Reduce STBL collateral factor to 0%


Reduce the STBL collateral factor from 80% to 0%. This change doesn’t impact the collateral factor of STBL-USDC LP tokens.

STBL can get de-pegged from $1 for an extended period of time. Moreover, an unlimited amount of STBL (up to the hard-coded platform limit) can be created instantly. This new money can then be used as collateral with the $1 hard-coded value. The combination of the above two facts breaks various lending protocol assumptions, which can lead to bad debt and unhealthy state. This is a risk for all users no matter if they ever use STBL or not. The proposal aims to remove this specific STBL risk from the lending protocol and for its users.

Right now STBL only represents ~222k of collateral and applying this change per the Algofi team’s opinion should be relatively easy.


  1. How is STBL de-pegging different from USDC de-pegging?
    Unlimited amounts of STBL can be created and re-used instantly, whereas that’s not possible with USDC, which makes it a much more severe issue. Moreover, the chances of STBL de-pegging are strictly higher as STBL depends on USDC for its pegged value.

  2. Why does the STBL-USDC LP tokens collateral factor remain the same?
    For these LP tokens for every STBL a USDC should be provided, which works as a speed bump, and makes the severity of the de-pegging issue for LP tokens close to the USDC de-pegging. Alternatively, depending on its feasibility, the collateral factor could be cut in half to be consistent with the generic collateral formula.

  3. What if instead of $1 the live price of STBL used in the lending protocol?
    The reliability of an asset price depends very much on the depth of the order book. The STBL liquidity pools are tiny for this purpose. Moreover, using a non-constant price may complicate both sides of the lending protocol.

  4. How is it not an issue with DAI?
    DAI has 50+% of its reserves in USDC.

Temperature Check Post


I voted against this proposal in haste. After a couple weeks of consideration, and reading the discussion ; I would vote “for”. I was wrong.