Collateral Selection - Autonomy / Predicitability (reducing risk) / Transparency

Full post here if interested

Providing tiers and the exposure they’re able to carry.

Exactly like a bicycle gear system.

or this
(http://www.dimitri.co.uk/still-life/still-life-images/silver-cogs-inside-circular-on-blue.jpg)

Would it be possible to just copy-paste the original posts (this and your other) to this forum?

It seems like it would be easier to digest that way. Without following the full-post link, I have no idea what this post is really about.

3 Likes

Sorry! Yes of course. Appreciate the critque, I’ll just make this OP moving forward.

Any feedback on the following would be awesome!

Is there ever a thought with the collateral selection to perhaps structure the collateral pool in something like 90% to the collateral, and then 10% for newly approved collateral types.

Vote —> Approved —> Now in the 10% portion.

Then over some set of metrics that matter most to the risk team, it would do a periodic (after n blocks, or Dai in circulation relative to ceiling) you then demote the worst two performing collateral types, and then the 10% pool.

Example

Dai debt ceiling set at $1B and MCD is in full flight. Highly anticipated newCollateral is voted in and assets have been voted in. Bullish. Over the coming 24 months, momentum has stalled, and we see the Dai in circulation hit 35% of the ceiling. Prior to this, the Dai circulating was $700m

So it’d look something like:

  1. Vote an asset into the audition pool (10% pool). 2) At the time of the next milestone collateral redistribution happens. Some mechanic I’m imagining if you have 10 types live as collateral, and there’s 4 types in the audition pool, the bottom two assets from the main get demoted, the top two from the audition pool get their entrance into the primary collateral pool.

Having some structure like this will help mitigate the impact to some networks that handle the response differently to others when a large additional use case for their asset is introduced.

Generating what would become similar to an assembly line into MCD. Strengthening the predictability of the protocol and how the market is responding through partitioning the pool is probably going to give the risk team another perspective to evaluate the system and optimization of their own processes etc…

Creating predictability (if done correctly) likely yields a faster path to capturing a larger commitment from CDP creators, Dai users, etc…

Btw, this is not something I’m standing and stomping my feet to have or anything, I just thought it would be worth circulating for discussion.

Practical example is a lot of liquidity aggregators are trying to find the best price for their customer and will implement these mechanics for trade execution. Incentivizes optimal performance other wise the probability of being routed an order to your exchange decreases based on last performance. Risk is never a silver bullet/destination, it’s a sliding scale where you can never really become risk free, but you de-risk repeatedly and those iterations get harder and harder to find.

Food for thought!

P.s Apologies if this has already been covered whether briefly or at length, it’s tough to keep up with everything over the bajillion random discussions.

Thanks for reading if you got this far!

If anybody is at EthDenver, Wyre will be there!Providing fiat to crypto ramps for platforms and have Dai listed hard pegged 1:1 for users wanting to buy/sell. We’d love to meet you and hangout! If we can be helpful in the hackathon too, we’re all about it.

Geez, I haven’t stopped f*ing typing. My god.

1 Like