[Signal Request] Adjust ETH-B line (IAM-Ceiling)

ultraschuppi · January 14, 2021, 2:32pm

Overview

Since IAM was activated for ETH-B on Monday, the DC has almost doubled in the first four days (from 10 MM on Monday to to 18.5 MM at 2020-12-18). Looks like the appetite for people minting day with a lower CR but a higher SF seems quite high.

We do have a lot of headroom since the line for ETH-B is at 50 MM, but next onchain poll is 3.5 weeks away - so why not start preparing for that Next onchain-poll will likely happen on January 11th and a lot of people might be half-active/in-active in the next weeks, so I thought it makes sense to start the signal right now

Pros

more fees collected
more DAI in circulation

Cons:

increased exposure to ETH volatility

`line` adjustment

Here are a few polling options with +50%, +100%, +200%. Please note that this is the IAM-DC, so if you are not familiar with the Instant Access Module - here is the MIP.

150 MM (+200%)
100 MM (+100%)
75 MM (+50%)
50 MM (no change)
Abstain

0 voters

Next Steps

Poll will run until January ~~7th~~ 14th and depending on the result will move on-chain assuming the outcome of the poll deems it necessary

Update

As expected raising the ceiling might have significant impact on the exposed risk and therefor potentially on the SF as well. Keeping the poll open for another week. Please have a look at the discussion, especially at @Primoz comment.

ultraschuppi · December 18, 2020, 9:02pm

I will also create a signal request tomorrow (?) about adding a new WBTC-Vaulttype with IAM with a lower LR and a higher SF - i think ETH-B proves that people are totally willing to pay more fees for getting a better LR and we should support this as it helps to bring more DAI on the market.

Jiecut · December 18, 2020, 9:00pm

Not sure if we should be increasing ETH-B exposure with the current liquidation system.

Will be interesting to see the Risk teams take.

ultraschuppi · December 18, 2020, 9:01pm

also keep in mind that IAM makes sure that the DC is not maxed out directly, as the DC gets upped only every 12h by 5 MM (at least with the current configuration):

Maximum Debt Ceiling (line): 50 million DAI
Target Available Debt (gap): 5 million DAI
Ceiling Increase Cooldown (ttl): 12 hours

ultraschuppi · December 18, 2020, 9:08pm

to be on the safe side I would not touch the gap and the ttl but apart from that I am not feeling uncomfortable with raising the line

Aaron_Bartsch · December 19, 2020, 12:15am

I think @Primoz will have to weigh in on this because I think we’ll have to increase the SF if we push the line past 50M.

hongbiao_li · December 19, 2020, 2:13am

If the debt ceiling of ETH-B exceeds 50 million, we must increase SF accordingly, which is necessary.

Primoz · December 19, 2020, 7:39am

Max Debt Ceiling that were based on thresholds where Risk Premium start reaching double digit values, in case of ETH-B around 14% at 50m. At current debt exposure Risk Premium would be around 6%. Rates Group did propose SF increase to 5% for this reason last week and this should be implemented when executive passes.

We’ll monitor situation and react at next possible executive. We are noticing though that ETH-B users are a bit more risk taking versus ETH-A (apart from lower collateralization levels used, they also seem to be using less collateral buffer to protect against getting liquidated). Since the box parameter is only 15m, this might be becoming an issue. Potentially we should also revaluate if box of 15m is still appropriate. This limiting factor is probably one of the biggest hurdles right now that makes me a bit uncomfortable against increased portfolio risk exposures despite higher compensation from fees.

mario · December 21, 2020, 5:39pm

The problem with increasing eth b vault type to double digit interest rates is that we are punishing early adopters for new customers and the end result could ultimately lead to a loss of revenue. I’d prefer to cap eth b vault to a debt ceiling that matches market rates and create a new vault type for further rate increases. I understand however this implies more work for SC domain teams and I’m not entirely sure how a new vault type affects auctions efficiency/liquidity in case of liquidation

ultraschuppi · December 21, 2020, 5:53pm

How does this reduce our risk w.r.t. to liquidations? If Eth vaults with a fixed low LR gets liquidated it does not matter if it is eth-B or eth-c

mario · December 21, 2020, 6:02pm

I mean whatever the market rate is, determine the cap on the DC.

