Decreasing Liquidation Ratios - Proposal

A week ago we published an analysis about the effects of lowering liquidation ratios for various Maker’s vault types. @Risk-Core-Unit decided to skip the Signal Poll on proposed changes, since we didn’t receive any objections in the mentioned post. We therefore propose to include Liquidation Ratio changes into an on-chain poll next week. If there are any last minute objections or good arguments to why we shouldn’t decrease liquidation ratios as specified below, please comment in the thread.

Here are the proposed Liquidation Ratio changes from our side:

VAULT TYPE DAI CURRENT LR PROPOSED LR LR DIFFERENCE
PSM-USDC-A 3,225,986,052.41 100% 100% 0%
ETH-A 1,490,899,135.68 150% 145% -5%
USDC-A 238,816,815.09 101% 101% 0%
WBTC-A 231,166,849.56 150% 145% -5%
ETH-C 198,127,476.74 175% 170% -5%
PAXUSD-A 67,292,624.28 101% 101% 0%
UNIV2DAIUSDC-A 49,990,490.24 102% 102% 0%
ETH-B 36,629,072.17 130% 130% 0%
LINK-A 34,810,469.23 175% 165% -10%
UNIV2DAIETH-A 28,609,753.49 125% 120% -5%
TUSD-A 26,538,594.88 101% 101% 0%
YFI-A 12,364,479.36 175% 165% -10%
UNIV2WBTCETH-A 10,243,791.44 150% 145% -5%
UNIV2UNIETH-A 10,003,111.78 165% 160% -5%
RWA002-A 6,312,771.91 105% 105% 0%
UNIV2USDCETH-A 3,087,401.23 125% 120% -5%
RENBTC-A 2,808,977.55 175% 165% -10%
UNI-A 2,500,840.70 175% 165% -10%
BAT-A 1,915,166.72 150% 150% 0%
AAVE-A 1,599,589.54 175% 165% -10%
MANA-A 1,502,136.98 175% 175% 0%
ZRX-A 1,406,983.16 175% 175% 0%
LRC-A 877,341.85 175% 175% 0%
GUSD-A 352,412.03 101% 101% 0%
UNIV2WBTCDAI-A 218,796.33 125% 120% -5%
BAL-A 171,115.06 175% 165% -10%
COMP-A 72,614.71 175% 165% -10%
UNIV2LINKETH-A 34,502.85 165% 165% 0%
KNC-A 8,856.41 175% 175% 0%
USDT-A 733.13 150% 150% 0%
USDC-B 438.6 120% 120% 0%
RWA001-A 0 100% 100% 0%
UNIV2ETHUSDT-A 0 140% 140% 0%
UNIV2AAVEETH-A 0 165% 165% 0%
UNIV2DAIUSDT-A 0 125% 125% 0%
RWA003-A 0 105% 105% 0%
RWA004-A 0 110% 110% 0%
RWA005-A 0 105% 105% 0%
RWA006-A 0 100% 100% 0%

Side notes:

  • 5% LR decrease for top tier vaults types such as ETH-A, WBTC-A, ETH-C and UNIV2 vaults types
  • 10% LR decrease for second tier 175% LR vaults types
  • No LR change for vault types that are intended to be off-boarded or have DC set to 0 (including pending executive for UNIV2ETHUSD-A and UNIV2DAIUSDT-A). The framework for off-boarding specific collateral is still in discussion and we didn’t want to impose sudden changes, because one of the strategies to off-board these collateral is to actually increase LR linearly and force liquidate vaults if needed.
  • No changes to ETH-B due to OSM delay risks.
  • No changes for UNIV2DAIUSDC-A vault type (LR was determined on yield)
  • No changes to RWA collateral
  • No changes to stablecoin vault types & PSM
15 Likes

Just confirming that this will go on-chain on August 16th. As the PAX-PSM stuff this comes under MIP41c2 - Facilitator Governance Powers.

Facilitators have the power to propose weekly non-standard Governance Polls (defined in MIP16) related to their Core Units.

1 Like

I’m in favor of these changes but am curious if we have a list of LRs of our competitors? Think this would be useful to review in combination with liquidation penalty rates.

1 Like

You can find them in D3M report for Aave - on average they are 20% lower. It unlikely we will able to go that low due to OSM delay at Maker. Yet it is true that penalty fees could be reduced, especially because Liq 2.0 is performing well. Hard to tell about penalty fee reduction effect on more usage and in theory you may want to attract safe users primarily.

1 Like

Looked at the analysis but honestly trying to figure out what this means for the system generally.
From my take basically no more than 50-75M could be reasonably minted. This is to address adding DAI liquidity on over 5B. Drop in the bucket amount of DAI - literally the surplus which I don’t believe is doing much in terms of limiting DAI liquidity. The issue is and has been demand due to farming. We can still see based on a report Maker Vaults Market Activity from @rema that liquidity farming (particularly on compound where all this started and players can just lever up) is still alive and well.

