[Signal Request] Raise DSR

Overview

We have set the DSR to 0% in September 2020 during times of being highly above the peg and relatively low stability fees to fix that.

Since then, we made really good progress on fixing the peg and fixing the supply side of DAI. Over the last weeks, the DAI-from-stablecoins is declining and being more and more replaced by DAI-from-ETH and other non-stablecoin vaults.

In this thread we started discussing bringing back the DSR to a non-zero level again, and it seems that there are at least some people thinking this is something we should consider.

Increasing the DSR

To give some numbers on how much of the fees are going to get spent on the DSR, it is probably worth the time to have a look at the different dimensions that needs to get factored into the equation

  • Fees/DAI: right now we earn 2.38% on each DAI minted. This number is influenced by the utilization of the different vault-types and their stability fee (stablecoin vaults are running on 0% fees)
  • Utilization of the DSR: not every DAI in circulation earns interest, just the DAI locked into the DSR. Compound had DSR integrated into its protocol, while we had a (un)reasonably high DSR (was once up to 8.75%) around half of the DAI was locked up there.

Given the current stats

DAI in circulation in MM Average fee per DAI Annual Revenue (Fees) in MM
1600 2.38% 38

We would end up with those costs / fee-revenues: (google spreadsheet to clone)

DSR DAI in DSR Annual DSR Cost in MM DSR % of Revenue Annual Revenue (Fees - DSR) in MM
2.00% 50% 16.000 42.11% 22.000
2.00% 30% 9.600 25.26% 28.400
1.00% 50% 8.000 21.05% 30.000
1.00% 30% 4.800 12.63% 33.200
0.50% 50% 4.000 10.53% 34.000
0.50% 30% 2.400 6.32% 35.600
0.25% 50% 2.000 5.26% 36.000
0.25% 30% 1.200 3.16% 36.800
0.10% 50% 0.800 2.11% 37.200
0.10% 30% 0.480 1.26% 37.520
0.05% 50% 0.400 1.05% 37.600
0.05% 30% 0.240 0.63% 37.760
0.01% 50% 0.080 0.21% 37.920
0.01% 30% 0.048 0.13% 37.952

Pros:

  • incentivize 3rd party apps to use DAI since they/their customers would benefit from the DSR (even if the chosen rate is low)
  • drives demand (pretty questionable, since rewards from farming are a lot higher than even 1% - but probably some farms will integrate DSR too)

Cons:

  • we lose some of the fees that would go into the surplus buffer, only taking in account DAI inside Compound a 1% DSR would cost 2.3M DAI annually (5% of actual lending revenues).
  • DSR is meant to be a tool for monetary policy to create upwards pressure on the peg - we aren’t below peg so this is basically a PR stunt

Please vote on ALL options you would support in an onchain poll. I know there are probably a bit too many options, but since this is uncharted territory, i decided for more of them.

  • 1%
  • 0.5%
  • 0.25%
  • 0.1%
  • 0.05%
  • 0.01%
  • 0% (no change)
  • Abstain

0 voters

Next Steps

The Poll will run until February 11th; its outcome will result in a on-chain-poll assuming the outcome of the poll deems it necessary.

5 Likes

Given the PEG mechanics are PSM dominated I am still waiting to see any thinking about whether and how to manage SF and DSR rates in the face of this.

Both the SF and DSR were PEG management tools predominantly.

Right now DAI demand is high. If we closed the PSM DAI PEG would probably rise.

Hence I am back to wondering exactly what the DSR is for at this point. In the past it made sense as a demand enhancer (i.e. PEG was below 1). That will only happen if the PSM runs out of USDC or vaults unwind and then we have SF, and then finally the DSR to raise.

DAI demand is already high without adding to this with a DSR rate to put DAI into a place where it serves absolutely no purpose (other than removing it from the markets). If we just want to pull DAI from markets to increase demand just increase the surplus buffer, raise SF. I don’t understand why Maker needs to compete with DeFI on DAI return at this point. To my mind with the PSM the DSR is actually no longer required ‘except’ as the last resort DAI sop up mechanism which means the DSR literally is the LAST thing to touch when the PEG is below the bottom of the PSM band or we have run out of USDC. imo ofc.

Hence I polled for 0 DSR until I see a DAI PEG below 1.

Honestly want to see a treasury/reserve in the 4-10% of outstanding float, or a DAI PSM exhausted and DAI PEG below .99x and high SF before this is even considered. Network costs just to deposit in the DSR even at .1% at this point probably don’t make sense either.

I see compound is paying 4.35% (not including COMP) so honestly would rather see people going there for the return vs. juicing the compound returns with Maker profits for no real market reasons.

