[Signal Request] YFI Vault Options

Hey all,

YFI-A vault type debt ceiling is maxed out and the community needs to decide what to do next. I noted here that we already need to increase fees from the current 4% to at least 7% according to risk premium calculations based on our model. If we were to increase DC for another 10m or more, fees might need to be slowly increased into double digit territory. Although we estimated maximum DC to be 46m DAI, the current structure of vaults is seen as very risky as there is one vault with 11.5m DAI debt that has only 204% collateralization (14% price drop and it gets liquidated). However this vault seems to be quite sophisticated and avoided liquidations a few times last week.

In the last Signal YFI-A DC increase to 30m has won. There are few possibilities how to move ahead so I am going to list them here.

  1. Increase DC to 30m, but also increase SF to 10%.
  2. Create YFI-B with a LR of 225%. Such vault type could have another 20m DC applied with a SF closer to what it is now (4%).
  3. Whitelisted vaults? We haven’t heard from the SC team if this is doable in a short term and also what it implies for Maker offering such products.
  4. Leave DC at 20m and increase SF to 7%. Some might feel the whole situation is too risky to additionally increase DC.

Note that despite the outcome here I am going to propose to increase YFI-A SF to 7% when we have a Rates Group meeting next week.

Which step should we take forward regarding YFI vaults?

  • Increase DC to 30m and SF to 10%
  • Create YFI-B with 225% LR / 20m DC / 4% SF
  • Whitelisted vaults when available
  • Leave DC at 20m and increase SF to 7%
  • Other - Comment below
  • Abstain

0 voters

Poll will remain open until this Friday December 4th and the winner will proceed to on-chain poll next week (including proposed SF increase to 7% for YFI-A if options 2 or 3 win)


I think that the recent continuous increase in the prices of BTC and ETH will generate price risks, and the stability fee (ETH\WBTC\YFI\LINK\BAT) should be appropriately increased to hedge the risks caused by price increases.

30M / 5% SF. No need to complicate and the 4%-6% looks fine to me. Would we be having this vote if the 11.5M user split their debt over multiple vaults?

EDIT: SF percents

1 Like

While the vault would get liquidated after a 14% price drop, you’d need another 43% drop for the liquidation to resolve at a loss. 51% total drop.

Binance YFI/USDT has more than $100m in daily volume. Coinbase YFI/USD has more than $35m in daily volume.

Raising the SF for the YFI-A vaults to 7% sounds like a good idea to make the interest rate more in-line with competitors. Makes sense to raise the rate closer to a market rate when the risk team doesn’t want more than 20M or 30M exposure for YFI-A.


30M and 6%. We have to remain competitive. We can underwrite the risk with stability buffer increases if we need to.


As I read the comments I am not sure everyone understands the option 2.
Option 2 is two vault:

YFI-A at 7% 20M with 175%
YFI-B at 4% 20M with 225%

So 40M in total.

I believe the reason behind this, came from a previous post.
Basically with 40M we hold 10% of YFI.
Even if you have a good turnover in terms of liquidity. You don’t dump 10% of the whole supply with out moving the market. Especially when the market is already down. The idea is first to decrease the risk and smooth the dump by having a spread on the dumping via the LR and the different vault.

1 Like

I think doing both 1 and 2 makes sense in the short term. I would love to see whitelisted vaults start to develop in the coming months.

1 Like

I was just informed by the @Smart-Contracts that it will not be possible to implement another YFI-B vault type by the end of this year. Which means the only other option in short term is to increase YFI-A DC to 30m and raise SF, if increasing debt exposure is preferred (so far it seems low LR vault type option has same amount of votes as 30m DC / 10% SF option).

I agree that 10% SF may not be competitive enough. I am simply stating the numbers from our model, but of course our model doesn’t predict exact numbers since it is based on assumptions. If we are confident that few whales who are keeping low CR can avoid liquidations successfully all the time, risk premium is lower than 10%.

On the other hand, the Rates Group can still propose lower SF next week and there is no rule that states SF needs to equal Risk Premium all the time, since RP mostly helps with guidance and is only estimated. We have actually violated this (under compensation of risk) few times this year when we kept fees very low.


In that case I would support an increase in the DC of 10m at a 7% rate which would be slightly lower than aave so we will be in the market and also protect ourselves from the increased exposure we would be assuming


I think it would be good to start thinking about risk tranches. Open up a secondary vault with the same LR but a higher SF. Would that work from a risk perspective or does the whole supply have to be taxed the increased SF on a higher DC?

A second vault with the same LR but a higher SF doesn’t make much sense.

1 Like

Why? First users get priority on lower fees while expanded supply costs extra.

I think it’s okay to be non-competitive if we’re uncomfortable with the risk. If half the vaults close their position, we’re left with 50% of the exposure and roughly the same amount of stability fees – seems like a win/win until the smart contracts team can potentially create a B-type vault to reduce liquidation exposure.


Yes, it would be a much fairer option since original vault owners don’t have to see their rates increased due to new vault owners, it would also enable more stable demand (probably). Not quite sure how complex it would be though in the short term.

Keep in mind, that while “Enterprise” is pushing heavily for whitelisting, it’s big players who are increasing systemic risks (coupling more and more protocols), not retail.

Dai is very important part of the defi ecosystem, and ideally, we would not treat this project strictly as business. Yeah, i am naive.


Good point.

1 Like

I don’t think we want to send the message to borrowers that we are uncommited to providing a competitive variable rate. I think its preferable that we subsidize the risk of YFI vaults via the stability fee until we are able to deliver a YFI-B vault type.

I would caution against this thinking. The risk premium charged should be inline with the underlying risk (to in effect insure against losses, thus the DAI minted is “risk-less”. Market competitors should help guide what is sane, but the underlying risk premium determination should be isolated from monetary policy and other lenders.


Who can I speak to about white listing our vault?

1 Like

You’ll have to submit a MIP application of some kind. Not to put him on the spot, but @charlesstlouis may have a good idea on how to progress?

1 Like