TIP-111: Increase stable backing of FEI

Proposals
1

Motivation

In the thread for PCV reinforcement, there is a community poll to increase the “stable backing” for FEI :

As of 16/05/2022 2pm CET, over 80% of forum participants want to increase the stable backing to 80% or greater :

image
image698×233 13 KB

Therefore, as outlined by Kirk, I think now is the time to discuss an ETH sale for acquiring more stablecoins in the PCV.

Specification

The PCV owns ~155k ETH (~310M$) and ~95M$ of stablecoins. Circulating FEI is ~230M.
The stable backing is currently ~42%.

Here are multiple scenarios to increase stable backing, selling ETH for stable values (stablecoins or delta-neutral positions) :

Scenario new ETH in PCV new Stables in PCV ETH sold Stables acquired ETH “Liquidation price”
Do Nothing 155k (310M$) 95M$ 0 0 870$
70% Stable Backing 122k (244M$) 161M$ 33k 66M$ 565$
80% Stable Backing 110k (220M$) 184M$ 45k 89M$ 418$
90% Stable Backing 99k (198M$) 207M$ 56k 112M$ 232$
100% Stable Backing 87k (174M$) 230M$ 68k 135M$ ~0$

Note that these numbers do not account for the ~40M$ PCV expenditure for Fuse hack repayment that has passed Snapshot voting. To achieve the target stable backing, the target number of stablecoins shall remain unchanged. But for hack repayment, the equivalent of ~20k ETH from PCV shall be spent, reducing the numbers of the “ETH in PCV” column. After hack repayment, the table looks like this :

Scenario new ETH in PCV new Stables in PCV ETH sold Stables acquired ETH “Liquidation price”
Do Nothing 135k (270M$) 95M$ 0 0 1000$
70% Stable Backing 102k (204M$) 161M$ 33k 66M$ 676$
80% Stable Backing 90k (180M$) 184M$ 45k 89M$ 511$
90% Stable Backing 79k (158M$) 207M$ 56k 112M$ 291$
100% Stable Backing 67k (134M$) 230M$ 68k 135M$ ~0$

The column “Liquidation price” represents the moment when, if ETH price goes down in USD terms, the protocol will own less PCV (in dollar terms) than there is circulating FEI. This situation may lead to large FEI redemptions, draining of the PCV, and the need to activate the TRIBE Reserve Stabilizer (that can inflate the TRIBE supply by up to 100%/year in order to capture circulating FEI and restore Collateralization Ratio to >1).

Voting options

Vote will be a Snapshot approval vote, with the 5 scenarios as 5 options, the option with the most approbation wins.

This vote is about establishing a formal DAO-wide policy for a target stable backing %, and should happen quickly.

The actual ways on how to achieve stability (which stablecoins, which hedging mechanisms, and in what proportions if multiple services are used) is up to discuss in this thread and will be the object of follow-up snapshot votes or Tribal Council actions directly.

3 Likes
2

I agree this is a unique situation and requires selling of ETH to maintain stable backing. But after this turmoil is done. We really do need to think about how efficiently to manage PCV. We are only loosing PCV by buying assets for PCV high and selling low. We should have planned to back more stables when ETH was going up and around 3800-4800. That is why we need a plan for the next time.

2 Likes
3

One other way to think about increasing stable backing is preferentially accumulating stables when demand picks back up, which is how the PSMs are currently structured.

Having 33k be the most conservative option is biased towards selling in size. I think there should be at least one more polling option in the snapshot for selling a smaller portion of ETH like 20-25k.

Another thing to consider is the Protocol’s obligation to acquire $10m in OHM in exchange for the 50m of FEI demand generated by Olympus DAO last week. This could be done by selling ETH, too.

4 Likes
4

the biggest profit generator for the DAO in the short/medium term will be eth price appreciation as it likely hits 10k+ sometime over the next 2-3+ years. Selling that to buffer fei more isn’t worth it at this moment… especially because we (the dao) haven’t yet hit a true flywheel of product/market fit for fei where it is making the dao a bunch of revenue. Said another way: We don’t have a good enough reason to cut off our main trajectory for a robust treasury until we figure out how to sustainably deploy fei out into the wild where we are paying LESS than the liquidity mining rewards. We’re basically getting rid of the real source of growth for the DAO for more fei buffer when there is no real demand for fei and likely won’t be until the market bottoms anyhow… this would be a major facepalm for the dao from my pov.