6 Likes

To reiterate some of the points I made in the other thread,

  • We still need to flush stablecoin vaults and the PSM that is at 0% SF.
  • Questionable what effect the DSR will have considering other options for yield.
  • Increased gas costs for other defi protocols to integrate the DSR.

And I also liked @SebVentures proposal to do some Airdrops if you wanted to do something.

I still voted for some other options other than 0% as I may still support them in an onchain poll.

1 Like

as @ultraschuppi wrote in his post: “incentivize 3rd party apps to use DAI since they/their customers would benefit from the DSR (even if the chosen rate is low)”.

I am not entirely sure this will be the case with, e.g., DSR = 0.01% or 0.1%.

But the point is that the DSR is a useful PEG-management tool if it is integrated (read: used) by other DeFi protocols, and when we want to activate it, it is ready to go. So I kind of agree that stimulating DSR-integration (to some extent) is important.

1 Like

It will be a demand enhancer when needed, which is not now.
I think the PSM will be a useful leading indicator. Watching its balance change will be useful information regarding when to adjust the DSR and non-Risk Premium portion of stability fees (which we might call Base Rate). Optimally, this would be done before the PSM runs out, to keep it in balance, rather than as a reaction to it running out.

1 Like

The only rational argument I see now for raising the DSR is the idea that it might spur integrations, but I doubt a tiny non-zero DSR would do that.

I think projects will integrate because they know there will come a time when the DSR is the best rate available (and lowest risk).

If DSR is less than 0.1%, it is best to choose 0%, because less than 0.1% has no meaning. Then let us continue to wait until most people reach a consensus.

1 Like

I guess anything above 0% will make integrators/apps believe that in some not to far future the DSR will make a difference and start favoring DAI over other stable coins - even if there is no direct effect right now.

@MakerMan great to see you back in action on the forum again :slight_smile:

maybe it is still the maxi in me and a total pipe dream: but longterm I would like to be totally free of USDC in standard situations and just have the PSM (with higher tin and tout) for severe peg-fixing and the DSR as the standard weapon to create upwards pressure.

Of course we should not raise the DSR to substantial levels before our dependency to 0%-vaults is removed.

Still, the DSR should not to be forgotten as a real feature, especially as a USP for merchants, apps and even farms - and that is the main reason I was thinking about making a move on the DSR now

1 Like

Nice of you to say. I still own and vote MKR so I guess I still feel the need to comment when I have time.

IF some tiny rate would spur integrations - fine. But honestly with main-net fees high I just don’t see the point and the vast majority of the DSR deposited comes from compound which is already earning 4.3ish%

I completely agree that it would be nice to widen the PSM and use normal SF and the relatively new DSR to stabilize the PEG. One thing I have hinted on is the idea of having a lot of USDC on the books as a tool to manage the low side of the PEG at various bottom PEG side spreads (can’t remember if this is tin or tout - i.e. PEG under .999 vs. just burning our buffers at .999).

I think my points were:

  1. PEG mechanics are different with the PSM
  2. DAI demand is already very high (why we are here with these LR101 vaults and PSM in first place)
  3. SF is used to increase DAI demand first (as system profits)
  4. DSR is used last.

Additionally with the PSM in place determining how the system measures and responds to demand is important.

I don’t want to forget about the DSR. I just think NOW is not the time to raise it since it sends the wrong message generally about what governance thinks about DAI demand and rates at 0 or .1% imo are not going to do anything since there are much higher returns on DAI all over the ecosystem.

Is this an integration thing - so people can test? Ok set it to the barest minimum required for that - but until I hear from integrators what they need as a rate ‘to test’ I have no clue what that should be set to nor how much I want the system to pay out generally to achieve some real level of testing and integrations. Where is the demand and the data gathering to test this? I just am not seeing how anything short of .1% is useful to anyone for anything at this point. Vault creation and DAI borrowing on a literal tear (almost concerning as DAI borrowing is very concentrated into only a few vaults). I also am not hot to send the wrong market message with what 500+M stuffed into stablecoin vaults with 0 SF and no fee recovery at this point. Did we DC all those vaults to 0 yet as any of those with available DC could be interest arbed to max out DC? I have not checked.

Closed the poll - Will craft the onchain poll PR now, will likely go into the portal next Monday.

Thanks everybody for participating!

1 Like

Poll for what? No change won no?

1 Like

was waiting for that :slight_smile: thanks!

so technically both 0% and 0.01% got >50% of the votes if you deduct “Abstain”. As 0% is the current state, we can just go with a binary vote here.

2 Likes

on-chain poll is live und running until 2021-02-18T16:00:00Z - please participate!

2 Likes

thanks everybody for participating in this thread, voting in the poll and helping on the exec. the exec just passed, so we are going to have a baby-DSR again starting in ~48h.

Update: spell has been cast

1 Like

DSR is dead.

1 Like