Is anything wrong with the DSR and Stability Rate calculations as planned for MCD?

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.

Current System

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.

Perceived Problem

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.

Example

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.

Discussion

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.

5 Likes

I think you’re spot on. I would like to hear more from someone in the interim risk team on how they plan on adjusting stability rates in response to fluctuations in DAI demand.

To me, the trade off seems to be an efficient centralized solution (risk team decides all risk premiums), and a less efficient decentralized one (MKR vote each risk premium change). Each solution comes with its own problems. While I think it is possible to change rates in a decentralized way, we may not be a stage in the project to allow this level of participation from voters. I think I will be okay with whichever route the risk team decides, but I would like to hear from them before launch about these valid concerns.

1 Like

there is no adequate tool

Wouldn’t the debt ceiling for each specific collateral type be sufficient to control for this kind of damage?

If there is a sufficiently low debt ceiling for Riskcoin, we can avoid having it cause damage to the general pool.

I think the currently proposed system is perfect.

The debt ceiling is a brute force tool that can resolve this, yes. But it is akin to saying ‘this problem is unfixable’ rather than fixing the problem. Hitting a debt ceiling means that we are leaving money on the table that we could be collecting by having a higher Stability Rate for that collateral package, while it’s perhaps correct to say that that outcome is ‘adequate’ is certainly not ideal.

This mirrors the discussions in SCD when the peg was broken. Could we have used the debt ceiling to maintain the peg when it drifted lower? Yes. Was the Stability Rate a better tool to resolve the problem? Also yes.

No system is perfect, blind faith in what is proposed is a fantastic way to get blindsided by a risk that could have been foreseen and mitigated.

If you dislike the phrasing ‘there is no adequate tool’ how would you rewrite the summary?

curtail a high demand

That’s one possible problem source, but another is simply needing to increase the DSR because DAI demand is too low. To me that’s a more fundamental issue.

Another attempt at summarizing the issue:

Equal price increases across multiple loan types due to exogenous reasons have unequal impact on how far out of market those loans become.

1 Like

The debt ceiling are set according to the risk tolerance of the DAO. If one collateral type is printing more it isn’t an issue assuming we’ve set the proper debt ceiling. It isn’t introducing any new risk to the system. The only way it would be an issue is if we’ve set the debt ceiling too high which we should obviously work to avoid. So in general I’d say its a non-issue.

Also

"Hitting a debt ceiling means that we are leaving money on the table that we could be collecting by having a higher Stability Rate for that collateral "

The execs at the foundation have explicitly told me they want to avoid milking CDP users, they’d much prefer to build trust with their users over a long term with better than elsewhere terms.

the answer is there is no good answer…

for a young volatile system, those massive swings will happen… and they will not be fun…

we need market forces / derivatives (and capital market bridges) to dampen the impact.

So the modified statement would be,

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. As such, capital market bridges and derivatives are needed to dampen the impact.

Perhaps it would be wise to split this into two statements, I feel there are two problems here:

  1. Flat-rate increases across multiple loan types due to exogenous reasons has unequal impact on the market competitiveness of collateral-packages.
  2. Under the proposed system and policy, there is no adequate tool for combating high demand for a specific collateral package without effecting the rest of the system.

Responding to your points on the debt ceiling:

This is true, there is no additonal risk, but there is additional Dai supply that effects the peg. Combating this with the DSR leads to either:

  1. Subsidising lower risk collateral packages (in the case where DSR > Stability Rate for those packages (this might be fine.)
  2. Increasing the base rate component of the Stability Rate for all collateral packages (I’d argue this is not fine, as it makes lower Risk Premium packages less market-competitive)

First of all, ultimately the DAO controls the governance parameters, not the Foundation execs, but you are right, we want to optimise for long term success, the DAOs image is important in that respect.

However, what I described is not ‘milking’ the CDP holders, it is allowing the Price Mechanism to work in the most efficient and fair way for the market. Take the following scenarios.


Scenario 1: We rely on the debt ceiling to curtail high demand.

A secondary market in CDPs develops shortly after we hit the ceiling and it becomes clear that governance consensus is to do nothing. The Price Mechanism works on this new market, CDPs are still available at a higher cost (a lump sum from buyer to seller).

