What happens if Angle goes bankrupt?

Bankruptcy is a term that I've very rarely seen in the context of decentralized stablecoin protocols or of DeFi protocols in general. Yet, stablecoins remain financial systems and it makes sense to also apply TradFi terms even if they just exist purely onchain.

In DeFi spheres, people prefer to use the term under-collateralized to describe insolvent stablecoin systems. Whatever the term, the situation remains the same, a bankrupt protocol is a protocol for which the value of the liabilities (e.g stablecoins in circulation) is greater than that of the assets controlled by the protocol (if you want a better understanding of stablecoin balance sheets I invite you to read this article I wrote).

The point here is not to analyze the causes that can lead a stablecoin to go bankrupt, we've also written a detailed article on the major risks at stake in DeFi protocols, but more to understand what to do when this happens and why it's important to have measures already in place before an insolvency situation strikes in a protocol.

First things first, Angle has never been this far from being bankrupt. Right now Angle is well-over collateralized and sits on an equity of 7.1m€: this is what would be left in the protocol if all protocol stakeholders removed their liquidity, burnt their agEUR or repaid their debt. Put otherwise, this means that the protocol can “afford” a loss of 7.1m€ and still ensure that there is 1 Euro worth of reserves for every agEUR in circulation.

Let's however imagine that someday because a collateral lost in value due to one of its underlying banks going bankrupt, like what happened during the Silicon Vally Bank failure, Angle ends up with less reserves than agEUR in circulation.

The issue within all bankrupt systems is letting the door open to first arrived first served situations.

Imagine you're in a setting where there are 90€ of assets to back 100€ worth of claims.

Then, in the absence of proper safeguards, the first claimants to arrive could redeem the 90€ of assets eventually leaving 0€ of asset for the last 10€ of claims. This is the thing you notice in any bank run scenario.

To prevent bank runs and pave the way to a fair outcome for all stakeholders, what matters is breaking this sequentiality mechanism: making sure that the first arrived are not the first served, but rather that everyone gets an amount of the reserves proportional to their stake.

When there are 0.9€ for every € worth of claim, the fairest outcome possible is to guarantee to every stakeholder 0.9€ for every € of claim.

The issue here again is that you don't want to have to activate your fair redemption feature. It needs to be here at all times otherwise you can easily be exploited during the time before which governance decides to pull the trigger. Any moment of incertitude can lead to unnecessary speculation, arbitrage and insider trading from players who have more information about what the expected outcome is going to be.

In March 2023, during the depeg of the USDC stablecoin which was making a significant chunk of the DAI reserves, the fair value of Maker's DAI was uncertain: people had to wait for a governance intervention to raise some limits and halt trading before knowing what would be the right price for Maker's DAI.

It is with the objective of solving all these issues that Angle Transmuter system was designed. Angle Transmuter comes with a redemption feature that is live at all times (it does not require any activation by a governance) and that ensures a fair distribution of the protocol's assets to its participants.

If the protocol is under-collateralized, everyone can redeem with its agEUR a proportion of what the protocol controls in the reserves. But also when the protocol is also well collateralized you can always get 1€ worth of collateral for your agEUR directly with the protocol.

The advantage here is that no one has to tell the protocol to activate a specific mode: the protocol just has to read into its balances and oracle values to automatically understand that some functionalities must be blocked.

If you see an asset in Angle basket of reserves that is depegging, it's also impossible for you to deplete the healthy reserves of the protocol leaving the other stakeholders with a backing only made up of a weak asset.

There is no case where you can exit with more than what other stakeholders would get. Even better, in case the protocol is slightly under-collateralized, a small penalty applies to the first ones to exit which means that they're leaving better off the stakeholders which decided to remain within the protocol.

Angle Transmuter is one of the first iterations of a stablecoin protocol implementing bankruptcy protection mechanisms in its core (Gyroscope is another example).

What worries me is that there aren't more protocols doing so. It seems that many people have forgotten about the USDC depeg and few protocols are adequately prepared for when a similar situation happens again.

There are a few other areas in DeFi where I believe that the concept of bankruptcy protection could be pushed further. For instance in borrowing/collateralized-debt-position systems, there have been very few attempts to enforce robust bad debt socialization mechanisms and fair allocation systems of potential losses.

Angle Borrowing module infrastructure comes with its own settlement mechanism, yet this one first needs to be activated by a governance.

In addition, Transmuter and all bankruptcy protection systems are well suited for the situations where the economic value of an asset in the backing is decreasing, but they don't work as well if a protocol becomes insolvent due to a hack.

In such a situation, regardless of the setup in place, the best option may be to keep an emergency multisig able to freeze the system before letting the protocol open to redeem at the face value of what's left.

One challenge I see with the greater adoption of such features in protocol is the advent of real-world assets.

Imagine a protocol owning in its reserves tokenized permissioned security tokens which cannot be distributed to every address. This in turn means that only the addresses with a proper KYC level enabling them to own these security tokens should be able to fairly redeem their assets with the protocol. It naturally reduces the scope of people who have access to these functionalities which should in the end be unbiased.

Down the line this shouldn't be an issue though because assuming an efficient market, this means that KYC-ed arbitrageurs could arb the secondary market to the fair value of the reserves.

In any case, providing the highest level of guarantees and protection for every agEUR holder is the #1 priority of the protocol. Even if our mission as a community is for this never to happen, in case of a black-swan event, Angle stakeholders would be far better off than they would be in any other decentralized system.

There would be no speculation possible on the fair value of the stablecoin and what to expect from the protocol would be highly and transparently predictable by every DeFi participant.

It's still the early days of bankruptcy protection in DeFi systems, Angle is at the edge today, but now that the foundations have been laid down, there is still some research to be done for making this even more holistic and comprehensive to everything that could lead protocols to become insolvent.

Regardless of this, having such features already today is a 0→1 for DeFi users, and we saw in 2023 how not having any measure in place could be dangerous and lead to millions of users value stupidly lost to predatory money.