I dont necessarily disagree, but what do you specifically see as the root problems?
It’s so premature to say this. In practice, we haven’t even seen the current design yet. Almost all of the Dai comes from ETH currently. The whole design was labelled Multi-Collateral and we have barely seen that in action.
Before we consider structural changes we at least need to see how far we can get with a mix of uncorrelated collateral types.
What about MCO as a collateral type? https://hitbtc.com/MCO-to-ETH
It seems relatively steady in price, * ##### Market Cap
I was thinking about MCO this morning. They provide a debit card where you stake this MCO to get a better deal on the card: crypto.com.
I was thinking that they might also see value in using MCD to offer a line-of-credit against MCO to back their debit accounts. I think MCO is worth investigating as collateral.
That being said, a quick glance at the token tells me there are a lot of undesirable interfaces that would need a very thorough review before we could sign-off on a collateral adapter.
Yeah, I have one of those debit cards. @cmooney Anything I can do to help with this?
Hey guys, based off the discussion in this thread and in the rocket chat over the weekend, I think it’s time to start polling for ideas on how to proceed forward and narrow our options. From what I gathered, these were the most discussed options that are feasible to be added in an emergency executive vote for later this week. Other viable options (such as Uniswap tokens or PaxGold) have some technical restrictions that make them hard to implement on short notice. As a final caveat, while these are individual polls, it is unclear if they would all be implemented at the same time (it’s hard to do conditional polling). Which of these are actually implemented should be discussed in this thread and the governance call.
- Are you in favor of adding LINK as a collateral type to MCD?
- Are you in favor of adding ETH with a lower liquidation ratio as a collateral type?
- Are you in favor of adding additional fiat-backed stablecoins (such as PaxUSD) as a collateral type
- Are you in favor of significantly reducing the stability fee and liquidation ratio for USDC?
I’d love to see two vault types for each stablecoin. One to help hold the peg and another with a higher stability fee for emergency liquidity. Bonus points if we can highlight this in the UI with snarky language.
@cyrus, can you provide some discussion on why this would be a good idea?
Thanks for your open mindedness. I see these as the root DAI problems that the current contracts cannot solve:
PROBLEM #1: No mechanism to deal with DAI above the peg and 0% SF
PROBLEM #1 DESC: The current design cannot get DAI on the peg so we add new collateral instead. Adding collateral is not meant as a measure to control the peg - we should be able to add collateral regardless of the peg and at our own pace. We can have a situation that there is no arbitrage opportunity for borrowers of USDC/any coin with 0% SF - what then?
- No separation of DSR and DAI + negative SF AND/OR
- Undercollateralized loans / giving up on promise to redeem 1 DAI for 1 USD temporarily AND/OR
- Incentives for borrowers when there is no arbitrage opportunity and DAI is trading above the peg
I’m not sure what the details of the solutions for 2 and 3 would be but I guess there are some solutions out there that other teams/blockchains are exploring
PROBLEM #2: Trust in the collateral liquidation process (flip auctions)
PROBLEM #2 DESC: People are afraid of borrowing because they fear that they wouldn’t get a “fair” price for their collateral in liquidation
PROBLEM #2 SOLUTION: The liquidation procedure should be extended to support DEX-es and include some notifications/delays… People must feel that the devs have done everything they could that they get the best price for their collateral - even if that meant a small possible loss/risk for the system as a whole. Maybe there is a way that the contract can decide when it is a time to ignore some large 1-day drops (delay selling) and when it is not.
PROBLEM #3: Trust in the governance system/decisions
PROBLEM #3 DESC: There is a known governance attack problem that needs to be solved thoroughly and I think we are on the right way to do it (dark fix). Besides that, we should solve the situation with bundling governance decisions and with SF decisions that sometimes make no sense. Some system parameters are chosen by the “risk team” instead of chosen by the MKR holders with a recommendation by the risk team.
PROBLEM #3 SOLUTION: Minimize governance decisions about the SF/spread and make them (semi-)automatic. Separate governance decisions that do not depend on one another. Let MKR holders vote for all parameter values and let the risk team just set the range for the poll and the recommendation.
PROBLEM #4: Transparency of the Maker Foundation
PROBLEM #4 DESC: I don’t think this problem is as important as the other ones and it is not connected to the DAI design, but it would be nice to have a roadmap with milestones, semi-annual financial reports, development progress, plans for their MKR/DAI holdings, jobs…
PROBLEM #4 SOLUTION: a roadmap, a semi-annual financial reports etc.
