• General

  • Draft OIP# Deploy treasury FRAX & DAI to Univ3 FRAX/DAI pool for FXS farming?

Summary:
This draft proposal seeks to start a discussion about whether we should put an amount of treasury FRAX & DAI to work earning trading fees and FXS rewards by depositing into Uniswapv3 pool FRAX/DAI pool and staking with FRAX Finance for FXS rewards.

Background:
FRAX Finance recently released [FRAX/DAI Univ3pool farming] (https://app.frax.finance/staking#Uniswap_V3_FRAX_DAI) - this is currently paying 17% apr (not including trading fees). The apr can be boosted by both time locking (for up to 3 years at 1x per year) and by up to 2x by holding and locking FXS as veFXS (similar to Curve veCRV locking). Further information on the boosting mechanisms can be found in FRAX Finance's docs here.
FRAX has also recently released [gauge weight voting for veFXS] - this allows for veFXS holders to vote on the allocation of 16,400 FXS per day across whitelisted pools. OHM/FRAX is a potential pool where veFXS holder could gauge vote to allocate.

Uniswap V3 pools allow trading between assets within specific ranges - the FRAX/DAI only exposes liquidity providers to those assets in the pool and allows them to generate trading fees.

Currently we have over 7 million FRAX and 20 million DAI sitting unproductive in our treasury.

Abstract:
This proposal seeks to begin a discussion about an appropriate number of FRAX and DAI to deposit to start earning FXS rewards and trading fees.

The minimum period for staking in the FRAX Finance farming contract is for 7 days - I do not propose that we should lock our FRAX/DAI LP for longer than the minimum period.

I propose that we would retain the farmed FXS and lock it as veFXS - fully boosted locked veFXS currently earns 45% apr claimable as additional FXS. I propose that we would also use our veFXS to gauge vote the FRAX/OHM pair for FXS rewards.

Risks:
The author considers the passing of the proposal adds the following risks to the amount of FRAX and DAI deposited.

  1. Smart contract risk from the Univ3 pools and the FRAX Finance farming contract.
  2. Cross-pollination risk from Univ3 trading pool exposure - where if either FRAX or DAI has some failure traders will come and trade the failing asset in UNIv3 pool within the range until we are only left with 100% the undesired asset (either 100% FRAX or 100% DAI).

Motivation:
This would allow us to collect FXS rewards and trading fees on an appropriate amount of FRAX and DAI from the treasury.

Voting period: voting period would be for 5 days from the date of snapshot voting commencement.

For:
Deploy the TBD amount of FRAX and DAI to UniswapV3 and stake in FRAX Finance faming contract to earn FXS rewards.

Against:
Do not deploy FRAX to the Convex FRAX pool at this time.

The values in the voting below reflect a total amount for e.g. 500K FRAX/DAI is a total of $500K USD of value.

How much DAI/FRAX from the treasury should we deploy for farming FRAX/DAI Uniswapv3 pool?

A couple of question I have in mind.

  1. What would be the price range that we will be setting for the Univ3 FRAX/DAI pool, and what would be the fees that we'll be setting for this?
  2. If we retain the farmed FXS and lock it as veFXS to use as voting power to gauge vote the FRAX/OHM for FXS rewards, what's the incentive for other veFXS holders to vote for this gauge too? Does this mean that we'll be moving some of our FRAX/OHM Uni V2 LP to frax.finance locking too?
  3. What do the values mean for the choices? For example for 500k FRAX/DAI, Is it 500k DAI and 500K FRAX? or is it 250k DAI and 250k FRAX? Or does this depend on the settings that will be used in creating a Univ3 FRAX/DAI pool?

    I think this proposal is a good way to support our partnership with FRAX. This fits in nicely with our existing treasury assets and doesn't give additional exposure to centralized stablecoins like a 3CRV pool for example. My perception of risk is fairly low on this given it's primarily UniV3 plus the Frax staking contract. Based on the 24hr fees vs. TVL in the pool it looks like trading fees work out to roughly 2% APY but I imagine that will shrink once TVL increases. A few points of clarification:

    • My intuition for deposit size is to match total Aave allocation of $1M, so $500k FRAX + $500k DAI
    • Given that we tend to hold our allocations, it might make sense to lock the LP for longer than the 7 day period to get the boost
    • Given the above, do you know what size veFXS position we would need for additional yield on this size allocation? Would we be able to use veFXS to boost this FRAX/DAI V3 deposit and also gauge vote FRAX/OHM rewards?
    • Can you give us any insight into the primary motivation for Frax launching this pool?

      FluctusLux it looks like Frax has a recommended price range and fee tier (0.05%) to set in order to be eligible for staking rewards:

        Hey FluctusLux thanks for your feedback - to answer your questions:
        1) The range is set by FRAX Finance in order to be eligible for rewards (it is pretty close to $1 see @tex post below)
        2) So the gauge vote just provides rewards to the pool voted for - the pool we would gauge vote would be our own Uniswap v2 FRAX/OHM pool and then we could just eat the rewards allocated by our boost
        3) The amounts are meant to be totals for example 500K FRAX/DAI is meant to be total $500K USD of value (I edited to clarify in the OP)

          Hey tex thanks for your feedback - you make a good case for a longer locking I am coming around to it.