When I look at this from a practical perspective minimum transaction sizes to take advantage of this are going to have to be 10K (minimum and this is really low, a more real number is like 50K) and this means the vault sizes to effectively take advantage of this are going to have to be 200K+ and really more like 1M+ and even then only some percentage will take advantage of this. What people are more likely to do is simply appreciate the lowered LR which is fine from a vault perspective, but not so hot from a systemic risk perspective, particularly against ETH. On the plus side if there is a ETH market downturn all vaults will get more room, and liquidations won’t come on line, but on the negative side the liquidation collateral value buffer is decreased by what 20% doing a 125-120 move, and 16% on a 175-165 LR move.

I personally don’t see this as a wise move for two reasons:

  1. It increases system risk for.
  2. Very little possible increase in DAI outstanding.

I am back to if Maker wants increased DAI liquidity lets get some new RWAs going with the 10-20M DC’s so over a few months as we get more comfortable we can up them. Also I really would encourage Maker to consider looking into competing with farming going on via issuing a MKR-R rewards token and dripping it out to borrowers based on interest paid. Decide whether how to exchange it for MKR, or how whether to participate in governance, and/or how to do a buyback on it. Or even the conditions to reward it (think kpi metrics)

Use this MKR-R to basically spur more liquidity to come and the real result to mint more DAI.

Honestly I am back to a basic set of questions?

  • Is the problem DAI growth or demand?
  • Is the problem DAI liquidity?
  • What is the problem this proposal seeks to address?

Frankly from what I can see the problem is not well stated and hence the proposed solution is unlikely to address the problem.

I believe the core problem is collateral. Maker isn’t getting much except through ETH price appreciation and added USDC but simply tweaking SF’s and LRs to put system in hazard isn’t a way to get even 1B more DAI much less 5-10B. Collateral and a huge chunk needs to be added and Maker right now isn’t in a good position to have that come. Not with current models, politics, etc. Now if someone has an idea here where next 5-10B is coming from that isn’t RWA I really want to hear it.

If I have missed something or are inaccurate above feel free to present arguments or data.

2 Likes

One problem from my perspective, and maybe this could be addressed by a yield farming initiative is the loss of ETH collateral market share to competitors, namely AAVE.

Over the past year ETH locked as collateral has doubled and while MKR remains #1 in amount of ETH locked, we are actually down from this time last year by 11% (in ETH terms). While we started the year off strong, we’ve been bleeding market share ever since AAVE launched their liquidity mining program and are actually down 800K ETH in ~3 months. This should concern us all even if these yield farming plans are unsustainable.

What about AAVE? Up nearly 12x and accelerating since liquidity mining was implemented.

Compound? Up 50%.

3 Likes

This is my main concern, too. I feel as if we’re putting all our eggs in the RWA basket, but not fighting hard enough for the $300B in ETH collateral that is still out there.

Over the next few months, some big chunk of that $300B is going to be moving to staking services; right now is the land grab phase. Our stablecoin competitors have offered staking services (Coinbase, Binance) for months and could decide to lend against deposits any time.

4 Likes

Nice post. Really glad someone has data @Aes here because it has been my impression for quite some time that Maker was losing marketshare to competitors.

I have been talking about paying attention to ‘market share’ in ETH ever since Black Thursday because every vault holder that was wiped out was ETH of a user that basically disappeared likely never to come back. Once liquidity farming came on line via compound driving DAI PEG high with high demand (which I wanted to counter with rate increases vs. decreases btw) I was also concerned that Maker by not at least looking at issuing a MKR-R for MKR rewards lost out on a clean, quick and easy means to achieve negative rates, create a new distribution metric, more robust MKR markets (not just in MKR but also MKR-R), as well as encourage more deposits and staying relevant relative to competition… MKR-R I saw as a win-win-win for the ecosystem and the earlier it was done the better.

If you look at Maker from a potential user we are offering zero return for collateral deposited while these other protocols offered at least some return on deposits, then juicing those up via farming rewards with basically out of thin air ‘governance tokens’ when you deposit and borrow. There really are so many reasons to start rewarding vault owners for depositing and borrowing it isn’t funny. It would be nice if Maker could get to something more like 5% of all ETH vs. what now 1%?! but without real focus on getting Market share I don’t really see this happening.

