New DEI Peg mechanism
About DEI peg:
First of all, we have been aware of DEI being below 1$ for several days now.
That is because, between 0.997 — and 1.003, we let natural arbitrage happen from market participants. Below that range, the AMO is constantly buying back.
A new incentive model between 997–003 so the AMO doesn’t have to be active to bring the price back to 1.
We want this to happen naturally, and the AMO is just stepping in at big sales that bring prices below the 997 range.
So we will be activating multiple measures that are currently being used in other protocols to keep the peg at 1$
Not only protocol controlled / arbitrage bots but also natural participants.
1: Stability Fee (inspired by MakerDAO)
The first addition to the DEI system is the model DAI uses.
If DEI is below 1$ for a longer time, the interest rates of overcollateralized Loans increase every second.
That mechanism pushes borrowers to buy back DEI below 1$ to pay back their loans more often.
2: AMO v2
will imitate natural user behavior, using lending, borrowing, repaying buyback, and collateralizing
3: AMOv3 the peg-savior
Additionally, we will implement that at a random time. Without any notice, we will use the AMO to bring the price back to 1$, but this will “punish” all borrowers by increasing interest rate by more significant percentages, so they will need to buy DEI above peg.
The new AMO is built directly into the protocol, and his task is to:
1- restore peg to 1$ at random times
2- help keep the peg in the desired range
3- increase interest rates
4- write debt from fractional system to over collateralized and vice versa creating a dynamic
5- burn outstanding DEI
Fractional to Overcollateralized ratio f/o
For that to work, we need to understand the current system:
DEI can be minted overcollateralized (isolated risk)
DEI can be minted fractional (global risk)
In our current implementation, both systems are isolated from each other:
The only connection is a variable we called “virtualReserve” this tells the PoolUSDC contracts that they should ignore that amount from their balance.
So the current F/O ratio is at
This means 2.16 times more fractional minted DEI than overcollateralized minted DEI.
The current PoolUSDC reserves are at $51,486,376.32
This means ~7M USDC was withdrawn for the AMO to keep its peg at 0.997
7m / 57m = 1.13 multiplier factor on virtualReserve but also on interestRates.
Those DEI will be added to the GeneralLender, increasing the O ratio in F/O.
With this dynamic, we will decrease the global risk by writing the balance to the isolated risk side when DEI trades below 1$; additionally, the increasing interest rate will naturally lead to increased demand for DEI below 1$.
With AMOv3 being launched, this will dynamically occur all the time.
It will write fractional to overcollateralized balances,
changing the F/O ratio constantly.
We will also activate the buyback / burn and collateralize that frax implemented into their stable system.
this new monetary policy atleast the part of the AMOv3 will be enacted right now
We deactivated the AMO today in anticipation of what happened at UST, to cooldown the market and not deploy further treasury assets.
AMO will get activated again over the night, with a new monetary policy model that will be rolled out in parts today and in the future.
With this upgrade we are implementing another feature on our mission to innovate in the decentralized stablecoin space.
An AMO that interacts with money markets and issues treasury bonds to ensure 1:1 backing of all DEI in circulation.
With the last events happening at Luna and UST, its important to ensure DEI will be able to maintain its peg further.
The main pillars of DEI
- collateral backed fractional minting similar to frax
- overcollateralized minting via non-correlated assets and P2P money markets with dynamic interest rates similar to DAI
- Modified FRAX re-collateralize TWAP-based called DEUS Treasury bonds with fixed yield.
- AMO technology is based on FRAX to enforce monetary policy combined with secondary money markets to supply or borrow.
The critical difference to UST is that there needs to be collateral injected into the system for every DEI going into circulation, collateral that can be used.
(this is very important going forward)
Hard collateral means either stablecoins like USDC fractional or volatile assets over collateralized.
Luna had one issue, it used volatile assets but not overcollateralized, meaning in times of high volatility, the peg becomes unreachable because price declines faster as liquidation can be achieved.
Let me explain that. Imagine you need 10 seconds to execute the liquidation, the oracle lags 1 min behind, and the price decreases by 10% per minute. It could end up where your collateral has less value than the debt because its price decreased too fast.
This leads to a heavy bank run that can’t be stopped.
How does DAI mitigate this?
Having overcollateralized non-project correlated collateral,
Implementing AMO that lends money from money markets or deploys funds at money markets based on the market situation to restore peg in volatile market situations.
A Bonding mechanism lets everyone buy DEUS treasury bonds, with the desired expiration day and fixed interest on top of them.
The fixed yield multiplier is 1.2; the multiplier rises with time if the demand to buy bonds is insufficient. It gets more attractive over time to buy bonds.
Bonds have an expiration day chosen by the buyer.
Ranging from 14 days to 365 days.
Buying one bond at 1 USDC price gives you 1 DEI at expiration day + fixed interest over time.
(that was repurchased before from the protocol via AMO or deposited into a money market)
Let’s say the current Interest is 1.2. you choose 365 days expiration.
You get 1 DEI + 0.2$ worth of DEUS over 365 days.
Choosing 14 days, you get 1 DEI after 14 days + 0.2 * 14/365 DEUS.
The DEUS is calculated based on an average 24h price and can be claimed daily.
We will be selling Bonds until there is a 1:1 USDC treasury for every DEI in circulation, whether minted fractional or over collateralized.