We all know selling eth below 2K is going to be such a mistake long-run that I think the best option is do nothing, closely followed by selling an amount less than 33k (Maybe 15k) and just live with the results.

1 Like
5

As Joey mentions here, the Protocol has an obligation to acquire $10m OHM. This could be done using ETH and would indirectly increase the protocol’s stable exposure and stable collateralization ratio. The issue with these metrics is that unlike Dai or LUSD, the OHM will not be used in a PSM. Thus this transaction would increase the stable collateralization ratio of Fei but not the redeemable liquidity.

I think an ETH for OHM sale through an LBP is a good first step to increase the stable collateralization of Fei. Would love to hear more thoughts under this broader discussion first and will be bringing this to its own proposal in the forums in the 24 hours.

6

“Another thing to consider is the Protocol’s obligation to acquire $10m in OHM in exchange for the 50m of FEI demand generated by Olympus DAO last week.”

Which FIP was this? Or, is this the liquid representative governance at work? if so, is there a way to see tribal council discussion or votes on this and other topics?

Are there any other obligations that the protocol has acquired recently?

7

Let’s take ~227m User Fei, subtract 95m (backed by stablecoins), then 57m (VC funds unlocking over 3 years) to have 75m ETH-backed user FEI. Considering this, the functional liquidation price is much closer to $550-600. With current backing it seems that we’re well collateralized. In my opinion a 5-10% APR bond program would be a far better option than selling ETH, as I still see value in reducing the amount of FEI that can be redeemed while ETH price is low.

2 Likes
8

While I agree the risk of insolvency (FEI permanently below peg) is low, the risk of equity loss (redemptions flow larger than stablecoin reserves at a lower ETH price than today) is larger.

Would be in favor of swapping ETH<->OHM rather than reducing the current size of the stablecoin reserve. An LBP is one idea, could also suggest an OTC of the ETH for OHM directly with Olympus at RFV, just like they were able to mint FEI using existing treasury DAI.

If we do the Olympus swap in ETH, reduces need for further open market ETH sale.

1 Like
9

Agree, considering the turmoil market condition, all actions should been taken ASAP.

Market might change 20% easily within 1 day.

10

Also Fuse hack repayment is going to cost ~30M in stables in the current execution plan that @klob is working on, where the hacker’s bad debt is repaid to be able to re-activate affected Fuse markets.

If the price of TRIBE decreases, more FEI enter circulation, the way this vesting is made (LP position).

Also this raises the question of : would anyone use FEI under these circumstances, if they know the PCV is not large enough to cover all FEI redemptions ? Partner DAOs might be comfortable holding through the winter, but it may affect the narrative the Tribe DAO is having about FEI, and reduce/delay user adoption.

I personally think we should act soon, and while an ETH sale is not the most elegant solution, it is something we can predict the effects of, and can execute quickly.

11

Soon always better than delay, No matter how much ETH decided to been sold.

The marco did’t give a V-shape opportunity for us, stay safe in winter is priority now.

12

I’m a bit hesitant on making a singular large swap of assets.

Is there a way to do this in tranches/steps or only sell ETH (or other assets) when some sort of trigger (say % backing or ETH prices) is activated?

13

Any update on selling ETH ? Also what is the consensus on selling may be 20% of more ETH right now, excluding the ETH sell to back the stables ? This way if ETH goes lower, we have dry powder to buy more ETH and if we rip from here then we can get on the train at a higher, which might be that bad since everyone here is in the ETH to 10k+ in the years to come train.

14

What was the result of the vote.

15

Following the veto snapshot vote posted by @OneTrueKirk that has not succeeded (proposal NOT vetoed), 2 hours ago the Guardian multisig executed a transaction to move 20k ETH to the ETH PSM.

So far, 223 ETH worth of FEI have been redeemed by a MEV bot.

The ETH PSM is in “redeem only” mode and the DAI PSM is in “mint only” mode, so there should be a sell of the 20k ETH over time, and possibly an inflow of DAI in the same period.

