cap, yet its protocol design was the main component of its risk profile.

When it comes to adding a new asset to Radiant, it’s important to look at what affects liquidations such a volatility. MAI is incredible stable, even by stablecoin standards, so it’s generally considered to be a good collateral asset. It is listed on Aave and 0vix, with no issues occurring on liquidations.

8

Every asset can be manipulated. We all saw what happened with AVAX on GMX’s Avalanche deployment a few months back. It’s important to set the right risk parameters to ensure liquidations run smoothly and to prevent manipulations from being profitable.

So why would Radiant add MAI and not just focus on blue chips? Well the answer lies in growth. MAI and its protocol QiDao have a lot of users. According to DeBank, QiDao has 16,897 users. This is a lot more than other bluechip stablecoin protocols. Following DeBank’s numbers, MakerDAO has 6,494 users and Liquity has 3,753 users.

So you have to ask yourself, are we building for the few whales or are we building for the bulk of users. I believe that Radiant can have a greater impact and long term growth if it onboards a large amount of users. That comes from adding assets that people actually use.

1 Like
9

Unless we can isolate the collateral like what Aave does, I suggest only to list blue chips and do more deployment into different chains with the tokens we had listed. It should be sufficient for the growth of the protocol and also mitigate the risk in this lending and borrowing protocol in crypto.

1 Like