Failsafe Reserve UNIT Buyback System
By our calculations, on a pessimistic theoretical basis -- assuming all Borrowers borrow as aggressively as possible, no Borrowers repay or recapitalise their Vaults to forestall the 16% tax of a Liquidation Event, and every other factor within the Protocol's control is as sub-optimised as possible going into a black swan event -- our protocol cannot incur any bad debt on an individual-Vault basis until BTC drops ~35% within an extremely short time. On a practical basis, the buffer will be significantly wider, as many users will self-liquidate or recapitalise to avoid a liquidation tax.
However, DeFi is famously unpredictable, and 'BTC-Fi' is nascent. There's a very remote yet real possibility that the Protocol fails to clear its unsterilizedDebt balance during a Black Thursday-type event, when Liquidators' available capital is swamped and BTC faces an accelerating plunge in response to some unforeseeable shock.
Programmatic Reserve Design
Depeg events are guaranteed in any stablecoin design, even fiat-overcollateralised ones. These can occur due to regulatory events compromising collateral liquidity (USDC in March 2023), exogenous under-collateralisation (UST), an unexpected dependency on now-compromised crypto collateral (DAI's USDC exposure in March 2023), network congestion and/or oracle issues causing the Protocol to deleverage on a flawed pricing basis (which happened to every stablecoin at some point in time), or other problems.
So regardless of any purported ingenuity of financial engineering, an extremely sharp deleveraging event could cause unsterilisedDebt balances to fail to clear from block to block as Protocol users' capital, in addition to their off-protocol "dry powder", is destroyed faster and faster, the unsterilisedDebt balance grows, and the Protocol's BTC reserves drop in value simultaneously, such that the USD value of auctionReserve falls below the total UNIT in circulation.
If this happens, the Protocol will need to buy time proportional to its availability of BTC reserves as it triages its design flaw(s), defends its UNIT peg, and provides some measure of redemption liquidity to those who want to withdraw immediately--all at the same time. That's what the Programmatic Reserve is supposed to do.
The Protocol will continue bidding for UNIT at a price of
auctionReserve/unsterilisedDebt, albeit at a peg of less than 1:1.The Vault's
unsterilisedDebtRolling Debt Auction will lower its bid price to repurchaseunsterilisedDebt, such that tranches:20% of
auctionReservewould bid at $.90, 20% at $.80, 20% at $.70, 20% at $.60, and 20% at $.50.The Protocol would assume that an unforeseen inefficiency caused a buildup of unhealthy or parasitic debt (debt that appears
overCollateralisedbut somehow isn't in practice) and would offer users exit liquidity at a deepening discount. At the same time, we solved whatever design flaw appeared.
Vaults' debts and Liquidation Thresholds will be valued based on UNIT's USD oracle price whenever the UNIT Rolling Debt Auction conversion rate falls below $.99, not simply the BTC:UNIT ratio.
The Protocol's interest in a de-pegging event is to liquidate UNIT borrowers when based on their BTC assets to USD liabilities, which would be declining during a UNIT de-pegging event.
If UNIT traded below one USD, we would apply the following formulas to calculate the true collateralisation rate.
Additionally, from a fairness standpoint, liquidating UNIT borrowers wouldn't be fair because our protocol failed to manage its BTC collateral efficiently enough to maintain a UNIT:USD peg.
For example, If UNIT were trading at $.90 to 1 USD, the value of a
U100: $130 BTC vault would be $90: $130. The relevant collateralisation ratio would be 130/90 = 144%, not 130%.
Last updated
