V2 Terra Mechanisms for a Collateralized Stable

Summary
This proposal is for a better stable model on chain. Moves away from algo because we don’t need the scalability. The true value prop. is an L1 built around a decentralized stable.

Motivation
Printing a new token and collateralizing it for a stable leaves 2 tokens without value, this is why an L1 asset built to collateralize a stable is needed. Otherwise there is no intrinsic value for the collateral, in this case that value would be the block space.

The stable won’t be created initially, as the new L1 token (referred to as Phoenix for now on) will be very volatile to start as many people will just be looking for exit liquidity, leading to excess liquidations early on. This would also allow the community to build out the necessary mechanisms but not have to wait to distribute Phoenix.

Proposal
Create a collateralized stable, mUSD, following Mirror naming conventions. It would be backed by Phoenix at a 110% - 200% collateral ratio, or 50% to 90% LTV. The borrow limit is 50% but the liquidation doesn’t hit until 90%, allowing for about a ~45% drop before liquidation, assuming no repayments.

Side Note: This opens up a huge market for Nexus and anyone redirecting yields

This is conservative for a reason. No need to have the stable out scale the ecosystem nor have a system with high chance of liquidation, as that would discourage usage.

The CDP would take a 0.5-2% borrow fee, with no interest. The fee and the low LTV disincentivize looping by lowering the amount of loops (LTV) and lowering the profitability of them (fee). Borrow fee goes stakers as a way to benefit from the growth of the stable as well as get a proportion of non-inflationary fees.

Liquidations
As the main mechanism for Phoenix, we would NOT want to rely on 3rd party bots for incentives. This is easier to accomplish when building things on chain as the procedures can be built into the block creation. I’m not well versed in chain tech but the idea here is that every block there would be a call for liquidations. If there are underwater positions, then they are liquidated, if not then it fails and nothing happens.

Instead of a bid mechanism like Anchor, I propose using only a Stability Pool (SP), similar to the one used by Liquity (Stability Pool and Liquidations - Liquity Docs). No premiums since a liquidation at 110% would provide a profit to the SP depositors. This would be incentivized by the chain initially but marginally less than what would go to stakers. The reason here is as to not accrue mercenary capital overtime since the SP has no locks. The true value prop. of the SP is discounted assets in bearish conditions. It would also be low in the risk category if built into the chain.

Worse case for liquidations where the SP is empty, the chain would distribute liquidated collateral to the SP and accrue bad debt. This alone should incentivize the SP to never be empty, as the users there would benefit from all liquidations no matter its size.

The idea here is that its better for the chain to accrue bad debt and have to pay it off than to sell Phoenix and cause more liquidations itself. The chain would repay bad debts 1st from undistributed mUSD staking rewards and 2nd from the borrow fee of new borrows until repaid. The thought process here is that bad debt lowers the valuation of mUSD, so in order to save that value and the value of future rewards to stakers, the mUSD is used as insurance. This provides a growing insurance fund in prolific markets for chain solvency protection in sharp downturns.

A possible attack vector for this would be someone buying up mUSD on the 2ndary market, loading up the SP, and trying to force liquidations. It would take a long time to accumulate mUSD without opening a CDP themselves which would get them liquidated as well. This would also assume no one is paying back debt which should admittedly be easier in a crash as people sell mUSD for Phoenix, mUSD depegs so people buy it to repay debts. The attacker could also benefit from this depeg. Another factor is the proportional amount of assets in the SP, if people start adding to it in anticipation for liquidations, the attackers profit metrics decrease.

Growth Thesis
Stay in IBC. All chains connected to IBC can use the liquidity found on the native chain, or anywhere in IBC for that matter fairly seamlessly. This allows users to feel like they aren’t trapped in the system which was a flaw for UST. The only escapes were bridges which adds attack vectors (low cross chain liquidity), or CEXs. I think we should take a page out of Osmosis and not worry about CEX listings, which guarantees users of our platform can navigate DeFi enough to calmly liquidate if they so choose.

On the same note about system exits, the main DEX should have robust liquidity with a bridged stable that allows users to get out. I advocate for an accessible stable that isn’t totally freezable, so FRAX MIM and BUSD come to mind. Long term once we regain some trust, we can get Phoenix added to Thorchain and users would be able to easily exit that way.

The community is still generally rallied around a stable and the growth of said stable doesn’t become artificial. If above peg, users can mint more to arb it, if below peg users would buy back to repay debts. Allows the chain to safely grow within itself and the broader IBC ecosystem.

The issues with the previous model for Terra were lack of non-mercenary cross chain liquidity (ie unreliable), artificial demand making the peg weak during large redemptions, and no easy DeFi exits out of the system which pushed people to CEXs which are frequently used for oracle prices.

Would love to hear what you guys think.

6 Likes

good idea

1 Like

if do kwon will be out of new project i am ok with everything
there is no future and no trust with do kwon aboard
his place is in jail

4 Likes

I do not think we can rebuild with TFL at the helm, that would be irresponsible on our part.

2 Likes

Somebody named @kteranisi in Twitter introduced the Terra V2 Upgrade proposal to @costapanayiotl2 (cp61) and he started asking questions on how to introduce a multicolateral stablecoin. Teranisi made two good points:

  1. There is an obvious demand for a fixed stablecoin and a bonding multicollateral system is already implemented and still working on Anchor.

  2. Apart from these assets who were and are bonded (bETH, bATOM, wAVAX, bBNB…) Binance may be very interested on BUSD being used to collateralize a decentralized stablecoin the same way that USDC and USDT are being used to collateralize DAI.

When you peg to one asset using a different asset as collateral, there will always be a chance for under collateralization or depegging. Even if it is over collateralized by 10x, the collateral asset can crash more than 10x. Nothing is 100% stable (relative to something else) in this world.

Some may last longer than other eventually they all will fail. The only way any Stablecoins will function in the long run is if you hold the underlying asset (USD) as reserve.