Why is this bad?

  1. It privileges CDP holders that made their CDPs first. Participants should all pay the fair market rate. Our image takes a hit because different sets of users are paying different amounts for the same thing (secondary market buyers paying a lump sum).
  2. Some CDP Holders are still being ‘milked’, only now the money goes to first-come CDP holders rather than to the DCS (Dai Credit System.)
  3. UX takes a big hit. Now, rather than being able to open a CDP at any time with no counterparty, you require a counterparty to sell you a CDP, and you have to pay upfront for the privilege.

Scenario 2: We increase the Stability Rate specifically for that collateral-package.

No secondary market develops, none of the problems from scenario 1 happen. Instead we have:

  • An increased market rate that effects everyone equally. Participants that are unwilling to pay the new rate can refinance elsewhere or close their CDPs.
  • Total revenue for the DCS is higher.
  • When CDP Holders complain at the rise, we explain that it is a variable rate loan, and that the system is the same in essence as lending on other DeFi platforms (Compound.Finance, DyDx, etc) The only difference is that the rate change is coordinated by MKR token holders rather than automatically calculated. It is still dictated by supply and demand in the market.
1 Like

Perhaps it would be wise to split this into two statements

So (breaking the spirit of holding off on proposing solutions here) point 2. seems relatively trivial to me. It’s just a matter of adding a free term “Refrigerant” to the Stability Rate equation. Then, if governance wants to “combat high demand for a specific collateral package”, there’s a knob right there ready to be turned. Also the year 3000 version of that knob is a function from total package debt to refrigeration level. Anyway it doesn’t look like a fundamental issue, unlike point 1. Maybe I’m missing something!

Not 100% following you with the term ‘Refrigerant’

In technical terms there is already a knob, it’s currently referred to as the Risk Premium. In actuality this could contain two components: Risk Premium (risk of asset) + Specific Adjustment (to counter demand if necessary).

This is more of a terminology problem, it’s why I mentioned ‘policy’ in the suggested sentence. If the policy is to use this knob, the problem goes away (to be replaced with governance overhead if we have hundreds of collateral-packages.)

It’s all trade-offs really, I’m just worried that when the situation comes up, no one is going to like the trade-offs we are stuck with.

Yeah refrigerant is not great, my bad! I was thinking about the asset overheating, as in there’s a frenzy of speculation, and so we could “cool things down” in terms of CDP activity.

Thanks for the reply, and yes, agreed. I was unsure whether there was a wider issue I wasn’t seeing. When it comes to point 1., do you agree that it’s a more “fundamental” (whatever that means) issue?

We also have another dimension of time… Most of the items you have addressed are solved with time and a mature market… the challenge is how to address the two issues when the system is young that doesn’t contaminate the principle that we want to promote long-term.

In that context, I am more of a fan of finding / building capital market bridges to off load the excess DAI (and thereby contain the DSR’s rate) than I am to want to change the RP for a given collateral package when the risk didn’t change.

Moreover, I am now leaning to being more of a purist in hard split between Risk Policy and Monetary Policy… and DAI minting from collateral packages leads to an excess that impacts monetary policy, then we better build some bridges and hold true to our principles on why we set the RPs…

The simplest case would be the RP for ETH when we use MCD (not related to the transition… but what the RP would be in 12 months)… if we use the same 150% collateralization metrics… as the charts showed yesterday with RP (in SCD) being all over the place (including almost 0%) with 150% collateralization ration, during a liquidation, the closest any collateral came to 100% before being sold was at 143%… as ETH has a liquid market to sell collateral, the system was able quick sell collateral and ensure there were no losses to MKR holders…

So what should the RP really be? This a risk question… and intuitively, it should not be 18%… and it should not be zero… that said, if we have a “perfectly” liquid market, we are only insuring against the black swan events… realistically it should be way closer to 2% than to 18%…

point being is that when MCD launches and if the DSR is enabled… and IF we use the properly set RP… the market has already shown its hand… if the market in SCD will support 19% SF… the total SF (in MCD) will also end up being 19%… rather it will be just constructed there (meaning that if we assume the RP should be 5% (purely an example)… then all of the other costs (DSR + Oracle + Risk-team) should equal 14%… for a total of 19%… where we are today…

so the question is from a principle level, do we change the RP to an elevated rate thus allowing us to have a lower DSR. or do we set the RP to what we believe it is… and let the DSR “do its thing”… and float to really high levels…

