IAM DC ETH-A Parameters, 27th Jan 2021

DAI debt of ETH-A vault type continues to increase and it is becoming hard for governance to react fast enough so that the debt ceiling doesn’t get not fully utilized. We believe Debt Ceiling IAM is needed and preferred for ETH-A vault type.

Given that there are no other contents for the executive proposal this week and that there is uncertainty over the demand for the ETH-A vault-type in the coming weeks, this change will be put up as an executive vote on Friday.

Parameters Proposed

line

We have a pending signal request on potential DC increase of ETH-A and currently the 1.5bn debt ceiling is winning with about 55% approval. As @Risk-Core-Unit already mentioned in forums and on Governance & Risk call, the ETH exposure is starting to become an issue for overall portfolio risk as VaR levels don’t match Maker’s protection at hand currently (low surplus buffer, low box, liquidation 2.0 not released yet,…). On the other hand there are valid concerns about integration risks of ETH-A if the debt ceiling gets fully utilized. Also it seems the community supports higher LR ETH-C proposal, but I am not sure if we got a clear signal whether potential ETH-C implementation also assumes limiting ETH-A debt exposure.

Having all this in mind and trying to satisfy community preference we can continue to increase ETH-A Debt ceiling but then mitigate risks differently, with higher SF. Here we want to make sure that people are aware that stronger measures may need to take place to counter the increased risk from ETH-A exposure. This includes, but is not limited to: increasing the ETH-A SF at a more rapid pace, increasing the surplus buffer and increasing the box. This is because the aim of higher SF would not only be to protect Maker, but also to lower the pace of ETH-A debt exposure increase. Note that higher SF doesn’t offer short term protection of increased debt exposure.

We propose line parameter for DC IAM ETH-A to be set to 1.5bn, matching current community expectations and also high enough to make sense using DC IAM.

gap

As already mentioned in our DC IAM analysis, we either want a combination of low ‘gap’ and low ‘ttl’ or combination of high gap and high ttl. You don’t want to be making low DC adjustments too rarely and you don’t want to be making high DC adjustments too frequently.

We believe a higher gap is preferred for ETH-A. We saw how whale Vault users can mint 20m DAI or more instantly and we don’t want to prevent this since Maker is uniquely positioned in DeFi lending space to allow such mints without affecting the rate instantly.

Downside of higher gap value is OSM risk, but since OSM risks are much lower for ETH-A vaults, we don’t think this could be a huge issue. Also we may want to make sure that enough unutilized debt ceiling can be available in case of DAI liquidity crunch when some of the keepers or market makers still prefer to use ETH-A vault to meet market demand for DAI.

Another reason for higher ‘gap’ is to prevent griefing attacks that we found for this IAM DC, but is less likely for someone to perform it since there are no benefits to it. Still, higher gap limits this attack severely.

We propose a 30m gap value for DC IAM ETH-A Vault Type.

ttl

Since we have proposed a bit higher gap value we want the ttl parameter to match certain conditions. If we assume governance can react and prevent further minting in about 48 hours (current GSM delay), we may want to set ttl to match a number that prevents too high accumulation of debt on ETH-A before governance can react and stop minting.

We believe ttl of 12 hours is a suitable number since it implies 120m max additional mints (48 GSM delay / 12h ttl x 30m gap) until governance can react and change DC. Further, this also enables 60m daily DAI mints on ETH-A, something we found to be quite common lately.

Proposed Parameters for DC IAM ETH-A Vault Type

line = 1.500.000.000 DAI

gap = 30.000.000 DAI

ttl = 12 hours

19 Likes

Seems pretty reasonable to me, and will cut down on a lot of the “human capital” involved with maintaining availability for ETH collateral.

One question in regards to this:

If the ETH-A Debt ceiling is raised all to 1.5M, does Risk anticipate a need for a higher SF?

I think the other options for adding security to the platform are preferable from a vault holder standpoint, but raising the SF does have a nice impact on revenue. I basically just want some clarification here as to the link between supporting this DC IAM and the result on SFs for ETH-A.

