DeFi liquidation mechanism protocol B.Protocol announced that it liquidated a 70,000 DAI debt from a MakerDAO vault that went underwater by utilizing a backstop system.
By doing so, the platform has shown that it is possible to efficiently run backstops in DeFi. The team shared the announcement in a Twitter post on November 29.
On November 26, the entire cryptocurrency market faced upheaval after a wave of high volatility. With Bitcoin dropping 15% in the span of several hours, numerous altcoins dumped by more than 20% as a result.
In such a situation with huge downside risks, the drop has affected lending & borrowing protocols the most. In fact, the CeFi ecosystem reached $2 billion in liquidations on that day. On the other hand, DeFi experienced more than $100 million in liquidations.
To compare their efficiency with MakerDAO, B.Protocol recorded a timeline of events. In a technical report, the team notes that their project so far manages 115 vaults and $7 million in deposits.
Moreover, they reveal that an ETH-A vault with a 70,000 DAI debt was successfully liquidated with the help of a backstop.
As with traditional risk mechanisms used in trading, the backstop prevented the vault from reaching an even bigger liquidation.
B.Protocol also notes that the backstop transferred 9ETH, accounting for 50% of its profits, to the B.Protocol Jar. This jar serves the function of collecting liquidation profits, which it then distributes to users.
B.Protocol backstop costs less, saves more
Within the technical report, the team stated that this event marks the first live demonstration of the backstop system. As recorded in the report’s timeline, the backstop behaved as expected and removed the overload that MakerDAO keepers would have.
Furthermore, we see multiple positive sides of using B.Protocol. Comparing liquidations processed by both platforms, B.Protocol reveals that their liquidations cost 250 gwei while 11 liquidations from MakerDAO cost 630 gwei.
The backstop protocol also had the chance to perform liquidations in ‘multiple chunks’ which allowed the vault operators to rebalance their position.
As Defiye reported on November 17, B.Protocol shared a new report regarding Maker’s liquidity engine. After performing a series of experiments, the team found out that the engine performs far worse than expected.
By dividing a total of $1 million into 7,800 vaults, B.Protocol successfully ran the vaults for hours before liquidation. In the report, the team states that it cost them only 0.5% of the total vault size to invalidate all future liquidations.
One more result of the experiments includes Maker’s liquidation model behaving randomly. As incentives become highly uncertain during periods of high volatility, the behavior of liquidations becomes unpredictable as well.
Maker’s liquidation cascade in March
As a reminder, the Maker protocol faced a gigantic liquidation cascade earlier this year. During the March crash, extremely high volatility resulted in a major loss of funds for vault owners who were undercollateralized.
While they did set $100 ETH as a price point that would liquidate their vaults, the oracles processed by Maker did not record the price at all. As a result, vault owners were liquidated at a far worse price despite setting precautionary measures.
The dramatic event concluded with a governance proposal in September, after which the community agreed to not compensate the affected March crash vault owners.