          1) So with the boost for Uniswap V3 FRAX/USDC: it is 4 veFXS to 1 FRAX (1 FXS locked for 4 years is 4veFXS) e.g. if we had 10K FRAX and 10K DAI in the pool, in order to get the full 2x boost we would need 40K veFXS (which is 10K fully boosted FXS locked for 4 years).

          2) With the gauge voting it is just something extra you can do with your veFXS whether you do it or not doesn't effect your LP positions - obviously you would want to gauge vote pools you are in but everyone in those pools gets their proportional percentage of the extra FXS allocation.

          3) In terms of their motivation as I understand they are trying to get more liquidity across popular trading pairs and will be adding more and more pools as they go along (for e.g. there is a governance discussion about adding FRAX/SUSHI pair)

            Thanks for the clarification @Mark11 .

            I voted for 2-3M FRAX/DAI as it represents approx 10% of the treasury, which I think is the perfect amount to put in for starters. Anything less imo is non-substantial, while anything more, is too much risk.

            I agree with @tex 's suggestion. If we're locking the LP anyway, might as well extend the locked period. I think its a non-issue and definitely less risk to explicit USD stablecoin exposure compared to putting the money in a curve Frax-3pool. At least with a longer lock, we stand to gain more veFXS. I imagine that we'll be swallowing FXS in perpetuity.

            I am curious what other benefits does veFXS give its holders, aside from price volatility, and the ability to gauge vote?

              FluctusLux Yeah so fully boosted locked FXS as veFXS:
              1) earns 45% apr claimable as additional FXS
              2) allows a boost to your yield farming pools up to 2x rewards - for Uniswap V3 FRAX/USDC: it is 4 veFXS to 1 FRAX (1 FXS locked for 4 years is 4veFXS) e.g. if we had 10K FRAX and 10K DAI in the pool, in order to get the full 2x boost we would need 40K veFXS (which is 10K fully boosted FXS locked for 4 years).
              3) gives increased governance vote 1 fully boosted FXS as veFXS gives 4x voting weight
              4) allows you to proportionally vote on allocation of gauge rewards to your choice of pools

                Thanks for the clarification @Mark11 . Definitely on board with this proposal. IMO we need to be more aggressive to low risk allocations.

                I would be in favor of a single sided provision of FRAX to the pool (understanding this would not qualify for FXS rewards) but would be strongly against any provision of DAI. Adding DAI would expose our treasury DAI to FRAX price - if FRAX were to lose peg, we would quickly sell all DAI and end up with FRAX only.

                Our treasury already has a strong interest in FRAX and, IMO, it would be imprudent at this point to increase exposure.

                  Thanks for the feedback JaLa - I discussed this risk in the risks section above - it also would work the other way with DAI, that is, we would be left with either undesired asset in the pool because of the mechanics. I think the best way to manage this concern would be determining an acceptable size rather than provision of single-sided liquidity - which exposes us to holding valueless DAI in the case of a DAI peg failure and not much in trading fees. I imagine there is some number, however small, that would be acceptable? I think that number is around 500K each side

                    I agree with @JaLa here, I don't see a reason to expose our DAI to FRAX risks, we lose the diversification aspect. I'd rather deploy FRAX separately.

                      FluctusLux

                      I had the same thought regarding 10% of treasury allocation to this pool, but due to our current RFV growth rate I voted for 3-5m FRAX/DAI. If we carry on with only linear growth in the RFV pool we'll cross $50m in about 2 months.

                        shadow @JaLa

                        Other than their own unique smart contract risks, I don't think FRAX and DAI off much (if any) diversity in net asset exposure or performance. They're both primarily collateralised by ETH and USDC. If either one of the those collateral assets falters, then every unit of FRAX and DAI in the Treasury is exposed to that same risk, whether they've been paired together in an LP or not.

                        IMO, for this reason, I think the proposal creates only a small amount of additional risk to the Treasury, and we should support it.

                          I think it comes down to, in my mind, FRAX being more risky than DAI today. We own an outsized % of the FRAX supply and don't think we should further expose our treasury to FRAX risk by co-mingling the assets. We love FRAX and have a great partnership but our treasury needs to be very careful in allocating assets. Agree that the risk goes both ways and that if DAI loses peg, we would end up with more DAI, but I view that as orders of magnitude less likely.

                          So, if the primary concern here is ensuring that we don't expose our DAI reserves to the various risks of FRAX then what about deploying this strategy and only using FRAX? This would mean we would need to swap half of the FRAX to DAI first.

                          This would obviously mean we would incur the slippage and fee cost when it is done. I wouldn't have originally proposed this idea for that reason, but the 333,333 FRAX+Convex OIP which has passed will incur the same costs to create the LP position. As such, maybe this is okay?

                          If so, that would allow the FRAX to be deployed at no risk to our DAI at some cost.

                          Write a Reply...