Improving the Monthly Governance Cycle
- LongForWisdom: Thanks, that was very interesting. Looking forward to being able to get rid of bundling the unrelated changes together in the voting cycle. One thing I would like to mention is that this new Chief gives us the opportunity of modifying how the monthly governance cycle works to make it more efficient. That’s another possible benefit of this.
MANA and LEND Smart Contracts Evaluation
- This is the Smart Contracts technical presentation on MANA and LEND, and I do have pretty good news.
MANA is an ERC20 token.
It follows standard ERC20 patterns.
It’s not upgradeable, so we would be able to use it with the
GemJoin adaptor with this, which is the standard adaptor we use for BAT.
There’s a mint function, which is mintable by the
Crowdsale contract. However, the
Crowdsale contract has been self-destructed. So the total supply for MANA is the current max supply.
There was an audit done by OpenZeppelin, and the team has given this one a low-risk assessment.
MANA - Collateral Onboarding Risk Evaluation - Forum Thread
It’s also a very low risk, basic, ERC20 token.
It can also use the
The only non-standard feature we found in this one is that the decimals return a
uint256 instead of a
uint8. Which doesn’t really affect our adaptors.
This does have mint functions as well. However, there’s a state machine built into the contract such that the coin is only mintable in the presale or ICO state, and we can’t go back from there.
There’s also an audit done by Victor Mezrin, and that was approved for MainNet.
LEND - Collateral Onboarding Risk Evaluation - Forum Thread
MCD System Stats
DAI 24hr VWAP Graphs
Maker Vault Stats
The State of the Peg
- A few significant mints over the last couple weeks, where it ballooned from 114 million to 154 million.
- USDC-A roughly 19 million Dai jump issued since July 1st. Primarily one to two large users. We saw it grow from then.
- After the initial 10 million steady-state utilization, it increased pretty really quickly to 20 million after the DC was increased.
- A significant dip in BAT after the March 12th events, and then another dip on July 1st timeline. As we all know, a few weeks ago, Compound Farming activity began, which impacted the Dai ecosystem in a significant way.
We see a significant amount of Dai being supplied in Compound. Obviously, there’s a significant amount of borrow, but given the nature of the re-wrapping, the biggest net impact here is the additional 85 million Dai sitting as reserves. For other assets such as USDC and ETH, there’s an even larger buffer in the amount that’s supplied versus the amount borrowed.
Given the nature of profitability of the different types of farming behaviors with various projected stabilities and yields, those are being used as collateral effectively for other borrowing behaviors on Compound.
That margin of the amount of Dai being borrowed versus supplied on a percentage basis is much smaller than other assets.
- In the last week or so, about $300 million in nominal borrow of Dai happened on Compound. That doesn’t necessarily reflect the amount of net borrow increase. That’s a distinction that we can touch on. The method of accounting for borrow and supply leads to virtual borrow interest that is effectively achieving a more aggressive
CR. For the amounts of loans outstanding, that’s a large looking number at around 800 million. But given that Dai supply outstanding is 154 million from ETH and another 40 or so million from a couple of other assets, we’re looking at a total of about 190 some million Dai issued there, and 88 million of that Dai is in the DSR(because of Compound.)
Net you’re looking at 100 million Dai floating out there, and you’ve got an excess supply in Compound of 85 million Dai.
So that’s a pretty significant chunk of the Dai supplied floating out there that is currently sitting on Compound as supplied assets. That’s something to be aware of, as movements in this pool of Dai could have significant impacts.
- We have seen an increase in positions living at the lower collateralization Ratios, which is more risk to be aware of and how to go about handling and managing that risk.
Even at an ETH price of $170, you’d be looking at 30 million Dai liquidated.
At $150, you’d be looking at a total projected Dai liquidated of about 70 million.
Those are pretty significant amounts, and those are not completely unrealistic potential price drops for ETH.
The ETH price is already falling a bit down to $235-ish. During the last week, the ETH has been falling, and the Dai price has stayed consistently at what a lot of people call an elevated level, around $1.02. In the grand scheme of things, this is a legitimate elevated Dai price.
Earlier in June, we saw Dai fluctuating between 1.00 and 1.01. Around the beginning of July, we saw the Dai price rise and sit at 1.02, and since then, it’s come down a little bit.
In the past, the peg was mainly driven by ETH price movements. It’s not 100% clear whether ETH price is driving what happens with the Dai peg.
Some of the fluctuations in ETH price would historically produce different effects on what’s going on with the Dai price.
When ETH started to fall from $246-ish to around $239, the Dai price did rise a little bit. But since July 13th or so, as the ETH price was coming down, you did not see a commensurate rise in the Dai price, it actually fell a little bit. That was a bit contrary to what we usually see. It’s important to think about all this external ecosystem around Dai and how it’s impacting supply and demand.
