Market Risk Monthly (February 2022): Updates and Review

This report builds upon Gauntlet’s Market Downturn Risk Review, with the goal of keeping the community informed of market risks pertaining to Compound. So far, Gauntlet has executed 7 sets of parameter recommendations as part of our Dynamic Risk Parameters engagement. Over the past two months, Gauntlet has executed 2 sets of recs and has implemented seven (7) parameter suggestions across seven (7) assets.

Old Collateral Factor New Collateral Factor
ETH (80%) 82.5%
UNI (70%) 75%
LINK (70%) 75%
MKR (55%) 60%
AAVE (60%) 65%
YFI (60%) 65%
SUSHI (55%) 60%

Gauntlet’s parameter updates align with the Moderate risk level that the Community has collectively voted for.

In addition to managing market risk for Compound, Gauntlet also has invested in creating educational resources to help the Compound Community, and the broader DeFi community, better understand market risk. Gauntlet has recently published the following educational documentation and advises that the community review these resources:

Gauntlet has launched the Compound Risk Dashboard and has invested in improvements to their infrastructure to enable daily updates. The community should use the dashboard linked above to better understand the updated parameter suggestions and general market risk in Compound. Gauntlet is keen on continuously bettering their dashboards and has been conducting user studies with members of this community and others to inform future iterations.

Highlights

Supply changes of key collateral assets from 01/25/2022 to 02/28/2022

Borrow changes of key collateral assets from 01/25/2022 to 02/28/2022

Price changes of key collateral assets from 01/25/2022 to 02/28/2022

Liquidations breakdown by collateral from 01/25/2022 to 02/28/2022

Liquidations breakdown by account from 01/25/2022 to 02/28/2022

Top 10 liquidations breakdown by account from 01/25/2022 to 02/28/2022

Insolvent accounts on 02/28/2022

New insolvent accounts from 01/25/2022 to 02/28/2022

This shows all insolvent accounts on 02/28/2022, which were solvent on 01/25/2022. Note that the only insolvencies are from negligible dust accounts, the largest of which has $1,600 supplied and the largest insolvent amount of $204.

Previous insolvent accounts from 01/25/2022 and differences up to 02/28/2022

This shows the 02/28/2022 state of all accounts which were insolvent on 01/25/2022.

Total insolvencies from 01/25/2022 to 02/28/2022

90-day time series of collateral-specific volatility

Using the 28 day trailing data and the Garman Klass volatility estimator, which utilizes both daily high H_i and low L_i prices, as well as daily opening O_i and closing C_i prices, the annualized volatility for each asset can be calculated as follows:

Screen Shot 2022-03-09 at 11.55.47 AM

If you are unfamiliar with the concept of volatility, please reference this writeup from Gauntlet.

60-day time series of collateral-specific protocol collateralization ratios

The average collateralization ratio for an asset is a valuable indicator of how risky borrowers are behaving. The recent price drop has led to decreases in average collateralization across certain assets.

90-day time series of liquidation volume

2 Likes

@pauljlei this is great. I am beginning to look forward to these reports each month :blush:

Can you please clarify the following?

  1. Looking at the graphics “Previous insolvent accounts from 01/25/2022 and differences up to 02/28/2022” and the next one “Total insolvencies from 01/25/2022 to 02/28/2022”
    Some accounts continued to become more insolvent. Is that because they are dust accounts and liquidators didn’t bother to spend gas money to liquidate those accounts? If that’s the case, in theory, dust accounts can continue to grow and accumulate losses for the protocol indefinitely, correct?

  2. One account sustained a liquidation of $12.8M. I think this would make an interesting story, and wondering if you can provide additional details – such as, the date of liquidation (high volatility day), collateral / borrowed assets, LTV used by the user (must’ve been very close to the CF requirement, with low margin for volatility) etc.

2 Likes

Hi @RogerS , great to hear you are finding these Risk Reviews helpful! To answer your questions:

Yes, a likely reason why these dust accounts were insolvent is because liquidators couldn’t liquidate these accounts profitably. In theory, these dust accounts could grow indefinitely (from a USD perspective, not # of tokens), but that would mean asset prices in general are likely to be growing. As such, these dust accounts are unlikely to present outsized systemic insolvency risk to the protocol.

As for the liquidation: liquidation events happen on-chain, so if you would like further details on this event in particular the details can be found via sources like a block explorer.