16

A Tribal Council action was proposed to:

  1. Sell 20k ETH for DAI over 2 days in a Balancer auction
  2. Tighten ETH redemptions for FEI to 60bp fee

This follows similar logic from the above post to increase stable backing for FEI.

The NopeDAO veto vote is on-chain here: Tally | Tribe NopeDAO Proposal

proposalId= 0x81c419dbb3c44645493c214eee0ceaf273ba870d1e7e1c48422e2762f60e0db4
execution timestamp = Jun-11-2022 08:30:06 PM +UTC

17

With the turbulence in the broader financial and crypto markets, Fei should consider diversifying the existing PCV further. Including less volatile assets in the PCV will allow the Protocol to maintain a higher collateralization ratio and to better endure periods of extreme volatility. TIP-111 is a great first step in that direction.

Following TIP-111, we would encourage the community to consider further methods to increase the stable backing of FEI, potentially to 90% or more (after accounting for repayment of the Rari hack). Doing so will enhance Fei’s resiliency and allow it to navigate the current market environment safely.

Jeff
a16z

8 Likes
18

Totally agree, I would go further and make an assertive statement: every FEI backed by a decentralized stablecoin. Imo this reduces the governance risk of the uncertainty on pcv management and provides more safety to the market in the current conditions. Following this, the team can have more peace of mind to focus on advancing the product roadmap that will create more use cases for Fei as Turbo and others.

1 Like
19

In another thread I shared a the spreadsheet of the current state of the PCV :

image

Given the current macro environment and state of the PCV, I believe we should move much more aggressively on increasing the stable backing of FEI. This topic has been in talks for over a month now.

A lot of comments are going in this direction in the community.

Out of the 137,538 ETH owned by the DAO :

  • 20,000 are currently being sold for DAI in a 2-day Balancer LBP auction
  • 49,713 are in Lido (stETH). Please note that Lido is suffering from peg problems (currently 4%)
  • 19,361 is in Aave
  • ~18,500 in Balancer FEI/WETH pool (staked in gauge to earn BAL rewards)
  • 16,389 is in Compound
  • 10,000 are in Tokemak, and will become withdrawable after the 21st of June if TC executes a withdrawal request in the next days
  • 3,115 are in the ETH PSM

If we consider Lido as illiquid (staked ETH is suffering from a depeg right now, and ETH staking rewards is the best income for the DAO right now), that leaves us with the reserves in Aave/Compound/Balancer/Tokemak to work with.

At the current market price of 1,544$/ETH, selling all 19,361+16,389+3,115 ETH in Compound/Aave/PSM (and considering the 2-day 20k ETH auction complete) would bring the FEI stable backing over 90%.

Over the course of next week, the TC can take action to prepare withdrawal from Tokemak and Balancer gauge, to make an additional ~28k ETH available and liquid for further action.

Following these actions, some ETH can be sold to finish bringing the FEI stable backing to 100%.

The privileges needed to execute this plan are :

  • Guardian can move Aave+Compound ETH to the Balancer LBP, to be sold over the 2 days following the current auction (i.e. between the 13 and 15th of June)
  • Meanwhile Tribal Council take actions to put Tokemak & Balancer ETH in a liquid position (week of 13-19 June)
  • Around the 18th of June, Balancer position becomes liquid following the TC action
  • Around the 21st of June, Tokemak position becomes liquid following TC action
  • Guardian can take action around the 18-21st of June to move ETH from Balancer/Tokemak (as needed) to start a final LBP auction of 2 days that will bring FEI stable backing to 100%

While some of these might be Guardian actions, and these usually don’t go through governance, given the impact this would have on the protocol, I will publish this plan as a Snapshot vote over the next 2 days to gauge the sentiment from all stakeholders. This execution plan will also be subject to the approval of other Guardian multisig signers, and Tribal Council members.

We have an amazing community of builders and the potential for our product is incredible. These are bad times and it sucks to have to take such decisions, the time is not optimal (is it ever?) but I think it has to be done. The current collateral ratio is below 150%.

3 Likes
20

The snapshot vote is live for the next 2 days :

https://snapshot.org/#/fei.eth/proposal/0x7e06e143e5e0f7c5655f2a1ebdc1f18b19638a397ad6e8077b3c12ccadff012a

1 Like