Users start by depositing one or more different proof of stake cryptos into their native chain’s validator consensus staking pool. They can then automatically use those consensus stakes in their QuiD portfolio to earn or borrow against, re-staking them in QuiD's "Live Pool", and walk away anytime without penalty.
For now, we’ll only discuss borrowing synthetic dollars, although synthetic dollars can also be pledged to borrow crypto against. Either way, the minimum over-collateralisation for borrowing is 10%.
As part of their service of backing the protocol, crypto deposits that aren't used for collateral to borrow against accrue debt (from those who borrowed, but got liquidated). These deposits will earn money until what they owe on accrued debt ends up costing more than the deposit earns.
The accrual happens in proportion to depositors' % contribution to solvency, thus those who are more able to accrue bad debt without becoming liquidatable, accrue more. Active borrowers also accrue some debt in this way, but it is much less as is their contribution to solvency relative to non-borrowing depositors.
As such, depositors make up an insurance pool that protects the protocol’s dollar peg against the risk of under-collateralised loans…in other words keeping the protocol solvent. As compensation, depositors earn the collateral of liquidated borrowers (on top of their APR paid) in exchange for absorbing their debt.
Pooling assets and liabilities in this way allows us to socialise losses from liquidations. Gains are also socialised in the same way. Unlike regular depositors who contribute more to solvency, active borrowers contribute more to potential insolvency.
To the extent that they contribute to potential insolvency, borrowers pay bespoke premiums at an annualised rate that's live-updated. Every depositor earns their share of these premiums paid.
Depositors vote on how much above 100% solvent the protocol should be, with the weight of their vote being the size and age of their deposit minus the time since their last vote.
When current solvency is off from the median of all these weighted votes, premiums for all borrowers are scaled up or down to induce reaching the target. Premiums also change over time as a function of stress testing at a global and individual level, and the relationship between the two.
Stress testing checks what all the protocol’s deposits would be worth considering the worst possible 10% of price drops that could occur over a given year (conditional value at risk) based on historical data.
Oracles constantly check every borrower to see if they’re liquidatable, and as part of this, compute each borrower’s share of the overall stress test (their contribution to risk).
For computing this off-chain and signing off on the result on-chain, they keep a cut of the borrowers’ consensus staking rewards.
The computation simulates a shock to the borrower’s portfolio, considering the individual volatilities and correlations between the different consensus stakes inside the portfolio.
The shock represents the risk a borrower would add to the protocol over a year. Options math prices this risk as the value of a put option, and premiums are derived from that.
The put is executed automatically on behalf of all depositors, if and when the borrower’s collateral becomes worth less than their current debt. Such a borrower would be “in default”.
Execution of the put transfers a defaulted borrower’s bad debt and collateral to all depositors, proportional to their % contributions to solvency.
The transfer doesn’t happen all at once. All defaulted borrowers’ bad debt and collateral sits in a Dead Pool. Whenever a depositor updates their portfolio, they pull in their share of the pool.
In the meantime, the Dead Pool serves as an ATM. Holders of the protocol’s synthetic dollars can redeem them against the pool for a crypto basket of equal value, burning the pool’s debt.
In the rare case when the Dead Pool is empty, such redemptions are fulfilled by burning debt from the least collateralised borrowers and subtracting from their collateral in equal measure.
This would result in zero net loss for impacted borrowers, and improve their collateralisation as well as that of the protocol.
The opposite transaction can occur (“inversion”) where crypto is paid into the Dead Pool while dispensing synthetic dollars.
Whether there’s more redemptions or inversions hitting the Dead Pool, depositors always get their fair share of liquidated debt and collateral.
Borrowers can also avoid paying premiums, by paying with extra collateral instead. They get to choose high APY and low collateralisation or vice versa.
One can get an APY cost to borrow, that's next to nothing by starting out more collateralised (e.g. 30%). If one accepts the same maximum drop of 10% that one would get from borrowing at the minimum over-collateralisation (10%), then one's cost to borrow is extremely low.
Alternatively, one could pay a higher APY for having more room to drop. Paying the max APY would mean one's collateral could drop 30% before one gets liquidated.