With just the significant amount of increase in supply, particularly in ETH-Dai supply in the last two weeks, it makes sense to consider that there is a portion that’s been supplied on Compound, but also a decent amount that probably made its way onto the market in terms of selling pressure.
All that minting is at a very low cost, so there are very few barriers at the moment for people looking to lever up, mint Dai, sell it, or use USDC vaults, despite cap on DC, for getting free hedge positions. If you look at a vault with USDC, where liquidations are not enabled at the moment, it’s a pretty low-risk option for minting and engaging in multiple different kinds of yield-seeking behavior. COMP farming is just one example, but there are plenty of others.
USDC-A Debt Ceiling Not in Use
- Vishesh: The USDC-A DC was raised. There’s the potential for that minting behavior. When we were looking back on July 1st, there was an immediate utilization of the amount of the available DC. That hasn’t jumped again to cap out the “increased leash.” It’s worth considering to see if the potential for utilization is still there. A few large users drove a huge chunk of that utilization. Whether we expect to see if another large user comes in and uses the remaining DC, how do we want to handle it? Sam suggested lowering the base rate. There’s a conversation to be had there.
Some Risks of USDC and its Concerns
- Vishesh: If you’re providing low-cost routes for minting Dai and getting access to a lower risk capital position and then using it for higher-risk higher-yield activities, there is a question of how low you want to make the rates to incentivize that, versus accounting for the risk that you’re inherently underwriting by facilitating this behavior. We’ve looked at the MKR-ETH collateralization ratio, and most people have nestled on the USDC collateral positions being a lower risk pool. It’s always something to keep aware of, that as the pool grows, the risk becomes different. It’s not necessarily a market risk. Given that liquidations aren’t currently enabled, we don’t see too many people concerned with that USDC collateral pool.
USDC Stability Fee
Vishesh: There are independent threads on how the fees are being managed and when the ceilings are moved. MakerMan is suggesting a more comprehensive conversation on that. I’m not sure what exactly you would do differently. Right now, it’s not been a risk-driven conversation on what the Stability Fees are, but more on the levels at which you incentivize different behaviors. The DCs appear to be moved when you need more slack. It’s not necessarily the wrong way to manage that, just stylistic. I don’t know if you would change the narrative of how to choose to set those DCs. Right now, it seems to be that when you want to incentivize more utilization, you provide more slack and when it’s not being used, you lower the SFs.
Vishesh: That’s the suggestion in the chat about the USDC Stability Fee. That’s a conversation to be had with the community. Not sure if I would be able to comment on that right here or now.
Risk, Peg, and Parameters
- Makerman: I haven’t posted on this. We are taking on risk, though. I don’t have a good vault analysis tool to look at why we have these big numbers of liquidations that have higher risk points. I think it’s just bigger players taking on farming risks and using their ETH to do it. You’d like to raise fees to capture some of these farming activities to mitigate the risk. You do that, and the peg might go up, you put more strain on it. Literally, half of available Dai is stuck in the DSR via Compound. Do you want to take on the risk behavior? I honestly don’t know. I don’t know if anyone knows. But it’s kind of the real question here. If the peg is number one, and three out of four people say peg is a big deal, including Rune, then everything else isn’t as important. I don’t have a good feeling about that, but it has implications for stretching other things. If we’re going to only grow Dai by growing assets, you have to think about how to do that. We have three key players. Unfortunately, we have two more on top of ETH: USDC and WBTC. WBTC wants more. We could raise the rate while we raise the DC and see if they take it again. USDC-A, if we want to help that, we got to come down on rate. Do we want to coordinate and try to get another 50 million to touch this risk and have risk behavior, and when the Compound thing switches, it’s going to switch hard and fast. The big money is going to move out of this, and we will have to step it up. Fortunately, we have a lot of good tools to do that. I just don’t know how we do the coordinated thing. I’d love to hear from Vishesh, Cyrus, Christopher Mooney, Rune, and others, about all the implications on what we do and when.
Consider the Lifecycle of the Dai and Market Fundamentals
- Vishesh: One thing to keep in mind is that increasing supply only impacts the peg when users are willing to mint that dai and sell it on the open market. That’s an important part of the conversation, for the types of supply created are fragmented into multiple use cases. You have to consider what is the end use-case and what is the path to removal from circulation for every marginal Dai that’s created. Is it being created just to lock on Compound, is it being created to be sold for something else? Considering the lifecycle of that Dai is a little bit missing, but it’s important to keep in mind that if the goal is bringing down the peg, you have to target use cases of Dai that will put more liquidity on the open market and help drive that price down. You also have to think if that’s something that will happen at this point in time, if there are other market conditions or smart contract-driven fundamentals that are putting a premium on Dai that makes Dai worth more than 1 USDC. That’s another important part of the conversation that is being under-discussed. For peg management, increasing supply is not 100% equivalent to driving down the peg. It only works when that supply makes it way onto the open market.