Lowering the LR on USDC has a hazard when we finally turn on liquidations. We need a LR buffer so people don’t get immediately liquidated when we do that.
I also want the SF on USDC to match the ETH one and .5% seems reasonable given market conditions. If everyone wants to go to 0 on SF across the board. I don’t think it will do much except take the small amount of revenue the protocol makes and take it away for probably very little PEG effect.
Rate isn’t the issue here (not between 0 and .5% on SF) the issue is DAI liquidity which if we lower the USDC to .5% should come in pretty quickly. (or at least we can see if it does).
I have made my statements regarding collateral type additions. I will be surprised if people use them to mint anything more than $1M DAI with exception of LINK and this isn’t enough liquidity for a hell of a lot of work. Get the USDC OSM and liquidations etc going normally on USDC and then we can talk about LINK. PAXG I think has too low of a cap to produce any substantial DAI and we are heaping centralization risk on across the board. If we are going to heap on centralization risk lets bag the big collateral dog on the block, BTC.
I also don’t see the addition of UNISWAP ETH-DAI token as a choice above (which I would support) I think this one is a very easy addition. Treat is almost like ETH and given 13% liquidation I really think we should not go much lower than 150% on the LRs on almost anything as after the 13% the vault owner at 120% isn’t going to have any collateral left imo. (LR-1*LF)/LR = 24.67% on ETH currently. Lower this to 120 with LF at 13% and this becomes 5.8%. Really think about your clients here and implications to them by making above types of changes… Personally I want to see LR go up a bit on ETH to maybe 180% and the LF go down to maybe 10% to give theoretical 45% collateral return on ETH vs. the 24.67% currently. This potentially is huge for our vault owners when it comes to liquidations.
We’ll need to do some research on how to best create a Uniswap Oracle. I’d love to see ETH/DAI right away but relying on on-chain data could potentially create issues if the pool is unbalanced via a trade. This is similar to what happened with bZx a couple of months ago, so we’ll probably need to work with Nik to get an Oracle stood up that medianizes the pool balances over time. It’s not anything insurmountable, but it will take a little engineering to do it right.
One thing that concerns me is the push to keep dropping the USDC SF. USDC is far riskier to the system than ETH, especially in light of the anti-stablecoin political climate, and the SF should reflect that. The system is vulnerable to state pressure to shut the adapter down, and I can’t iterate that enough. USDC was added as a way to provide liquidity to keepers in a crunch, and it served that purpose. Stablecoins are not the long-term solution to restoring/maintaining the peg, they are not speculative instruments and so they create almost immediate repayment pressure as soon as they are issued. More (uncorrelated and speculative) collateral types are the answer here, but we need to be aware of the inherent risks of the collateral types that we do have. I’m worried we’re tossing levers just because we have them, without considering the long-term effects.
I think Maker does have a trust issue among Vault participants right now, and it’s going to take more than new collateral types and lower fees to restore that trust and confidence. We’re also in the middle of a global financial crisis that’s creating a lot of uncertainty in the markets. It’s my personal position that the platform is sound as designed, but rapid changes made lightly will jeopardize that stability. Multi-Collateral Dai was designed to have more collateral types, so I think it’s inevitable that we’ll add high-cap tokens like LINK. We just need to make sure we add them with the right risk parameters and that governance adequately manages those risks in the future.
I agree that Maker can and will add collateral types (MCD was all about that right?!). The idea that this has priority over improving system mechanics and other important topics to my mind sends the following message to the community “well we did a few things to the system, we hope will work and stop other problems in the future” along with another message “don’t get liquidated you could lose it all” and oh btw “we are onboarding PAXG, LINK, and whatever else” just doesn’t seem prudent when one is looking at getting a bottom line here of profits needed to fund the DAO in the 5-10M minimum/yr by system DAI growth to approx $.5-1B DAI and a stable thriving community.
I completely agree with your “I’m worried we’re tossing levers just because we have them, without considering the long-term effects.” I will add I am worried we are basically leaving open huge holes in this system exposing not just DAI and vault holders, but the entire DeFI ecosystem and Ethereum itself to a financial risk event that literally lays waste to the entire ecosystem. Everyone will throw up their hands and start blaming everyone else. Personally I find the risk of such an event growing dramatically and steadily reducing my own exposure in this space to get out of the way.
