[Signal Request] Offboarding MATIC

That’s why I raised QiDAO, it is easier for the community or the Polygon Foundation to issue a stable currency of equal value but with more chains to operate as MAI.

No doubt that what you mention about the inoperability of different chains is a yoke for Maker. Not everyone has the possibility to open vaults of 10k or 25K DAI.

Maybe what I recommend and I hope, is that you publish it in the forum, agreements what is working and so, so avoid passing those posts offboarding Nik and the Maker community learn what is doing the Matic Team.

1 Like

I can see why there may be confusion generated through miscommunication. I apologize for that. I can only speak to our internal communications which went from discussing 100M Dai to a lower figure of 30M Dai due to the liquidity profile of MATIC. The 30M figure is essentially a minimum amount borrowed the Maker Protocol needs to make a collateral profitable. The Core Units refrain from onboarding any asset that will generate less than this minimum threshold. MATIC was prioritized because it was believed a deal was in place that would guarantee that floor.

So I agree with you, maybe the expectations for both sides were unclear, so we can hardly hold that against you. I’ll edit the original post to remove any references to a deal and frame it exclusively from the economic point of view.

That said, it doesn’t change the situation that the Maker Community finds itself in where MATIC is unprofitable. This isn’t singling Polygon out, Maker Governance has been very active in offboarding collateral types that fail to meet this floor including USDT, KNC, BAL, COMP, AAVE, ETH-USDT, DAI-USDT, LINK-ETH, and AAVE-ETH. This is likely to continue with the community already having MANA, UNI, and UNI-ETH as future possible candidates for offboarding. If we can get the MATIC vault above the 30M threshold then the vault would be profitable and the Maker community would have no reason to offboard.

Maker currently does not support borrowing of Dai on other chains, and doing so would be a huge undertaking. Whatever may happen in the future with regards to issuance on Polygon PoS is out of scope for this particular poll which exclusively is limited to MATIC on Ethereum.


I think there’s a misunderstanding because Polygon team has been actively engaging with Maker protocol even before years before the Foundation dissolved. They also have been introducing various projects on Polygon to Maker Growth team to help to integrate and promote Dai.

So unlike many other collateral types, MATIC is one of few protocols that clearly has benefited the Maker protocol in other ways even if their vaults might not be profitable at the moment.


Remember that there’s 400 million + Dai on Polygon. Even if we consider their contribution at 10%, that’s already 40 million (which is more than 20 million to make the vault profitable) and as more and more projects have been on boarding to Polygon and they continue help to connect with Maker Growth team, they will continue to increase the demand for Dai.

Therefore, the community should consider long-term consequences


The current DC is set to 20 MM and the gap is also increased to support minting all of the DAI in just one TX.

Zooming out, onboarding of MATIC hasn’t been profitable so far. So I do agree to the basic statement: if it is not utilized in a way profitable for the protocol, we should offboard.


Understood, thanks for sharing the thoughts @NikKunkel

I understand, and good to get the perspective from the Maker team.

We would work more to promote it, and see Polygon community support the MATIC on MakerDAO.

1 Like

Can you walk us through the math a bit? At 3% fees, that would equate to 900k in oracle costs to cover the MATIC vault.

This seems contradictory to the estimates you give in the MIP for oracle gas costs – where OSM + medianizer + spotter is around 500k in estimates.

It also seems likely that YFI should have comparable oracle costs, and is a much older vault that is not currently at all-time-high useage. Given that MATIC is only 3 months old, should Maker consider offboarding YFI? Or is there something that makes MATIC specifically a more expensive ilk to provide oracle services for.




Based on comments:

Given my understanding here I believe a reassessment would be appropriate. It looks like the Polygon team wants to move to satisfy Maker demands, but also has goal objectives it would like to meet.

I honestly have been watching this vault for a while and wondered why the full 20M DAI wasn’t drawn. I think there is significant potential with Polygon being missed here.

I will vote no to offboard MATIC as I think this would send the wrong signal from Maker to Polygon. We really need to straighten this out. Right now we are earning perhaps 1/2 the 600K/yr needed. It will cost us 75K DAI to give the teams 3 more months to work out some arrangement. I think this 75K and up to even 150K would be well spent to try to secure a 15B Token ecosystem with Maker. Polygon could easily become a 100B MC and is one of the largest side chains in operation.


