Market Downturn Analysis: November 2022

As of 11/8 11:59pm UTC, Compound II has experienced no new insolvencies despite the recent market crash. Prior to the market crash, Compound VaR had greatly decreased this past week due to the market uptick and risky users updating their positions to be safer. Even after the crash, as of 11/8 11:59pm UTC the Value at Risk (VaR) in our simulations is $0, despite Liquidations at Risk (LaR) increasing to $169.59M.

Simulation Statistics

Value at Risk (VaR) time series since 2022-10-01

Liquidations at Risk (LaR) time series since 2022-10-01

Market Snapshot Stats

Total Liquidations by Assets since Nov 1st

Compound market liquidated approximately $747k in collateral assets yesterday. Asset cETH had the largest liquidation amount totaling $605.3k.

Liquidations on Nov 8th by Assets

Compound had liquidation repayment amounts of $580k yesterday with no large insolvencies.

Liquidation Repaid Amount by Assets

AVG Asset Collateralization Rates

AVG Collateralization Ratios represent the entire collateralization ratio of each user and calculate the average ratio by weighing the ratio by total asset supply. Collateralization Ratios did experience some significant changes from Nov 7th to Nov 8th. We would note that Sushi experienced a 30% increase in collateralization ratio.

Percentage Change in AVG Collateralization Ratios from Nov 7th to Nov 8th

Risky Users Analysis

In the charts below, β€œ11/9” refers to data from 11/8 11:59pm UTC.

User: 0xe84a061897afc2e7ff5fb7e3686717c528617487

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/9

User borrowing power breakdown 11/9

User borrows breakdown 11/9

User borrow usage 11/9

This user has updated their position multiple times today to increase their WBTC supply as WBTC price has continued to decrease, but their position is still riskier than it was prior to the crash (86% borrow usage).

User: 0x99e881e9e89152b0add27c367f0761f0fbe5ddc3

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/9

User borrowing power breakdown 11/9

User borrows breakdown 11/9

User borrow usage 11/9

User: 0xceef57f6c40a7cb2392eaad101ee0440aa43ba42

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/9

User borrowing power breakdown 11/9

User borrows breakdown 11/9

User borrow usage 11/9

This user previously had not updated their position since 2022-06-18, but this user updated their position today to decrease their supply to $7.67M WBTC and $32k COMP, and decrease their DAI borrows to $4.49M DAI and $423k USDT, which results in a borrow usage of 91%.

User: 0xedd54cc255ac1fe407f4113ec22a1d82fd5a2f71

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/9

User borrowing power breakdown 11/9

User borrows breakdown 11/9

User borrow usage 11/9

As of 2022-11-09 07:30 PM UTC this user increased their WBTC supply but still has a borrow usage of 91% due to WBTC price continuing to decrease.

User: 0x7e6f6621388047c8a481d963210b514dbd5ea1b9

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/9

User borrowing power breakdown 11/9

User borrows breakdown 11/9

User borrow usage 11/9

This large SUSHI supplier recently added $2.7M worth of COMP supply to increase the safety of their position.

5 Likes

Great analysis, @pauljlei and team!

2 Likes

As we’ve done with other Risk Analyses from Gauntlet, we add our user level credit risk assessment to β€œRisky Users” as defined by Gauntlet, using our credit scores.

The credit scores represent a borrower’s likelihood of default, or liquidation in the case of Compound. For example, strong credit score borrowers (1-4) are likely to β€˜top up’ their position’s collateral rather than being liquidated, and vice versa for weaker credit scores (7-10).

Below analyzes historical activity across all of DeFi for each β€œrisky borrower” – not just Compound:

Borrower: 0xe84a061897afc2e7ff5fb7e3686717c528617487

  • Credit Score of 3, strong credit score
  • 108 total borrows, all-time
  • Wallet age of 707 days
  • $125M in total debt repaid, all-time
  • Zero liquidations, all-time

Yesterday, borrower repaid $450K USDT and deposited 400 WBTC onto Compound, increasing health factor to 1.24. Price being 24% before liquidation coupled with strong credit score make this position less than risky than perceived.

Borrower: 0xceef57f6c40a7cb2392eaad101ee0440aa43ba42

  • Credit Score of 9, weaker credit score
  • 60 total borrows, all-time
  • Wallet age of 422 days
  • $2.3M in total debt repaid, all-time
  • 14 liquidations, all-time
  • $27M total value liquidated, all-time

Yesterday, borrower deposited 387 WBTC bringing their health factor to 1.15. Price being only 15% before liquidation coupled with weaker credit score make this position more risky.

