Create a FEI PSM


Thanks for starting the discussion but I believe this would be not make much sense at all.

When we introduce a PSM, we’re assuming that the token in the PSM is more likely to stay pegged to $1 than DAI. With USDC and PAX, this is reasonable because those tokens can always be exchanged with a centralized entity for $1 all the time. This is not the case with FEI.

Secondly, while we are exposed more heavily to USDC than we’d like, exposing ourselves to another stablecoin should only be done if that stablecoin’s ability to stay pegged, not get blacklisted, etc. is at least comparable to USDC. Diversification is good but only when the alternative option is as good and secure as USDC. Unfortunately, FEI’s mechanism has not been tested sufficiently and the protocol is too new to make this claim. Here’s FEI’s entire history (six months).

I find the risk too high to introduce a FEI PSM with the data we have so far and what we know about FEI’s mechanism.


Some use cases of a FEI PSM include the Fei DAO utilizing the PSM to market make FEI-DAI. For example, our DAO could use FEI to mint DAI, and provide additional liquidity to the existing Gelato UniV3 FEI-DAI pair , or other platforms like Curve and/or Balancer.

I am confident that the Tribe community would be open to providing TRIBE incentives.


Hello @PaperImperium thank you for your reply and relevant questions.

  1. FEI clearly has a weaker peg, but has a more decentralized collateral base and desire to leverage the PSM. The benefits include: trading fees, more decentralized collateral and growing DAI supply

  2. Can’t hurt DAI peg due to USDC PSM and relatively small size

  3. Has similar characteristics to D3M, I’m sure FEI holders would be open to a covenant

  4. No, PSM usage is great for us

1 Like

Echoing what Bri said above, there are some pretty clear synergies if the Maker community is open to this addition!

IF MakerDAO’s goal is to only add assets with strictly greater than or equal to USDC in terms of peg and Lindy that would invalidate pretty much all other decentralized options. Totally respect that viewpoint if that is the case, and FEI would not make much sense to add.

However, I want to clarify some things about the data to enhance the argument as to why FEI is both secure enough and brings meaningful value to MakerDAO.

  1. The v1 peg mechanism did not work as you can see from the chart. We migrated to effectively a 1:1 redeemability model in June, with a current global collateralization ratio of over 300%. This has led to the tight peg seen recently. Impressive considering the almost exclusively volatile-asset collateral base (no Fei-equivalent to a PSM)
  2. Fei V2 will be announced soon and released in the coming months, further improving on the existing mechanisms and managing tail risk. MakerDAO and. DAI can play a huge role in this upgrade.
  3. Fei provides trading fees, decentralization, and DAI liquidity for the relatively small risk incurred by the capped exposure.

Basically MakerDAO would be helping the FEI peg plus issuing a line of credit in exchange for these benefits. I’m personally excited about the potential here and think FEI would be a logical add for MakerDAO.


How does this model work in practice? Are there docs available? Why is this only effectively 1:1 redeemability?

How will V2 related to V1? Is the token changing? The stabilization mechanisms?

Assuming FEI does have a robust 1:1 redemption mechanism that allows it to keep its price, it might be worth increasing the out fee in order to keep it in the protocol given that we need (decentrally backed) supply.

Feels like guarantees as to the backing in the future would be important as well. Will FEI always be overcollateralized? Would DAI be kept below some percentage? What other assets would you consider adding? How does FEI recover from an event that leaves it less than 100% collateralized?


PS: I hold some $TRIBE.

Having the other protocol stablecoin on our balance sheet is kind of lending to each other at ~0% cost and creating a counterparty risk.

Therefore having a 125M FEI-PSM is extending a $125M credit exposure to FEI.

If we abstract the details, both protocols can provide such a facility with a Uniswap V3 pool. You have 52M of $DAI and an infinite amount of $FEI for that. We have an infinite amount of $DAI. the real question is who should do it and who really needs it?

I can see why having a tight peg is of strategic importance for FEI. But it’s not really a problem for MakerDAO. Moreover, you have all the tools to make it happen. You could even do a 1:1 exchange rate (as you can use the DAI on Aave Compound to generate fees).

125M is quite a big credit exposure (2.5x our equity buffer) to a new protocol that is quite leveraged and yet it does not achieve much on our side in terms of reducing the USDC-PSM. On the other hand, you can easily have a 125M credit exposure to Maker (and it would provide you 50% of your stablecoin liquidity). You can even profit by putting it to work on money markets.


I guess I’m just not clear why this would be appropriate for a PSM rather than a more traditional crypto collateral.

I’ve not yet had my coffee, so maybe I’m missing something obvious?

There would be no demand for it as a regular collateral. At least not FEI by itself. It’s meant to be stable, so there is no incentive for anyone to lever up on it.

Like, depending on the LR ratio you could do something similar to what we did with USDC initially - but at that point the risks are very similar to the PSM and the PSM is a neater mechanism.