I also agree with you that I wonder if people really are taking into account the macroeconomic picture here. I mean even if Maker could command 1% of all crypto capital this amounts to no more than 2B collateral at this time. At 300% LR this means maybe $.6B. We were pushing $140M just on ETH on perhaps 2% of all ETH before getting smacked down 40-50M DAI on approx 190K ETH and are at 1.66%.
I think I am just going to leave my whole - lets improve the system first - before jamming it full of other collateral and trying to grow it like mad by 10’s if not 100’s of collateral additions aside. My impression is many folks subscribe to a view that somehow magically adding collateral is going to fix any of the other larger issues here and nothing else really needs to be done other than that. My personal view is that if we fix what people percieve are the issues with the system currently we can grow much more with ETH, LINK and other correlated assets because the system will be ready and made more robust against these correlated asset moves… It is also my view that IF Maker as one of the most important DeFI participants can help provide the confidence in the crypto DeFI space that it could lead to a massive market cap growth phase in Ethereum based tokens as well as other coins creating the environment to safely and sustainably grow and support these protocols. Literally to get to where Maker wants to be we need a crypto market cap 5x to $1T and if Maker can be the leader in this - we will have 5-10% of all ETH ($5 -10B value at $100B ETH cap and $1T total cap) and god knows what ever other coins deposited in our most highly valued system for a most highly trusted and valued DAI stablecoin.
Sure IMHO, sure this is opinion, maybe or maybe not an expert in anything (sure people can toss that out if they want opinions vary). My real concern here is that instead of good solid steady growth of the entire crypo-economy people are going to push for the get it quick and drive this entire ecosystem and its participants into one or more thursdays that will be even blacker than 3/12-3/13 and last a hell of a lot longer. As I like to say. I will be happy to be wrong on this as this is how I usually learn something new and fortunately i am still learning. Unfortunately I keep learning over and over that usually when I think something is bad it usually is worse and not better.
Echoing some of the more recent posts here – from a long-term perspective, the current deviation from the peg is unfortunate and warrants attention, but the solution is not an emergency action.
I would urge us to fight the instinct to react too quickly and risk overreacting or compromising on fundamental principles. While it often seems that things move very fast in crypto, highly liquid, trusted, stable monetary systems are built on the scale of decades and centuries.
We are in the midst of what is likely a once-in-a-century economic event. Even the most stable monetary systems out there are being challenged. Only a handful of weeks have passed since Black Thursday – an event which posed unprecedented challenges to the core liquidation systems that are the key backstop of DAI. In such an environment of a sharply increased perception of risk, are we surprised that we are seeing challenges in DAI liquidity?
Over-focusing on the competition between DAI and USDC (or other centralized stablecoin) is not productive. From both the decentralization and liquidity perspectives, this is apples and oranges. If there are projects out there that are considering USDC over DAI for liquidity purposes, what they are really saying is that they are choosing to give up the decentralization benefits of DAI for the liquidity benefits of USDC. Fundamentally, DAI’s unique value prop is its value being backed by decentralized assets. With the thus far limited value captured by decentralized assets (compared to fiat systems), it should not surprise anyone that scaling liquidity continues to be a challenge. In the foreseeable future, both systems will continue to exist, and applications that require liquidity more than decentralization will likely increasingly pick centralized stablecoins.
IMHO, the path forward is to continue to focus on the fundamentals of building a decentralized stablecoin system, and to address the core problems illuminated by recent events. I don’t claim to have any answers, but here are some fundamental questions that should be the discussion going forward:
- What are the key improvements to the liquidation system that would prevent the kind of failure modes we’ve seen recently?
- Is there a mechanism for the system to provide a DAI liquidity backstop in extreme scenarios? How can the system more predictably handle sudden spikes in demand for DAI?
- Are there key metrics that could illuminate the overall nature of liquidity in the system, that would allow better modeling of risks around cascading liquidations on inability to acquire DAI to close CDPs?
- How does the system grow liquidity when being backed by risk-on assets in a generally de-risking macro environment?
We are very much in the early days for DAI, and several weeks of lessened confidence in the system will be but blinks of an eye when put in the context of the many years it will take to scale liquidity to billions. Attempting to add incremental liquidity by quickly onboarding potentially problematic new collateral types is, at best, kicking the can down the road on fundamental issues, and at worst, adding new types of risk into a system. The more quickly we can start addressing fundamental risks, the faster the system can restore and grow trust, and eventually liquidity.
