The past two weeks have been a major stress test for the Ethernet Decentralized Finance (DeFi) ecosystem, which has users worried due to the high volatility of ETH prices, high gas fees and liquidity. At the same time, the past two weeks have provided a very important stress test for the various organic components of DeFi. For example, by many metrics, the past two weeks have seen even more congestion on the Ether network than 312. Furthermore, assets held by Compound and Aave have risen more than 100 times since 312, which could lead to a more dangerous spiral of liquidation. This article will analyze some advanced statistics from the Aave protocol and plans for the future. Overall, Ether’s on-chain lending protocol performs well under these extreme conditions!
Liquidations occurring on centralized exchanges
To understand what is happening at DeFi, let’s take a look at the liquidation of centralized exchanges. The most liquidated day on a centralized exchange during the recent market sell-off occurred on May 19. In Figure 2, we can see that centralized trading platforms liquidated a total of $70.9 billion on the long side (Firecoin reached $2.95 billion). This figure is higher than the total value of liquidations that occurred on April 22, and also exceeds the $9 billion liquidation that occurred on April 17. $4.7 billion worth of positions were closed out on April 17, which could partially explain the significant drop in liquidation volume of $870 million on May 19. In addition to liquidations, several centralized exchanges have also experienced technical difficulties.
Figure 1: Amount cleared by centralized exchanges in 2021. (Source: https://www.theblockcrypto.com)
Figure 2. centralized exchange long clearing volume 2021. (Source: https://www.theblockcrypto.com)
On-chain protocol clearing
From the analysis above, we can tell that May 19 was the day when the bulk of the liquidation occurred. While this was not the lowest price for ETH in recent memory, it did see the most dramatic intraday volatility. eH fluctuated about 41% from a high of $3,400 to a low of $2,000. (On centralized exchanges, some even saw prices as low as $1,700.) As a result, both Aave and Compound experienced the largest single-day liquidation totals in their history.
From May 17 to May 23, Compound liquidated collateral totaling approximately $260 million and Aave liquidated collateral totaling approximately $310 million. At that time, Compound and Aave had $10 billion and $13 billion worth of collateralized assets, respectively. Please note that the value of liquidated collateral assets is priced at the closing date of the liquidation and may vary if the price closest to the liquidation is used as a reference due to the sharp price movements. the largest single day liquidations occurred on May 19 for Compound (approximately $160 million) and Aave (approximately $170 million). Below we take a closer look at the specific collateral and repayment assets of the Aave liquidation.
Figure 3: Liquidation of Aave’s collateral assets and reimbursement assets, May 19, 2021.
In the chart above, the total value of liquidations is broken down by repayments. As expected, most of the liquidations are unstable cryptocurrency assets collateralized by stablecoin borrowings. Another calculation that is also important is to convert the total liquidation to the total supply of each asset. The ranking of liquidation volumes on a pro-rata basis is
Gas fee spike and network congestion
From May 17 to 23, the top 50% of the highest gas bids were above 100 gwei and the top 1% of gas bids were around 1,000-4,000 gwei. On March 11 and March 12, 2020, the top 50% of gas prices were above 60 gwei and the top 1% were above 400 gwei. High gas fees will have a negative impact on the Aave agreement (see our report for details). In general, both top-up and clearing transactions close to the clearing amount account are very important for the solvency of the agreement. During periods of high gas and network congestion, these transactions may not be processed in a timely manner, resulting in users’ inability to pay their debts. To assess the impact of network congestion, we looked at the Gauntlet model for clearing and borrower behavior. Note that in Figure 4, most Aave clearings on May 19 do not require high gas quotes. However, there were more than 30 liquidations in the 2,000gwei of gas bids, which is worrisome. Since liquidators are willing to pay such high gas fees, this will eventually lead to more liquidations and a worse user experience.
Figure 4: Gas prices corresponding to liquidation events, Aave agreement, May 19, 2021.