I voted NO for the following reasons:

  1. Polygon team committed to open a vault using $50M worth of MATIC, and they did that (currently, there are $54.6M locked). I understand that for us, what’s essential is the minted DAI, so before approving their vault we could have asked for that. They never committed to mint 20m DAI nor 30m DAI.

  2. It Is not clear what “unprofitable collateral” is. Is it collateral with less than 900k annual fees? After how many weeks/months should we evaluate each collateral and classify it as profitable or not? Are we going to offboard all collaterals with less than 900k annual fees?

  3. Once we are clear on classifying collateral as unprofitable, we could think of raising the SF to make it profitable or giving the team that represents the ilk x amount of weeks to increase their debt to make it profitable.

  4. We need clear rules and a procedure to offboard collateral. This will help our community and partners to determine when to initiate this process.

  5. Polygon is a strategic partner for Maker. Currently, there are almost 300M DAI in their network, and we are working with them to integrate DAI in the Dapps on polygon


@monkey.irish , is this something that @collateral-core-unit is looking into? I want to make sure there’s no duplicate work.


Yes, as we’re getting more familiar with the current processes our goal is to identify where we can make improvements. Providing clear guidelines for collateral offboarding is a key part of this work that is already on our radar and has been discussed a bit in the team.
Having clear guidelines on collateral offboarding will also help clarify what type of collateral is “good” collateral in Maker, in terms of minimum requirements, which can help set the right expectations for MIP6 applicants and current collateral partners.


I can say after experiencing the Aave offboarding myself, that the process needs to be revamped. I’ll keep this short and would love to discuss further with the MakerDAO team as the current expedited and abrupt process has caused great financial loss.

How many of the total offboarded vaults were liquidated vs. self-closed? I think the numbers will show there was a lack of communication and hurt those that were attempting to help diversify the system. Sure there was no liquidation penalty, but those positions weren’t given many opportunities to minimize impact.

I don’t disagree with the decision to offboard those collaterals given the “costs” but there NEEDS to be a more orderly process once approved.


I had some thoughts on this about a month ago. The main sticking point I had was coming up with a way to get votes on chain in a timely fashion outside of the Signal Request process. I discussed this with some other GovAlpha members at the time. I’ve subsequently been distracted by some other things so haven’t really returned to it.

I’m happy to discuss further in DM here or on discord.

1 Like

One issue I encountered is in regards to Defi Saver, which I heard from Nikolaj is that “The thing with these offbarding processes is that dfs automation cannot react to them, as the system cannot recognize this “next liquidation ratio” as it keeps going up”.

Overall though, there needs to be a longer period before the dss-lerp begins to increase. You gave a very minimal window to vault holders from the time of the Executive vote being passed. I would recommend a “communication and preparation period”. What are the checks and balance differences on onboarding vs offboarding assets?

Just in this thread alone, based on the confusion between Matic and Maker, it’s clear that communication and process are lacking and improvement required.


I do think there needs to be a revamped process around communication of offboarding to users. As someone who was impacted by this, do you have any suggestions on what method would have been best to inform you? Keep in mind we don’t control the front-ends for the Maker Protocol.

I don’t quite see why your comments re DeFi Saver are relevant. These vaults aren’t getting liquidated because the price is dropping, they’re getting liquidated because the collateralization ratio is being lerped higher. With the liquidation being set to 0, it makes little difference whether DeFi Saver Automation liquidates one hour earlier than the Maker Protocol.

I only brought up DFS as I had it activated yet it did nothing. I wanted to bring light to why it did what it did as other users may be in a similar situation. I was initially stuck wondering why it didn’t sell my collateral in chunks to maintain ratio, but the dss-lerp would be moving too fast and DFS is a fixed %.

I’m in the process of writing a bit of a briefing / impact statement and will be happy to post it.

1 Like

As promised.


Interesting developments from the Uniswap community, as they look to deploy v3 via Polygon with an LM program:

1 Like

This signal request has concluded unsuccessfully and no further action will be taken at this time. Thanks to all who voted.


This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.