Yeah I think your about right. Thanks for typing it out, not that I thought you didn’t have specific points to back it up, I just was curious.
Trust Trust Trust, so important. Its what alot of this boils down to.
One thing that concerns me is the push to keep dropping the USDC SF. USDC is far riskier to the system than ETH, especially in light of the anti-stablecoin political climate, and the SF should reflect that. The system is vulnerable to state pressure to shut the adapter down, and I can’t iterate that enough.
In this scenario, there would be plenty of advance warning for people to wind down their CDPs or have them liquidated before the adapter was shut down. It would require a court order or new law to make this happen- not something that would occur overnight. A graceful wind down would be the only way to protect innocent users. So the main risk here is people needing to buy back the dai and having to find other sources of liquidity to keep the Dai price from going up again.
USDC was added as a way to provide liquidity to keepers in a crunch, and it served that purpose. Stablecoins are not the long-term solution to restoring/maintaining the peg
I don’t agree with this. Stablecoins are the only collateral that’s not correlated with ETH right now. USDC provides a great way to arb the peg (i.e. mint dai, sell for usdc, buy back dai when peg is reached) - its the only way to do it without worrying about the crypto markets. I could see easily getting another 10m DAI minted for this reason. Agree that we have to maintain a reasonable ceiling, but I don’t see risks to MKR holders higher than with ETH collateral in terms of black swan events that could result in a major loss of the collateral value.
Hi small MKR holder, first time poster. General comment on collateral options from perspective of stability.
As well as uncorrelated collateral it’s also important for collateral to have low volatility. The issue in crash across markets generally was correlations went to 1 AND volatility spiked (and has stayed high).
I’d push back on notion LINK is uncorrelated to ETH. i haven’t modelled but looks positively correlated to ETH to me, just very high volatility.
I think PAXG would be very useful collateral for stability as while gold does have higher volatility than usual right now it’s not like crypto levels of volatility.
how about unorthodox approach. Correct me if I’m wrong, but from technical standpoint it is possible to vote in by governance minting unbacked DAI.
How about minting let say 1 MLN DAI selling it off for MKR effectively changing amount of collateral behind each DAI to ~0.988 USD
(in case of Emegency shutdown)
and see if that will drive price down. This is a method commonly used by central banks (SNB defending EUR/CHF rate at 1.2 for years being one notable example)
maybe the real problem is in the fact, that
Amount Basking each DAI + Utility of holding DAI instead of collateral == Price of DAI
Amount Basking each DAI = Price of DAI
and You cannot have those two equal under all circumstances.
Sure expanding collateral base should to some extend push away a problem, but I guess problem itself is much more fundamental.
I am definitely for minting unbacked DAI. I’m not sure what you would do with MKR and would the buying drive the price up.
If we assume that ETH will be i.e. $280 at some point, we could mint some DAI, buy ETH and deposit as collateral and then restore backing by selling ETH for $280 in a few months or whenever.
I don’t think $1M will be enough - maybe even $10M won’t be enough because the demand will rise again and borrowers are scared of black swan events. The minting should be followed by improving the liquidation process.
Would this work:
- Mint 1M DAI with no collateral
- sell all DAI for ~5555 ETH (1 ETH=$180)
- deposit all ETH as collateral (DAI is now 100% collateralized)
scenario 1: ETH goes to $90: sell ETH and end up with 500k system debt
scenario 2: ETH between $90 and $180: do nothing (the system is undercollateralized)
scenario 3: ETH between $180 and $234: do nothing (the system is overcollateralized)
scenario 4: ETH reaches $234:
- sell all ETH for 1.3M DAI (realize 30% profit)
- burn 1M DAI, 300.000 goes to system surplus
What You describe is quite interesting variation of what I suggested
My reasoning was that since MKR holders takes the risk they should also benefit from the operation (that is DAI used to buy MKR)
Your suggestion is quite different. What you are suggesting is for MKR governance to be market makers.
Generate DAI and sell it for ETH and put that eth as deposit to the CDP which is governed by MKR governance not by external actor. That is quite interesting proposal I have to say!
One thing is that here MKR holders take additional risk of bad debt defending the peg and not benefiting from it, so there is incentives misalignment.
In what I suggest there are both sides. carrot and a stick. I guess You can even build external system where every MKR holder can choose to lock MKR or not (let say for 12 months) and participate in both profits and possible loses (if DAI drops below the peg ) sure there is some complexity in finding when DAI needs to be buyed back and burned, but I guess it can be solved.