2 Likes

So if you are asking about debt exposure increase, definitely yes, unless we improve risk profile by launching Liquidations 2.0, increase box, have much higher surplus buffer, etc. All seems less likely short term.

1 Like

Lol sorry about the typo there.

Makes sense, thanks for the clarification!

@Primoz —going to try to push back here—try, but I won’t have any luck :slight_smile:

What about the analysis that perhaps the overall markets are currently trading with:

  1. Wounded economies
  2. More borrowing to throw into the Mkt = “how could I lose???” attitude.
  3. Less Doubters, or NO doubters at all. Seems like the believe is that the Fed + numbers only go up and up
  4. Seems like nobody thinks Fundamentals count.
  5. The burst of Euphoria? Sure, Bubbles don’t pop right away–but they always POP in months and not Years.
  6. Realistically? Can the Euphoria continue with the exceptional opportunities we have seen in the last 9 months?
  7. Unprecedented Fiscal spending with a Fed who is ready to bend the knee to the Biden Administration. Hey, it’s not the People’s money–it’s the Federal Reserves :slight_smile:
  8. History books enjoy moments like these… fewer people working, less productivity, yet Markets are higher. Does that feel right? :thinking:
1 Like

really would like to have used IAM on more collateral types before the “main” one. I also “get” that we may need to do so regardless.

Is anyone else a little uneasy about raising ETH-A DC by 50% without additional adjustments (higher SF, increase box/surplus buffer)? ETH-A is up 70%+ YTD which is more growth than most businesses have in a year. I know we just increased the SB to 10M, but that post was made when the ETH-A exposure was at 466M and now we’re at 867M, nearly double! If we raise the DC to 1.5B we could rapidly see ourselves in an untenable position if the markets turned and the SB yet to be filled. We should be working to avoid a BT situation where MKR has to be issued at dirt cheap prices, especially while we wait for liquidation 2.0 to be implemented.

I agree with @ElProgreso 's concerns regarding the overall market and would encourage @Risk-Core-Unit to push proposals to mitigate the risk of the higher DC. I didn’t see a single post in the SR to raise the DC supporting a 500M increase which is unfortunate given that option is currently leading the polls at 39%.

I know we want to stay competitive with current market rates, but the rapid growth implies our SF is too low, I would rather increase SF to temper growth until risk is more comfortable with our position. I fully support @Primoz suggestion to move faster with ETH-A SF increases and the other risk mitigation measures. Moving quickly is an advantage we should leverage and will allow us to better manage the increased exposure.

3 Likes

Why not couple this with a higher ETH-A Stability fee?
That would help fill the Surplus Buffer faster and potentially slow down Dai growth, both of which would help reduce system risk.

ETH-A stability fee is low vs. other sources of Dai, by several percentage points.

1 Like

I am very happy to see the arrival of DC IAM. After we automatically increase the DC, how to increase SF IAM accordingly? Is there any progress in this regard?

I think we should increase the surplus and increase box to decrease risk, before raising SF above what is competitive in lending markets. Maker has to remain #1 and slowing down loan growth during the defi land grab is a mistake.

Even purchasing secondary market insurance is preferable to overshooting with the SF. I don’t want a repeat of the YFI-A scenario, this time with our biggest by far collateral.

3 Likes

the main question here is if the risk materially changed… if not, then why change the fee? there are other tools in the toolbox we can use when needed…

The above is the proper reaction and concerns. Having vaults with different risk parameters and SFs also important too.

I am only going to support this if it goes with the following:

  1. Increased ETH-A SF - closer to 4%
  2. Increased surplus buffer - we got 10M now. I want to see this closer to 2% of float (now 30M) at a minimum.
  3. Work on building a reserve/treasury minimum 2% float (30M)
  4. Something like @SebVentures has done using real vault data and running them against real world price drop scenarios to model loss profiles and get a real risk analysis going on not just this vault but the other large and building vaults (wBTC in particular).

My concern here is with PSM governance is getting stuck into this: “we need to constantly raise DC to accommodate demand” and for some reason we are shying away from increasing SFs, building multiple vault tiers (with difference CR, SF, DC profiles) on particularly ETH but generally with a focus on our largest and growing vaults.