Borrower: 0xedd54cc255ac1fe407f4113ec22a1d82fd5a2f71

  • Credit Score of 8, weaker credit score
  • 38 total borrows, all-time
  • Wallet age of 456 days
  • $44M in total debt repaid, all-time
  • 4 liquidations, all-time
  • $8.1M total value liquidated, all-time

Borrower deposited 67 WBTC bringing their health factor to 1.17. Price being 17% before liquidation coupled with weaker credit score make this position more risky.

Borrower: 0x7e6f6621388047c8a481d963210b514dbd5ea1b9

  • Credit Score of 9, weaker credit score
  • 104 total borrows, all-time
  • Wallet age of 822 days
  • $522M in total debt repaid, all-time
  • 5 liquidations, all-time
  • $3M total value liquidated, all-time

Borrower deposited 3.1M COMP bringing their health factor to 1.66. Despite the weaker credit score, the significant health factor makes this position less risky to Compound.

1 Like

As of 11/9 11:59pm UTC (which we refer to as 11/10), Compound has experienced no new insolvencies despite the recent market crash. Prior to the market crash, Compound VaR had greatly decreased this past week due to the market uptick and risky users updating their positions to be safer. Even after the crash, as of 11/10, the Value at Risk (VaR) in our simulations is $0, despite Liquidations at Risk (LaR) increasing to $182.67M. Also note that the market has since had an uptick, and as a result, the protocol is more highly collateralized than when we ran our simulations. However, we will still show the analysis as of 11/9 11:59pm UTC.

Simulation Statistics

Value at Risk (VaR) time series since 2022-10-01

Liquidations at Risk (LaR) time series since 2022-10-01

Note that the LaR is substantially higher than realized liquidations in the past couple of days. This is because 1) more accounts are closer to their liquidation thresholds after the most recent crash, and 2) LaR corresponds to liquidation amounts from simulations that model even more drastic market crashes than what we’ve experienced recently.

Market Snapshot Statistics

Asset price changes from 2022-10-07 to 2022-10-10

Total Liquidations by Assets since Nov 1st

Compound market liquidated approximately $8.8M in collateral assets yesterday. Asset cWBTC had the largest liquidation amount totaling $7M.

Liquidations since Nov 8th by Assets

Compound had liquidation repayment amounts of $7.3M since Nov 8th.

Liquidation Repaid Amount by Assets

AVG Asset Collateralization Ratios

AVG Collateralization Ratios represent the entire collateralization ratio of each user and calculate the average ratio by weighing the ratio by total asset supply. Collateralization Ratios did experience some large changes from Nov 7th to Nov 9th. We would note that COMP experienced a 30% decrease in collateralization ratio.

Percentage Change in AVG Collateralization Ratios from Nov 7th to Nov 9th

Risky Users Analysis

Below we analyze 5 risky users in Compound. All 5 of them have updated their positions through the recent crash, thus avoiding liquidations. Most of their positions as of 11/9 11:59pm UTC were still riskier than prior to the crash. However, the market has since had an uptick, and we mention their updated lower borrow usages as of 11/10 9:00pm UTC.

In the charts below, β€œ11/10” refers to data from 11/9 11:59pm UTC.

User: 0xe84a061897afc2e7ff5fb7e3686717c528617487

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/10

User borrowing power breakdown 11/10

User borrows breakdown 11/10

User borrow usage 11/10

Borrow usage as of 11/10 9:00pm UTC has decreased to 73.5%

User: 0x99e881e9e89152b0add27c367f0761f0fbe5ddc3

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/10

User borrowing power breakdown 11/10

User borrows breakdown 11/10

User borrow usage 11/10

Borrow usage as of 11/10 9:00pm UTC has decreased to 76.9%

User: 0xceef57f6c40a7cb2392eaad101ee0440aa43ba42

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/10

User borrowing power breakdown 11/10

User borrows breakdown 11/10

User borrow usage 11/10

Borrow usage as of 11/10 9:00pm UTC has decreased to 80.0%

User: 0xedd54cc255ac1fe407f4113ec22a1d82fd5a2f71

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/10

User borrowing power breakdown 11/10

User borrows breakdown 11/10

User borrow usage 11/10

Borrow usage as of 11/10 9:00pm UTC has decreased to 76.9%

User: 0x7e6f6621388047c8a481d963210b514dbd5ea1b9

User supply time series since 2022-11-07

User borrows time series since 2022-11-07

User borrow usage time series since 2022-11-07

User supply breakdown 11/10

User borrowing power breakdown 11/10

User borrows breakdown 11/10

User borrow usage 11/10

Borrow usage as of 11/10 9:00pm UTC has decreased to 55.2%

1 Like

@pauljlei any chance you could post the link in the image that further defines VaR or a reference to the VaR methodology you’re using? Thanks!