Adding a poll to prioritize collateral applications

There are currently 7 different collateral proposals that have been greenlit by the community and are pending feedback from the foundation in one form or the other. Additionally there are another six or so applications that are pending community greenlight.

It seems that we may soon find ourself in a situation where the foundation will have a large backlog of potential collateral to evaluate and greenlight. I wonder if process changes are needed to ensure that the domain teams can focus attention on collateral applications that maker holders view as having the most potential for ecosystem impact.

One potential solution to this problem might be to have a regular governance poll that goes out to determine which collateral application is a top priority for maker holders. This way the domain teams can be sure to focus their efforts on those applications first.

I would suggest that we add a regular possibly semi-weekly or monthly poll where we ask maker holders via a ranked choice poll to determine what is the highest priority community-greenlit application to onboard next. Having a ranked choice poll should hopefully also give us some signal into what is the second and possibly 3rd most important collateral applications as well.

Will add a poll to judge if people are in favor of this idea. Let me know what you think in the comments!

Are you in favor of adding an additional governance poll to prioritize collateral applications?

  • yes
  • no
  • abstain

0 voters

If people are in favor i will add some additional polls to hopefully determine what might be the appropriate interval to poll MKR holders.

2 Likes

This gets sort of tricky, because it implies that collateral that didn’t pass the initial greenlight poll is off the table. This week’s poll results (so far) are really throwing me for a loop, I think we may want to explore how MKR voters could revisit assets that previously failed the community greenlight poll as well.

Considering that some MKR whales have bags/portfolio investments in other projects, it might also make sense to agree on more objective criteria for prioritizing inclusion to avoid manipulation of the governance process. Possibly AUM, liquidity, token gini coefficient of distribution, if the token/project actually exists already or is just an idea, etc.

3 Likes

Yeah i mean just working with the greenlit applications seems sensible to me, but polling on non-greenlit applications is definitely something that could be considered in later polling rounds supposing that the idea is favored generally.

While i definitely agree that objective criteria should be used to inform prioritization i also think that prioritization is often an art and not a science, and FWIW In my opinion having an existing token should be a pre-req for greenlighting but that is just me :stuck_out_tongue:.

Basically i think codifying the rules for “what is the most impactful token” is going to be a pretty complicated exercise and likely would result in picking a few metrics that don’t really take in to account everything that should be considered.

I don’t really have a solve for the fact that MKR whales might have conflicts of interest. Seems like the a risk that might just need to be accepted with plutocratic voting systems, but definitely something worth considering here.

I am not in favour of adding a prioritization poll until prioritization proves to actually be a problem.

Right now the first-come-first-served approach seems to work just fine and this also allows for planning and predictability at the foundation. Reordering the schedule based on a poll would most likely not contribute to overall quality.

2 Likes

Some points in response to this thread:

Community Greenlight Polls
The community greenlight polls already achieve the goal of providing rough prioritisation for the domain teams. The greenlight polls produce scores that can be used as one way to rank the potential collateral assets.

Directing Domain Teams
As it stands, domain teams are specifically not required to work on the collateral that MKR Holders most want them to work on (at least at this stage where they are not being paid by the protocol.) Community sentiment towards an asset is meant to be one of the inputs into their decision of what to work on, but not the only input.

There are three reasons we made this decision when writing the collateral on-boarding process (which the collateral onboarding MIPs are ultimately based on).

  1. The domain teams have a much better understanding of the risks and ease-of-onboarding within their domain. This lets them prioritise more sensibly to achieve maximum throughput.
  2. In the current state of affairs the Maker Foundation pays the domain teams salary, realistically this gives them control of what the domain teams work on, and you can’t hand-wave this away regardless of what you write in a MIP.
  3. Asking people to work on things they don’t believe in is a bad idea. It hurts morale, and you end up with work of a lower quality. In the future there is also the possibility of multiple teams within a single domain that have different specialisations. It doesn’t make sense to force a team to work on something they have no experience with.

Repeating Community Greenlight Polls
The most recent amendment to MIP9 changes how repeating polls works to some extent. The previous version was somewhat ambiguous. The new amendment gives the governance facilitators leeway to repeat greenlight polls if there is community demand to do so. This was viewed as a compromise between ‘Repeat all the failed polls every month’ and ‘never repeat the failed polls’, and allows us to keep a lid on the number of greenlight polls happening at once.

On the current approach
The current approach is not ‘first come first served.’ The current approach leaves the decision up to the domain teams as to what they work on each month.

5 Likes

@LongForWisdom great response. Some interesting points regarding “Directing Domain Teams.”

