Adding a new debt-based Minting Module

ETA: Around end of May

Overview

Angle’s current model to issue stablecoins is very robust when gathering all the requirements. However, doing so can be difficult which can make expanding agEUR in a sustainable way under the current model challenging.

To help boost this growth, we are currently thinking about deploying a new additional minting module for agEUR that would let people borrow the stablecoin as debt from a deposited collateral, similarly to the Maker - DAI model (Liquity-LUSD, Abracadabra-MIM, …). This wouldn’t require the need for hedging agents or SLPs, and would allow users to keep their collateral exposure.

The new module would work hand in hand with the current one, and would also have minting rights on the agEUR contract. One of the big advantage is that it could be deployed on multiple chains in a safe and modular way: this could be the first step in allowing native minting of agEUR on other chains.

On top of that, this module would be part of the larger Angle ecosystem, such that it would provide value to veANGLE holders through potential revenue redistribution, and that they would be the ones leading its governance and upgrades.

Debt-based stablecoins ELI5

This on-chain model to create decentralized stablecoins has been created by Maker (DAI), and the same general mechanism was applied by other USD-pegged stablecoins like Liquity (LUSD) or Abracadabra (MIM). Users deposit tokens such as ETH in a vault as collateral, and can then mint stablecoins as debt. The borrowed/minted amount has to be low enough to ensure that the $ value of the collateral backing those stablecoins remains higher (i.e. that the stablecoins are always fully backed). In case of exceptional market conditions (sudden price drops), agents called liquidators come with stablecoins bought on the open market to pay back users loans (the stablecoins originally minted/borrowed). In exchange, they get the collateral that was backing the loan at a discount for ensuring the stability of the system.

Potential Features

While Angle New Module has a similar logic, we are working on some additional features to improve the overall system and the synergies with the current one and its governance token. Let’s have a look at how this new module could work below :point_down:

Multi-collateral, Isolated Loans, and Fees

First, such model can support various tokens as collateral, with different parameters for each of them. This means that a more stable token could have a higher LTV (loan-to-value) than a riskier or more volatile one.

On top of this, loans from users would be independent from one another, meaning that they could borrow different amount of stablecoins from different collateral tokens with different parameters.

Liquidations

As explained above, this system relies on liquidations to ensure that the circulating stablecoins are always backed by collateral in the system. These liquidations can be triggered by liquidators when loans become “unhealthy”, meaning that the value of their collateral is below what is required by the value of their debt.

Once a position becomes “unhealthy”, liquidators could receive an increasing discount depending on how far from a healthy situation the position is. This would create some sort of dutch auction to increase the liquidation incentives while the health of the system degrades.

Liquidators holding veANGLE (or having tokens delegated) would be able to earn a bonus on these liquidation incentives. This would allow to give more utility to veANGLE in the system, and increase incentives for liquidators.

The amount of debt being liquidated could vary depending on a target health factor. As long as this target can be reached within the collateral parameter, the liquidator will liquidate just what is required for the position to reach that target.

Debt Transfer

One interesting feature of Angle’s implementation would be for users to have the ability to transfer debt between vaults.

Imagine that you deposited DAI and ETH into the protocol, and have borrowed agEUR out of the ETH vault. If ETH price decreases, you might be in risk of being liquidated. In this case, you would have the option to transfer some or all your debt from one vault to the other, reducing your risk of liquidation.

Whitelisting and Volatile Assets Vaults

In general, only pre-defined tokens will be accepted as collateral. However, in some cases, the protocol’s governance could decide to whitelist an address to open a vault with a specific collateral and parameters. This would allow DAOs for example to open loans with their tokens and borrow agEUR from it.

Reactors

Expanding on this whitelisting possibility opens another interesting use case inspire from Tribe’s Turbo. This new module could allow someone to deposit riskier collateral, but give only to the protocol or other whitelisted addresses the right to borrow agEUR on top of this third-party collateral. The protocol would then use the funds in pre-defined ways, subsequently sharing the yield with the depositor.


Discussion

This is still a work in progress, and we want to get the community’s opinion and thoughts about it. What do you think? Are there any features you would add or remove?

We also created a #new-module channel in Discord.

7 Likes

Could be the next big thing for agEUR.
Could also allow to safely deploy new stablecoins, and on different chains. Overall, I believe we should all push for this

1 Like

Love this proposal, esp. excited about ability to permit minting on other chains and simplifying the path towards issuing different stablecoins.

The collateral transfer feature’s a cool innovative idea as well.

1 Like

Indeed, cross/multi chain agTokens will be great for the protocol!

If/when the expansion of these modules on other chains comes, we’ll have to keep in mind that it requires efficient liquidators to maintain a tight peg on the stablecoin.

I think this is a great, ambitious roadmap. The only hang-up I have is the proposal to add flash loans. Flash loans, along with oracle manipulations have been the primary attack vectors for protocol hacks. In my opinion, the addition of flash loans needs A LOT of research to identify possible attack vectors, and even then, it needs to be evaluated if the added revenue/better UX of adding flash loans is worth the risk. I am more in favor of “almost” flash loans, where you require at least 1 block between txns. This simple element, while increasing gas costs (who cares), would have prevented hundreds of millions of dollars in exploits over the last few years.

Without a huge amount of testing, auditing, and financial incentive to add a flash loan module, I would vote against such an addition.

Other than that, I love the roadmap. Some details still need to be worked out wrt ETH being used to mint agEUR, since the protocol will be net long ETH after mints, and ETH is quite volatile. So protections will need to be put in place to keep protocol level ETH exposure within reasonable risk bands.

2 Likes

Agree on most of this!
I don’t think the flash loan feature adds some risk for the protocol, more to other protocols interacting with agEUR, which will need to be wary when using it.
Flash loans may induce some risk, but they force all protocols to be super wary of what they do and what they use, so overall they’re good for the ecosystem I’d say.

It is going to facilitate liquidations for many users as well as the efficiency of the market for agEUR.

Flashswaps on pairs, direct minting and redemption are already :star2: for liquidations.

2 Likes

We should be open about discussions with MakerDAO as they have some quite vocal members supporting a possible euro DAI - but with no success in the horizon:

Could be a great way to get some smart minds to join Angle’s community, or even reach friendly-fork levels of collaboration with MakerDAO.

It’s true you’re right! Spent quite some time looking into this discussion! I don’t think though that this discussion has moved forward by a lot in the past few months, and from my understanding, it was put out of the table.

Totally, which strengthens Angle’s possible position as THE euroDAI equivalent

Yes that’s it, and that’s how I think we should position ourselves