Thank you. Correct, why would someone lever with a stablecoin for another stablecoin except to yield farm. Time for that coffee…

1 Like

True. I can tell you with or without DAI the FEI community will expand their bonding curve via RAI, LUSD, etc., per the community proposals I’ve come across. So, lets see if the Maker community cares, or not.

Although I like the project, it is too early to talk about making an PSM for a project that is still very new.

It’s a risk, and as Paper says here, it’s so new that I see its only use in.

You may be right and it certainly is exciting but even if I was completely agnostic to the protocol behind FEI itself, the age and corresponding untestedness of it still makes it far too high a risk.

My personal opinion is that FEI needs significantly more time to mature before it can be used as collateral that backs DAI whether it be through a PSM or a traditional vault. I encourage you to start a poll or even a signal request, if you’d like to hear the broader community’s opinion on a specific question.


Can you expand on “will expand their bonding curve via…”? I don’t know much about the FEI mechanics and having some info here relevant to Maker and this PSM proposal would probably enrich the discussion.

1 Like

@swakya FIP28 will provide an idea of the Revamped FEI Bonding Curve: FIP-27: Revamp FEI Bonding Curves - Proposals - Tribe

1 Like

My issues with FEI PSM were mostly pointed out by others already, but they are;

  • FEI PSM would help stabilize FEI more than dai, as dai has a premium issue (more demand than supply), while FEI has a discount issue (more supply than demand).
  • In instances of large market drawdown demand for dai would further increase as vaults want to repay their debt - in such cases reliable PSMs are crucial to match the sudden demand for dai and prevent an even larger dai premium. Since FEI itself is vulnerable to large market drawdowns as it can become undercollateralized, I don’t think it is suitable as a PSM asset.

I also want to point out that the statements made in this post are contradicting each other, at least from how I see it. USDC backing dai presents a substantial regulatory risk, and to move forward confidently, DeFi must be decentralized and permissionless; and decentralized stablecoin FEI is going to reduce it, while FEI itself wants to increase its exposure to dai and therefore increase its own exposure to centralized assets.


Are there alternatives to a full blown PSM? Maybe one with a low DC? If FEI deviates too hard from peg a huge PSM would not be great for Dai, but smaller scale collaboration between the protocols would still be nice.

1 Like

I’m glad to see so much engagement from the Maker community because adding FEI is not strictly positive, its more of a give-and-take arrangement in terms of the benefits to each side.

I largely agree that this integration is early, and given our upcoming v2 we will focus on that rollout and come back to MakerDAO after a successful launch in a few months. I’ll go ahead and address many of your concerns here:

  1. Adding FEI to the PSM will stabilize FEI not DAI:

This is not in dispute at all. By adding a non-zero spread MakerDAO would profit off of FEI volatility in exchange for this extra stabilization. Framing this as a “peg stability benefit” to MakerDAO is incorrect, even though the name is PSM.

We plan on announcing V2 very soon and updating our docs accordingly. FEI will be redeemable for PCV collateral with a spread (50bps in the case of ETH). This part of V2 is currently live on-chain.

The token is not changing, it is mostly an additive upgrade for managing the scenarios where PCV is undercollateralized and algorithmically rebalancing asset allocations.

This seems totally reasonable

FEI will always be collateralized over a governance controlled-target which will likely be 100% but could be slightly lower. DAI can be kept below a certain percentage. Any ERC20 is open to be added by TRIBE holders. PCV would algorithmically shift towards stablecoins as it gets closer to 100% collateralized, and if it gets under then TRIBE will inflate similar to MKR.

To be clear, Maker adding FEI would both decrease the centralization of DAI and increase the centralization of FEI. I see this as the clearest mirror to FEI gaining additional stability at the expense of additional risk for MakerDAO.

The Fei community is actively deepening its usage of LUSD and RAI as collateral. DAI is currently the most liquid and trusted decentralized stablecoin, but other alternatives are growing fast.

A full-blown PSM would make the most sense given that FEI is a stablecoin and demand would be almost 0 for a normal vault.

We will aim to get a broader opinion as we roll out V2 to see if the Maker community has any interest at that time.

Thanks everyone for the engagement so far! Truly impressed by the thoughtful and diverse perspectives being brought to the table


I like this proposal but I would change the total FEI exposure to less than 20M, same for other algo stablecoins (the ones that kinda work), I want them in the PSM with a low debt ceiling.

1 Like

I’m not up on FEI’s mechanisms, but is there a reason any not to just park a bunch of your ETH in a vault, generate DAI, and use that to manage your own PSM, or even a pool on Uniswap and Sushiswap? Or use the DAI you already hold to create a PSM?

I guess I just don’t see how this can either 1) support the peg for DAI, or 2) generate fees.

If you really wanted Maker to do this (our approval alone seems likely to help stabilize FEI’s peg some), maybe there can be some kind of regular payments or additional collateral to cover those times that FEI is under peg to $1 USD?

Trying to think outside the box, and I reserve the right for the above statements to be stupid for some obvious reason.

I agree with this.

It seems FEI would benefit more from deploying a DAI PSM than we would benefit from deploying a FEI PSM