Figure 5: Liquidation size vs. gas price for all liquidation events, Aave protocol, May 19, 2021.
It is also important to compare the behavior of borrowers during the same period. In Figures 6 and 7, we look at the top-ups (or additional deposits) on the Aave agreement. Note that in Figure 6, all gas bids above 2000 gwei are included in 2000 gwei. Although the number of recharge transactions is about 10 times higher than in a normal period, the number of transactions over 2000 gwe gas bids is significantly lower during the same period.
Figure 6: List of gas prices used for recharge, Aave agreement, May 19, 2021.
Figure 7: Collateral size vs. price of all rechargegas, Aave protocol, May 19, 2021.
Figure 8 is a valuable comparison of the relative difficulty of the combined recharge amounts relative to each collateral asset. The average gas bids used to replenish certain assets (LINK, CRV, YFI, UNI) are significantly higher than others.
Figure 8: Total recharge volume and average gas bids by asset replenishment, Aave protocol, May 19, 2021.
Prior to the failed recharge event, there were 18 different liquidations on the Aave protocol. Supplemental loans are attempts to repay debt or increase collateral prior to liquidation. This metric definitely does not take into account users who are discouraged (in terms of deposits or repayments) from topping up due to extremely high gas prices, but it does illustrate how network congestion affects users of the protocol.
Figure 9: Liquidation size of failed recharge accounts vs. gas prices, Aave protocol, May 19, 2021.
In addition, more than half of the outstanding loans are due to insufficient gas fees. About $35 million of collateral was liquidated due to failure to repay.
Figure 10: Liquidation behavior of accounts with failed top-ups on the Aave protocol, for May 19, 2021.
Figure 10 depicts liquidators who face liquidation of their accounts due to failed top-up attempts. Note that these liquidators are quite sophisticated in terms of fine-tuning (increasing the price of gas in small denominations) and private transaction behavior.
A key focus of our market risk assessment is to understand and model the resilience of the agreement. One of the key metrics used to determine resilience is net asset insolvency, i.e. underwater debt.
In early March, when the deposit size and total lockup value of Aave V2 began to outpace V1, there was already approximately $14,000 worth of insolvency debt on the Aave V2 protocol, but $10,000 of that insolvency debt was an account borrowing from ENJ, UNI, BAT, and ZRX. most were very small positions that the liquidator could not profitably close. As of today, there is approximately $60,000 worth of insolvency debt on the Aave agreement. There is only one account over $3,000 that cannot be repaid. Most of the collateral used in the account is WETH, LINK or YFI. with a total locked position of over $10 billion, this is less than 0.05 bps of the agreement insolvent! A very promising statistic for the future of the Aave protocol and its users!
The next step
According to Gauntlet’s key metrics in terms of protocol security, protocol insolvency and insurance fund reductions, the Aave protocol performs well. In addition, Gauntlet now has an overabundance of data to test and validate the key assumptions and agent behavior in our simulations. At the same time, the large number of liquidations that have occurred does not indicate an optimal user experience. While Aave has withstood stress tests, these events have exposed the problems facing the protocol:
How can the user experience be improved in the future by reducing liquidations (total liquidations and loss of liquidation bonuses)?
Besides collateral and health factors, what are the key indicators of account solvency that we can continue to monitor?
How has debitor behavior changed over the last few weeks and can the protocol update parameters to facilitate beneficiary user behavior?
One way to address these questions is to allow the protocol to update the risk parameters based on market conditions. In the past, analysis has focused on solvency and reducing insurance fund reductions, but this is only one aspect of protocol optimization. As Aave evolves and matures, balancing safety with maximizing user returns will become an increasing priority.
Posted by:CoinYuppie，Reprinted with attribution to:https://coinyuppie.com/how-did-the-aave-protocol-perform-as-the-market-plunged-in-may/
Coinyuppie is an open information publishing platform, all information provided is not related to the views and positions of coinyuppie, and does not constitute any investment and financial advice. Users are expected to carefully screen and prevent risks.