Multiple Tiers of ETH to Discover Risk Appetite
Christopher Mooney: Let’s imagine the peg was at 1. We would all be concerned with the amount of risk building up. Just a month ago, we were looking at that histogram where liquidation ratios wherein the 500%-600% range. We had a lot of people that were risk-off because of Black Thursday and the overall macroeconomic sentiment. COMP has incentivized ETH-A collateral holders to go risk on. They’re pushing towards lower bounds where liquidations are extremely risky. And thanks for calling that out, Vishesh. If we were at peg right now, we would probably want to suggest multiple ETH collateral types, where the effective 0% SF would be reserved for something that had a liquidation of maybe 300% or something that pushes people to have a higher CR. Maybe we need multiple tiers at this point for ETH. Naturally, people that want the lower rate that have higher liquidation ratios are going to move into those other tiers, and we start to get a good idea on who’s “risk-on,” and how much would we have to liquidate of those people. Maybe not ideal.
- MakerMan: I agree with that. We’re coming to “more tools better.” The idea of having multiple tiers of the same thing: why do that? It’s to manage risk by setting the liquidation ratio. Also, to understand where the money and risk spread out on a spectrum. The best bottom one will always be taken for the longer haul, but the higher risk stuff, to see where it’s landing, collect some fees, deal with the risk by having liquidation ratios managed, I think it’s all a plus. We should do a MIP on it actually and have it as part of the Protocol. Not just one, but a couple of tiers in each of them. We could pick the ones that are working right now, like ETH, and maybe WBTC, and USDC-A.
Subsidizing USDC-A and the Base Rate
- Sam MacPherson: As Vishesh said, we should maybe consider lowering the base rate to a negative value to encourage more supply coming from USDC-A. It shot up pretty quickly when we were at 0%. And now it’s staying flat. That’s more Dai that we could bring to fill this COMP-driven demand. I think we need to fill it in whatever way we can. We can’t be too concerned about recouping risk premiums because we’re in a really hard place right now, and we need to subsidize it through MKR. I don’t think we have a choice.
Liquidity into COMP
MakerMan: I really want to re-echo what Vishesh said. We’re taking on risk here, so keep in mind how we’re managing this. I like the idea of getting fees out of different tiers to deal with taking on these risks. If these markets will do it, I would rather do it that way. ETH-A as it sits right now, those are our core people. We could add another at 1% or 2% for all the new people that want to take their ETH and cycle in. We have probably missed out on 50 million worth of ETH that would have paid us 1% or 2% to have access to a facility with lower ratios. When people that are in the “good facility” leave, that money will jump there. There’s a real issue with how we manage this. It’s still the same point: You would like to get some fees but anything you try to collect, is it going to affect the peg?
They’re progressively eating up liquidity. If you graph the USDC supply and borrow, it keeps going up slowly. I’m not sure if it’s new money turning up into the bigger accounts. We can add liquidity, but if it gets buried somewhere, it’s not going to do anything to the peg. How do we deal with the peg in the face of COMP farming? You can throw 50 million at this, and the peg doesn’t even flinch until the COMP thing changes. Throw 500 million Dai at it, maybe it changes. I see it as a blessing in disguise, as it makes us think about the size of the market. Everyone wants to grow, and we can determine how we should grow. This demand could come from anywhere. Any time and any amount. Other places could pull the same trick and make Dai the big deal.
Market Making and System Risks
Vishesh: You can continue to increase the supply of an asset, but it’s not going to impact the peg unless users are willing to say, “I’m going to mint, sell and relinquish the Dai above peg, and I’m ok with maintaining the risk and the fees, in exchange for that short-term profit.” Unless you have that market-making behavior, that arbitrage behavior, you’re not doing anything to the peg by increasing supply. That’s a really important point that’s completely overlooked because people in their minds think “more supply, lower price,” but we’re missing the Rube Goldberg in-between that gets you there. Everyone is focused on the peg, and we tend to overlook system-level risk, and I showed earlier that it’s growing. Not just in percentage terms, but also in total dollar terms. How much of that nominal risk can and should MKR be able to underwrite?
- Sam MacPherson: COMP farming will not lower the price of the Dai peg, but it removes upward pressure.
- LFW: Fantastic discussion around the peg and the current issues facing us. I encourage everyone to continue the discussion in the forums. I know it’s cliché, and I say this every week, but that’s where governance happens. I would love to see this discussed in more detail. Specifically, more concrete proposals about what we could or should do.
Vault Compensation Plan
MCD: The Multi-Collateral Dai system
CR: Collateralization Ratio
DC: Debt Ceiling
ES: Emergency Shutdown
SF: Stability Fee
DSR: Dai Savings Rate
MIP: Maker Improvement Proposal
PSM: Peg Stabilization Module
Tim Black produced this summary.
David Utrobin produced this summary.
Artem Gordon produced this summary.
Galla Guillen produced this summary.
Juan Guillen produced this summary.