Further increases that imply higher rates to cover risk to be integrated through a new vault type with a different rate and so forth.

That way you keep the customers and can adjust your premium accordingly through the new vaults.

ultraschuppi · December 28, 2020, 3:49pm

update: we are now at 39MM dai-from-eth-b, if we stay with the current velocity we are going to hit the DC at the end of this week

ultraschuppi · January 3, 2021, 6:43am

update: we lerped up to the upper boundary of 50 MM ceiling, 45MM DAI minted. have been a little optimistic with my prediction, but lets see what happens today

LongForWisdom · January 4, 2021, 12:00am

I’m not seeing a lot of consensus in this poll so far. Please remember that you can vote for multiple options.

I also feel like this is one of the situations where voters should consider modifying their votes based on the input from the Risk team. It’s naturally hard to tell if this is happening, but yeah, worth consideration if you haven’t thought about your vote since you made it.

Primoz · January 6, 2021, 8:53pm

I don’t particularly like increasing ETH-B DC at this stage to be honest.

One reason is that until liquidations 2.0 are released we are relying on 15m box parameter which is very low considering current DAI debt exposure and this is particularly dangerous for ETH-B. I am not sure if we have good argument to increase box, but it would be very beneficial knowing that we have more support from keepers and increase it.

Second reason is low Surplus Buffer and the outcome of the Poll where people aren’t necessarily in favour of increasing it. We can’t have Surplus Buffer representing 0.3% of DAI debt and meanwhile aggressively increasing debt exposure of both ETH-A and riskier ETH-B.

Third, are we ok to come into situation where debt exposure to ETH-B increases towards 100m and then realize we need higher SF to compensate for risk? Are we going to be ok by suddenly charging 10% or more for ETH-B? We had similar situation with YFI-A and it didn’t turn out good.

Fourth, I know nobody wants to miss on new DAI mint, but ETH-B users could very well be same users who would use ETH-A otherwise. It is possible Maker is stealing users from Compound and Aave currently with ETH-B, but don’t forget that Maker has OSM and others don’t, so these people might prefer to use Maker even at slightly higher LR. This is only speculation though.

Fifth and lastly, maybe we should just slightly press brakes during this insane growth. But it’s me representing Risk so maybe I am too cautious for some.

alexis · January 6, 2021, 9:23pm

I think you are alright we need to consider that as fast as it went up, it can go down even faster and the ETH-B is not really the type of vault we want to have wildly open during this type of event.

Thanks for being this too cautious guy. Changed my vote.

mario · January 6, 2021, 9:44pm

Agree with this, much preferable, even if it takes more time to create a new vault type with the same LR and higher SF in order to avoid a similar situation as with YFI, difference is ETH holds 75% of our revenue. If it cannot be done I’m fine keeping it at 50m

smaugho · January 6, 2021, 10:03pm

After seeing how the ETH have gone easily from $750 to $1200 in almost no time, the speed of that movement and current market behaviors, I’m starting to worry much more about a big market correction to occur at anytime, I don’t see at this point a 25% or even 30% correction in less than 24h like something rare, that’ll be more than back in March?

Not saying not to adjust the ETH-B line later after proper assessment by the risk team has been done, but in this precise moment, it is a huge risk IMO.

Added to Primoz points above: I changed by vote to no.

Aaron_Bartsch · January 6, 2021, 10:09pm

I think if we are going to do an ETH-B line increase it should be married with a larger SF increase as well as a substantial surplus buffer increase. That should be enough to satisfy all parties I think.

mario · January 6, 2021, 10:12pm

That is as long as the 20% gap in the LR invested exceeds the gap in the rates which with the new parameters it’s 3%.If not people will switch to eth A