From an external perspective, I believe the only path is the second one. With the second one, we allow for the value transfer from the borrower to the saver and MKR gets its “insurance” risk-premium… but it should not be more than is needed…

if the DSR is unusually high, that is a reason for individuals to buy DAI and lock in the DSR… and be motivated to build capital market bridges…

Even right now on dYdX, a DAI deposit can earn 16% (and 15% on USDC)… If we assume the contracts are legit, earning that return on cash in a “risk-free” model, is insanely high.

the only challenge here is that it is so elevated that it actually has the reverse impact… people believe there is no way that you can earn that with there being some mysterious risk…

This should be a market oppty…

I don’t think we will have a system in which we can only increase the DSR. Ultimately it has to be funded by the Stability Rate.

Having the Stability Rate solely based on risk, and using the DSR to absorb supply only works until the DSR needs to go higher than the Stability Rate. At this point we have to increase both together (adding the DSR component to the Stability Rate) to keep the system solvent.

I agree the Risk Premium should be set solely on risk, and that the other components, DSR + Governance Fee + Oracle fees should make up the rest to the current ~19% rate that the market is willing to pay. This works fine if demand is uniform across collateral-packages.

However if demand is not uniform, then we may want to use specific increases to the Stability Rate for a particular collateral-package, in order to prevent the negative effect the DSR funding component has on other packages.

This is largely a terminology problem, the technical lever labelled ‘Risk Premium’ can (but doesn’t have to) include two components: the actual Risk Premium and a Collateral Specific rate modifier to counter demand.

Hmm, having some thoughts on classification generally. Perhaps we should setup a generic scale for evaluating the risk of described problems, such as these ones and -ve DSR vs centralized stable coins as collaterals!

Perhaps a 1-10 scale where:
1 = I am happy to live with this risk forever
10 = I believe we will have to Emergency Shutdown because of this at some point in the future

Can we use the number rating poll type for this?


Flat-rate increases across multiple loan types due to exogenous reasons has unequal impact on the market competitiveness of collateral-packages.

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10

0 voters


Under the proposed system and policy, there is no adequate tool for combating high demand for a specific collateral package without effecting the rest of the system.

  • 1
  • 2
  • 3
  • 4
  • 5
  • 6
  • 7
  • 8
  • 9
  • 10

0 voters


Edit: Well, these polls do exactly what I wanted them to do, which is a nice surprise.
Edit the second: Closed these polls as I’ve created a separate topic for that purpose.

I 100% get your point. I do however disagree… and it is for this fundamental reason… when you were to bifurcate the Risk Premium into Risk and Risk Adjusted Demand… the ultimate risk premium gets paid to MKR holders… and that only makes sense if the actual risk went up… if the risk did not go up, then it is a monetary issue not a risk issue.

(I dont understand this part “Having the Stability Rate solely based on risk, and using the DSR to absorb supply only works until the DSR needs to go higher than the Stability Rate.”)…

As outlined the SF is a aggregate of the DSR + Oracle + Risk Team + RP… so if the DSR goes up, it will increase the SF by the same amount. but there into a scenario where the Oracle / Risk Team / RP would be negative… so the DSR can never go higher than the SF.

all the more reasons to get DAI out to the masses and get it locked into a DSR… and get points of sale accepting DAI as a method of payment with the default feature of depositing in the DSR right after receipt enabled… there are dozens of simple applications that will help offset an elevated DSR…

Hmm, actually I think it’s me that was misunderstanding things. So please ignore this part.

Having the Stability Rate solely based on risk, and using the DSR to absorb supply only works until the DSR needs to go higher than the Stability Rate. At this point we have to increase both together (adding the DSR component to the Stability Rate) to keep the system solvent.

I had some strange idea that Governance pocketing the RP meant we were earning more than we should be, but in actuality we’ll be ‘earning’ nothing, because the point of the RP is to offset the expected loss such that total net change is zero.

1 Like