Liquidators' Incentives

The preferred protocol outcome is when new bidders mostly assume a Borrower (Bob)'s liquidation. This lowers Protocol churn and replaces Bob's capital with new capital.

Continuing from the previous example, Alice has several sources of profit: the free arb of Bob's vault triggering a Liquidation (5%), the discount at which Alice bid on Bob's debt (10% in Alice's case, a maximum of 16%), and the pro-rata liquidatorIncentive, which will vary by Auction but which we are targeting to be 0-5% of the total value of the liquidated debt. Note that these rewards are all denominated in BTC-overcollateralised UNIT in the Liquidator's final custody and don't assume any rewards for Liquidators in the form of governance token airdrops/emissions in the future.

Thus, a Liquidator can expect an arbitrage opportunity of 5-25% per Liquidation, a handsome risk/reward over any conceivable 10-minute (1 BTC block) time interval.

Why the Protocol incentivises Liquidators so much

Liquidators recycle new capital into undercollateralised Vaults and onboard new capital into the Protocol.

Any Liquidator can pay down a newly acquired debt immediately. From the Protocol's perspective, this is still a win, albeit less so if the Liquidator's capital stays with the Protocol.

The Rolling Debt Auction, or centralised liquidation, is more profitable for the Protocol in the short term because the Protocol pockets a 20% liquidation tax at the cost of very low duration risk. However, the Protocol also loses Bob's capital, loses the TVL represented by Bob's Vault, and presumably loses Bob as a user. Furthermore, the Protocol undertakes (very modest) duration risk when liquidating debt.

By contrast, the Rolling Debt Auction use case will have simpler UX, higher liquidity, higher visibility on bids getting filled, and higher velocity of participants' capital. Therefore, it should require proportionally fewer rewards to achieve a critical mass of user participation.