This is a consolidated version of my topic on reddit, many thanks to those that commented there. I feel this could use further discussion though, so if you commented before, please use this as an opportunity to summarise your objections or agreements and try to form consensus.
There has been some discussion recently over the potential for problems surrounding the DSR and the Stability Rate calculation for MCD.
These are not simple concepts, so I think it would be useful for Governance to generate a problem definition that we can get consensus on, or at least, that no one actively objects to.
With that in mind, here is the relevant information on the currently proposed system. Please correct me if I am mistaken with these statements:
- Collateral Package : A combination of a collateral type, along with attached values for the Risk Premium, Collateralization Ratio and the Debt Ceiling. There will be one or more collateral packages for each collateral type where these values are balanced differently.
- Risk Premium : A collateral specific, semi-fixed value solely determined by the output of the risk model when applied to a specific collateral type. This changes only when the underlying risk of the collateral changes.
- DSR : Dai savings rate, a system-wide payout to holders of locked Dai. Proposed to be used in the Stability Rate calculation as below.
- Governance Fee : The fee we take for governing the system, which eventually ends up buying and burning MKR.
- Stability Rate : The APR that CDP holders pay to mint Dai, which varies with the type of CDP.
The current plan is for the stability rate for each collateral package to be calculated as:
Stability Rate = Risk Premium + DSR + Governance Fee
The DSR is planned to be the only lever used to control Dai supply/demand balance, and is used in the Stability Fee calculation as stated above.
This is my understanding of the issue, please chime in if your understanding differs, maybe I’m wrong!
- DSR directly affects the utility of holding Dai, and therefore demand. The DSR is a single term equation: Utility of Holding Dai = DSR .
- The Stability Rate however is only one factor that controls the utility of maintaining a CDP of a certain collateral type, it is a multi-term equation. Utility of holding a CDP = Average perceived utility of Leverage with that collateral / (Stability Rate * Collateralization Ratio)
- Under the proposed system there is no dynamic, collateral specific component to the Stability Rate equation. The Risk Premium is collateral specific, but it is not proposed to be used dynamically. The DSR component to the equation is dynamic but it is not specific. The result is that there is no good option to react to a collateral-specific dynamic change in leverage demand.
- We will be forced to fix a collateral-package specific CDP demand problem by using a tool that effects general CDP demand (as well as Dai demand). This has knock on effects on the balance of collateral in a way that causes the average risk of locked collateral to increase as the DSR increases.
For simplicity’s sake, assume the following:
Stability Rate = DSR + Risk Premium
The DSR is 0%.
We have three assets with different risk premiums (and since the DSR is 0% these are also their Stability Rates): SafeCoin at 1%, Mod(erate)Coin at 10%, RiskCoin at 40%.
If we increase the DSR from 0% to 5% then we have the following effects:
SafeCoin goes from 1% to 6%, a 500% increase over the previous rate.
ModCoin goes from 10% to 15%, a 50% increase over the previous rate.
RiskCoin goes from 40% to 45%, a 12.5% increase over the previous rate.
If possible, I’d like to keep discussion and comments solely focused on the nature of the problem (if one exists). If you think you have a solution, great! Write it down and save it, we’ll get to that part later. What I would love for this post to achieve is a consensus problem statement: something no longer than a sentence that we can agree on and use to describe this problem. Here is an initial attempt:
Under the currently proposed system and policy there is no adequate tool that can be used to curtail a high demand for a single collateral package without negatively affecting other collateral packages, especially those with lower risk premiums.
Please suggest changes as you feel necessary, and we will try to come to consensus.