My vote may not mean much but I may start tapping on the breaks here at least with my MKR rejecting increasing DCs and reiterating using the other tools (SF in particular, and building surplus, reserve/treasury) as a better reaction to managing demand than just letting DC run and hoping to hell we don’t get trapped into a bad situation before we can bring on-line.

Liq2.0
information regarding keeper status/state
DAI liquidity sourcing (flash mint may work but will all of this work with gas at 2000gwei and above?)

I am happy we have demand. I think it is a great problem to have. But we need to stop constantly and blindly raising DC and mostly ignoring SF as a demand management tool. It took almost 15-20% in the SCD SF to finally tamp down demand which also coincided with a ETH price downturn in 2019 if I remember right. We already are stuck between a rock and a hard place causing governance to put into place stablecoin vaults with huge DCs and LRs of 101 and finally the PSM to get the PEG under control.

I have maintained a thesis since yield farming came on-line that once we could provide liquidity we literally needed to RAISE SF rates to tamp down demand (or get it under control). At the time this was rejected because the PEG was running 1.02 to 1.04. Now with PEG under control we can finally address demand with SF. We need to be slow, but steady and consistent with approach given the completely new PSM situation.

My biggest concern with Maker is that it has yet to seek to achieve any kind of steady state. DAI either grows or declines. I have never seen a real stable DAI demand situation. But these markets are constantly changing and this requires US to constantly change as well. We now have 1.5B outstanding, collateral what 2-3B under management we need to have some courage to do the right things both for the DAI and DeFI economy and develop the tools to monitor economic changes so we can respond more quickly and more accurately. My own research indicates that Maker generally lags by at least 1-2 months what markets need, and often also tends to overshoot on each side (both with keeping rates too low for too long, and raising them too high for too long). This was predominantly why I wanted some dynamism with rates relative to facility usage so the system could instantly and dynamically respond to market demands within bounds set by governance.

5 Likes

I support the proposed parameters for the ETH-A DC IAM but I share the same concerns as others about our overall risk exposure on ETH-A. Notwithstanding the extra smart contract work required to activate an ETH-C vault-type, in the ideal world here is what I would want to see (not in any particular order but hopefully in close succession):

  1. Set ETH-A DC-IAM to 1.5B with proposed parameters.
  2. Increase ETH-A SF as needed to keep total ETH-A exposure somewhat below 1.5B.
  3. Activate ETH-C with higher CR and SF 1-2% lower than ETH-A.

As demand for ETH-A increases I think more risk averse Vault-holders (like myself) would be happy to migrate to an ETH-C vault type for a lower rate.

1 Like

Could we have a higher line? 2-3 billion ?

3 Likes

I support the parameters as proposed without raising the SF. We have to stay competitive and build a large user base as 2 new serious stablecoin projects are expected to go live in the next month or two (RAI, FEI).

1 Like

Why not lock ETH-A at current DC and start a new ETH-C with higher SF?

2 Likes

“Deficits are skyrocketing due to COVID-19. Average yield to maturity on TLT (20 Plus Year Treasury Bond ETF) is just around 1.7% at a time where inflation is running above 3% (represented here by PCE excluding food and energy). It is likely that the US economy is in a stagflation now (economic activity is subdued while simultaneously prices are rising).” --Tomasz Zaton via Seeking Alpha

:man_shrugging:t4:

“Pfizer CEO: High Possibility Future Variant Will Elude Vaccines.” –Delta One

:gorilla:

2 Likes

We are living in QE to infinity. The money printing is not going to stop. We need to make access to dai easier not more difficult. Ether should basically have free reign so that more dai is created and we grab as much land as possible while we have no competition. Lower the ether security fee to 1% and throw everything in the buffer. We dont need to burn now we can burn later. the roof should go to 3-5 billion not 1.5 billion. No more playing baby games — for the spice must flow!

4 Likes

This topic was automatically closed 91 days after the last reply. New replies are no longer allowed.