This is something that Growth/Business CUs really needs to be looking at. I honestly think Maker has been too busy dealing with dissolving the foundation, and trying to pull itself together financially, organizationally, operationally that this kind of stuff just dropped through the cracks. This typically is indicative of a general lack of any unit taking a larger view of Maker through all the CUs effectively - and/or a general lack high level longer term planning. I think to be fair dealing with the community as a DAO is much more difficult than these other centralized coins and organizations - just popping ‘decentralization’ via ‘farming tokens’ out of their asses and then bolting those rewards on their ecosystems. This is really a difficult topic to discuss because as my wife says it is easy to armchair comment on driving a system, a lot harder to actually drive it.

I do think some idea of a CU that looks at longer term issues, has access to all internal data and people that can utilize Maker expertise as well as community ideas to form a longer term grander vision for Maker is something well worth pursuing. There needs to be a way to focus on the bigger issues and get out of the muck of smaller stuff. I remember over a year ago debating with people about skipping vaults on collateral types which were likely to not mint much DAI and focusing strictly on what will get us real TVL and DAI minted. People were more concerned getting to $1B with 100 vault types vs. focusing on 2-5 that would get us to 5B. Well guess what here we are with 2 vaults accounting for what almost 90% of the TVL backing almost 6B DAI. TWO. Let that sink in. 1 of which only pays a 1 time fee (the PSM).

Contrast this with your data showing us that Maker is continuing to lost market share and perhaps MakerDAO will finally start to take seriously how to sustainably build a model to drive the next 100B TVL here. Now maybe this is via a focus on RWA but I don’t see any reason we should not try to look at how we can most effectively compete for TVL market share and build greater networks.

I have written up some ideas that are even more relevant for Maker that I was pitching to 1Hive regarding DAOnetworks and building DAO community DeepLiquidity.

Maker basically has the stablecoin for DeepLiquidity, as well as the rep and capital to be able to cross-invest in projects and build its own MakerDAOnetwork. I really think people underestimate how democratic and effective Uniswap V2 AMM structure is and how pairing community tokens with stablecoin cash to be the liquidity backbone of trading for the entire community can be amazingly powerful. Heap in the idea of LP provider rewards (not from one but entire community of assets), brand, loyalty building, combined with PR I think a MakerDAOnetwork with right players could be a significant force to be reckoned with.

I see DAOnetworks being the next future. I also see Token-stablecoin v2 type AMM type democratization of liqudity and markets to honestly have a huge potential if pursued with the right focus of constructing community value backed with hard cash (stablecoins) along with a trading venue where communities focus their liquidity, trading, and earning being a way to create a new decentralized network effect via DAOs and the people associated with them.

Personally once I understood Uniswap V2 LP and what it means to create value by pairing a Token with a Stablecoin in terms of liquidity, price support, and democratization of risk/rewards via a LP unit of currency and adding sqrt risk into the mix I was floored by how powerful this could be. If you think about it we have companies with Stocks that we all buy. Imagine if instead of buying a Stock you bought a LP Stock-Cash pairing that was basically guaranteed to get return from fees on the trading of that stock on a broadly accessible decentralized exchange where that liquidity was provided, was completely transparent and accessible to all, and with crypto security (your wallet - your assets) The beauty of this type of pairing is the volatility of your holdings is the sqrt of the price of the token in cash and you earn the fee returns on top via trading and you can split your LP into cash and stock/tokens whenever you want. The community becomes the exchange, the LP, the LP holders, and everyone participating in DAOnetwork governance, with all DAOs and the exchange basically sharing and redistributing fees to all the stakeholders openly with full transparency. There is an elegance to this ideologically, financially, politically both from a financial, and fairness perspective and an efficiency that DeFI and crypto has still yet to realize.

The question is how can Maker lead such a larger initiative and vision looking towards building a more integrated DeFInetwork community vs. everyone playing by themselves trying to be the king of a increasingly fragmented DeFI ecosystem.

4 Likes

This data is straight from DeFiPulse but I’d like to build something more comprehensive using Dune and get deeper into the business analytics. There are so many questions I can think of in my mind leading down rabbit holes regarding the ‘business’ of the protocol but getting my hands on the data is typically not very easy.

Maybe this was before my time, but has the community considered this type of program? For simplicity we could just lower rates to be in line with the net rates of competitors after liquidity rewards are included but I do like the idea of giving users of the product a stake in the ecosystem (if they don’t have one already). Adds another incentive to drive new users.

Couldn’t agree more. I deeply believe we need to form some sort of consensus as a community of the direction of Maker and aggressively invest to achieve that direction. Just because we are a DAO doesn’t mean we shouldn’t share a collective goal or purpose. That topic may deserve it’s own post.

Would you like to spearhead this initiative? #MakerManCU. Cross-investing in projects seems interesting but feels like it would be quite difficult to govern effectively. It’s challenging to get the vote out consistently in Maker’s protocol so hard to imagine MKR holders wanting to have governance rights to other protocols (unless this was delegated to MakerManCU to manage).