Care to elaborate here? Just thinking about it I could see a few potential problems with trying to infer priority based on the greenlight polls. It seems like any inference that you make is going to be heavily skewed based on voter turnout on a given week and would potentially suffer from the problem described by @Adam_Skrodzki here:

I.E why vote on the greenlight poll for say LINK when it is already clear that it is going to pass?

Ultimately, it just seems to me that as long as we ask one question to infer a different one we stand to make the wrong inference, so given that we have the facilities to ask this question directly why wouldn’t we just do that?

Poll is closed. 18 responses 72% yes 22% no.

Given the lower engagement not sure if this is something that I want to push forward with for on-chain polling.

If people want to see that happen just let me know in here and I can start polling the forums for on 2 outstanding questions here.

Outstanding questions are:

  • Poll frequency
  • If the poll should only contain greenlit collateral applications.
1 Like

Couldn’t we prioritize in basis of the following rule>> Collateral with bigger approval % wise and then filter it with collateral with higher amount of mkr staked? Or apply a % between the 2 to ponder sum them up and rank it that way? Also the priority should not be mandatory and the different teams can gauge community sentiment from that data.

For example I’d love to have real world assets as a priority, however, there are many aspects related to onboarding them that I may be unaware of and if oracle team, risk team go through with that it may actually take longer than other assets. I think that the priority should be to maximize diversification while increasing dai minting and I don’t think we can get that from polling.

I think the best course of action in order to proceed with this is to propose an amendment to MIP9. Informal polling prior to that is potentially also a good idea.

Given that the current behaviour is currently defined by a MIP, that makes it problematic to modify or extend it using the signalling process.

Feel free to reach out to me if you want to discuss how to go about doing this @Andy_McCall

1 Like

I wanted to bump this thread as I think it’s a pretty important point, and it hasn’t really moved into MIP9.

It seems like the backlog of collateral is growing, and if the community can signal which assets to prioritize, it may help. How is the smart contracts team currently prioritizing collateral?

One speculative hypothesis I have for LRC being rejected was because it was deemed less important than other collateral types (I don’t have evidence for this). Additionally, it seems like onboarding a wBTC alternative (like renBTC) is more dire than adding other ERC-20s that are being worked on right now.

A forum poll would be the preferred method IMO, as we don’t have to make it binding, but just get a general pulse on what people think is important. Instead of mandating something be worked on, it can simply inform the process.

Listening to today’s call, we got a glimpse into why they choose specific collateral types. Based on oracles, some collaterals are not technically possible yet due to existing infrastructure. Some collateral types are more “low hanging fruit” risk/technical wise and I think that flavours the Domain Team’s decision making. RenBTC will also be difficult to assess due to it’s closed source nature but I would also love to see it added somehow. Having said that, having a more granular priority poll would definitely be a good thing.

1 Like

When the price of LINK was $4, the community voted to pass it, but so far, we have not seen it. I am against LINK now. The $16 LINK is more risky.

We missed LINK, we missed MKR, we missed… Finally, these all went to AAVE. Now as long as someone maliciously attacks the MAKER, they only need to borrow tens of thousands of MKR from the AAVE platform. If the MKR is placed on our own platform, there will be no worry.

1 Like

@hongbiao_li
Maker did not miss LINK - it is being worked on.
Maker did not miss MKR - it is simply unwise to have your own coin as collateral.
This is lesson number 2: https://www.youtube.com/watch?v=OJM8yJTn_I0

A poll for collateral prioritization? If it is onchain it will cost me USD 3 per vote and still not deliver more collateral types. If it is a forum poll it is simply a beauty contest that adds nothing but noise.

1 Like

If someone borrows 50,000 MKR from AAVE for an ESM attack, how can we protect ourselves?

If someone borrowed 50k MKR and wanted to use it for voting they would first have to lock up their MKR in the governance contract.
Leave MKR in contract->they will have to vote against their own selfish interest
Withdraw MKR from contract->it is going to be very easy to overturn anything they voted for.
Read more here: https://www.reddit.com/r/MakerDAO/comments/b7gz4k/faq_governance/

1 Like

MKR being available to borrow on aave is way more dangerous than having it as collateral in maker. Assuming that the price of MKR drops more than 10x in an emergency shutdown (really it will go to 0), then it becomes profitable to do an ESM if you can borrow at least 55k MKR. The total supply on aave is not that far from it

1 Like

This is an existential problem I have been banging the drum on. I’d like to see some actionable measures taken to make sure this doesn’t happen. I’d do it myself but alas I have neither the brain power nor time.