2 Likes

Agree on this, these changes won’t drastically improve the current picture if you make calculation based on current exposures, but it is a step towards more competing vault offering and maybe we see some refinancing from other venues.

As said above, I partially agree on 2., but remains to be seen. As for system risk, it does of course increase it, but not significantly and this is why only 5% LR drop for largest vaults was proposed. Lower LR doesn’t necessarily lead to more liquidations (unless users suddenly start behaving more risky, keeping less buffer above LR), it mostly means probability of loss during liquidation is higher due to slippage applied to lower LR, but as said this is limited by 5% difference. Put it simply, if you have liquidation at 145% instead of 150%, you need 31% slippage from auctions instead of 33.3% to start creating losses for the system. Not a significant change, but it does affect risk premium we calculate for sure.

I am hearing some concerns lately about this topic, though only few voices seen so far and more related to the DAI minting effects this will have. I may reconsider putting this straight into on-chain poll and instead make a Signal Poll on Monday to have full support on it, because it is still important parameter change however you look at it. Also to be clear, Risk Core Unit is probably the last Core Unit in line to propose riskier system parameter changes, this proposal was made after others or say more growth oriented units have voiced support.

2 Likes

Lets try to focus on risk as this is your CU responsibility. Growth can make their own case for the need.

I am coming at this looking at what happened on BlackThursday. This isn’t a ‘slippage’ issue but an oracle to market pricing issue. Put simply if one is at liquidation of 125 and the market price drops 20% by the time the liquidation system hits this price you basically are at 1DAI market value for $1 collateral. (no 13% fee) before you hit the loss point, no return of collateral to vault holders. Now drop this to 120 and the collateral buffer to loan recovery value drops to 16.66%. With this example you have cut the collateral buffer value by 1-.1666/.2 = 16.66%. When you factor this in against liquidation fee of 13% this collateral buffer is reduced from 7% to 3.66% or by about 1/2. You also are really eating into potential returns of collateral here IF a liquidation occurs. Maybe this is good, maybe this is bad.

This assumes my math here is correct.

The numbers for 10% at 160-150 run similarly. A price drop of 37.5% takes us to loss condition at 160 this drops to 24.5% with 13% fee buffer. At 150 this is 33.33% which drops to 20.33% with fees. This isn’t as bad even with the 13% fee because of the high LR requirement. I am less concerned about the 160-150 vs. the 125-120’s because of these numbers.

If there was anything to think about here for competitiveness one should think about the liquidation fee.
From a user vault side a move from 125-120 basically means a effective increase in leverage from 3-4 (4-5 is only theoretical). 160-150 basically means theoretical leverage change of 1.6-2, effective probably 1.5 to 1.9.

If we really want to encourage growth I see doing some sort of MKR-R type vault borrow reward as a more interesting proposition that could drive a lot of TVL than the above LR modifications which increase system risk.

The simple answer here is that to increase competitiveness with the above changes (these changes won’t make much more DAI unless TVL significantly increases) risk to the system is increased by some amount (the quantitative value here probably difficult to nail down and would have to be done via risk analysis scenario via ETH pricing scenarios)

Here is a question that really should be looked at? What does the bin histogram of vault sizes look like since I don’t see these changes doing anything to vaults with 200K or less of collateral. What are the vault sizes of our competitors? Isn’t the onus on growth to make the case for some significant TVL change here via some analysis of competitors user/vaults and these proposed changes. Have I missed a report document regarding this? I am trying to understand why risk did work on this if growth hasn’t made a growth case for this?

1 Like

The drop from 125% to 120% is proposed only for stable/volatile UNI V2 vaults where price volatility of these collateral assets is about half as low.

You are right though that due to auction slippage and market/oracle prices risk, there is a lower probability of collecting full liquidation fee, which is quite high compared to competition. There were discussions about reducing it after successful Liq 2.0 rollout and we plan to address this after LR adjustments.

Not sure how much it would benefit us making competitive analysis where other venues are subsidizing borrowers, enabling them negative real rates. But this is definitely on our radar lately and we started evaluating user growth patterns at Maker, see our latest post on Maker Vault Activity Metrics. This is something that falls under Growth though, but we did it because we want to understand our user behaviour better, next step is analyzing their vault protection habits.

We made this proposal and analysis beforehand after other mandated actors supported the idea to evaluate lower LR.

1 Like

I agree with this sentiment. It’s not a long term solution, but getting more ETH in the system in the short term will de-risk the system and buy us more time to onboard RWA safely and at scale.

3 Likes

One thing that slightly worries me is the fact that we can’t adjust CRs back upward without potentially liquidating Vault users.

If we wanted to reverse the move without inadvertently hurting users we would have to launch a fresh vault type and limit the DC on the old one.

2 Likes