[Proposal] Update for DeFi+L Yield Strategies

Summary
The DeFi+L Pie uses the novel Pie Vault architecture to bring together eight blue-chip DeFi tokens for investors. The Vault allows for the individual tokens to engage in yield-bearing strategies while being part of a broader index. The current yield strategies are heavily reliant on lending via the Compound and Aave platforms, but the APY for these strategies is very low, usually less than 0.5%. I propose to transition the underperforming assets away from lending to vaults at Yearn.Finance to improve the yield.

Background
Let’s take a look at the current asset APYs in the DeFi+L Pie:

Sushi is currently the best performing asset by a wide margin due to the unique xSushi staking strategy. Aave and MKR have no yield bearing strategy currently, while the other assets range from 0.00%-0.29% APY with their lending strategies.

Yearn.Finance advertises itself as a ‘yield aggregator’ and specializes in low-risk yield strategies which commonly involve lending optimization or single-sided staking. Yearn vaults do not typically offer the highest yield options for a given asset, but importantly they offer the best yield among low risk strategies.

Specification
I believe five assets in the DeFi+L Pie, if transitioned to their respective Yearn vaults, would significantly outperform their current lending strategies.

  1. Aave currently has no earning strategy as it cannot be borrowed natively on Aave markets. The yearn Aave vault fluctuates in APY but generally has performed at around 2-3% APY, and recently the debt limit of the Aave staking strategy has been increased, which is why the APY is trending up to 5.56% today as of this writing.

image

The Aave vault is currently using the following strategies:

  1. SNX is currently earning 0.29% through a lending strategy on Aave. The yearn vault’s APY has recently increased significantly, related to the 1 year locking period of SNX staking, and now the vault is able to routinely harvest staking rewards.

image

The SNX vault is currently using the following strategies:

  1. YFI is currently earning 0.00% through a lending strategy on Aave. The yearn vault recently performed very well due to a staking strategy on LeagueDAO which is no longer active. There are currently no high yield strategies utilizing the YFI token so the APY will continue to fall over the coming weeks. However, YFI is in the process of upgrading their tokenomics and a transition to an xYFI token with revenue sharing is anticipated. There is some thought among the community that the current yvYFI vault could be rolled into the xYFI implementation when it goes live.

image

The YFI vault is currently only using the MakerDai strategy:

  1. UNI is currently earning 0.15% through a lending strategy on Compound. The Yearn vault is currently utilizing a staking strategy through Vesper Finance though this is likely temporary. This is a lower yield strategy than is available to some other tokens, so the harvest is typically once per month, which you can see on the below graph:

image

If the Vesper strategy becomes no longer viable the vault would probably transition back to the MakerDai strategy.

  1. LINK is currently earning 0.00% through a lending strategy on Aave. The Yearn vault is currently employing a higher yield (though likely temporary) strategy via LeagueDAO staking. The joinwido data is inaccurate so was not used here, but the yields can be seen below:

The COMP Yearn Vault does not have any yield-bearing strategy, so I haven’t considered that token for transition at this point. MKR does not have a Yearn vault, and the yield on Aave is 0.00%. It should be noted, however, that Aave liquidity mining incentives on MKR could make lending the asset reasonable to consider for the 0.73% APY:

This would require active management from PieDAO however, to claim the stkAAVE rewards and then either convert to MKR or AAVE.

Benefits
The current total APY of the DeFi+L Pie is 1.01%, which is mostly from xSUSHI:

image

I recognize that if the DeFi+L assets are moved to Yearn vaults then the vault yield would be slightly diluted, and therefore lower than is currently calculated. However, by rough estimate I think the blended APY would be ~3.32%, which represents a more than 200% increase.

Risks
Additional smart contract risks associated with integration into a different protocol
Additional gas fees associated with minting new Pies that involve assets in more complex strategies
If a yearn vault has 100% of the assets deployed into a strategy the gas cost to withdraw is significantly higher, so if attempting to redeem a pie for the underlying asset the gas may be exorbitant.

Sources
Images and data used in this proposal have been gathered from the following sources:
Wido
Yearn.Watch
Yearn.Vision
Aave

If the sentiment for this proposal is positive I will work with PieDAO leadership to move to a formal PIP. Please share your thoughts and comments.

5 Likes

Hey! Appreciate your time for looking into this.

Here’s my 0.0000001 Eth thoughts.

Basing on what you presented, coming strictly from an APY generated POV, all the other assests ‘could’ be categorized as underperforming.

Which begs the question if it would make more sense / better use of time & effort to concentrate on reviewing weightings and rebalancing/phasing out and incorporating new assets?
im not talking about a complete overhaul.

Link currently takes up 27%. Though there is no doubt about its utility across in the ecosystem, i wonder if taking up more than a quarter of the allocation is justified. This might however be due to the role of market cap being one of the methodology of Defi+L and its wieghtings?

Also, what kind of narrative do we want to put out for Defi+L? Is it just a basket of blue chips + any APY generated just an added bonus?
Or a basket of blue chips that is somewhat propotional to both?

Or this might not matter at all if the BasketDAO merger goes through.

Thanks for sharing your thoughts. Two things:

I think this proposal should remain separate from a rebalancing or change of assets. Personally I don’t have much problem with the current distribution of the DeFi+L assets, though I think the DeFi+S assets are past due for an update and I will begin a proposal for that when I have time. I think one of the admins mentioned they were thinking about a unification of the two products as well.

Concerning the APY of the assets, I think the answer depends entirely on the tokenomics of the underlying asset. For tokens like YFI or MKR that do not have emission schedules I don’t think earning an APY is a necessity. However, for tokens like SNX or SUSHI that are inherently inflationary, if the pie is not earning yield to at least keep pace with inflation then the holders of the pie will experience dilution over time. I think this idea will be even more important when we discuss adding different assets to the pies, as many of the newer DeFi 2.0 tokens are highly inflationary at this point. As a side note, I think that we should find a way to integrate CRV into our DeFi Pies and that is obviously also inflationary.

PieDAO’s primary differentiator from other index protocols is the intrinsic productivity of its indexes. Yearn is a battle-tested, low risk integration. With that in mind, I’m strongly in support of this proposal.

The primary differentiator is a mixture of the mechanics of the pievault + strategies of the underlying assets that make up the index no?

A pievault comprised of assets with no stategy is akin to a pie with no filling?

i am not good with yield farming… thou i thought having SPELL in there might make sense as well… thou it might more belong to the DeFi + S part

Let’s put assets to work!

2 Likes

Go :radio_button:

Hey @Wave
Would be great if you could include which vault you suggest want to use and their addresses so that we can check for compatibility

Otherwise, the proposal seems on point

Sorry for that oversight. Here are the vault addresses:

  1. yvAAVE: 0xd9788f3931Ede4D5018184E198699dC6d66C1915

  2. yvSNX: 0xF29AE508698bDeF169B89834F76704C3B205aedf

  3. yvYFI: 0xdb25cA703181E7484a155DD612b06f57E12Be5F0

  4. yvUNI: 0xFBEB78a723b8087fD2ea7Ef1afEc93d35E8Bed42

  5. yvLINK: 0x671a912C10bba0CFA74Cfc2d6Fba9BA1ed9530B2

3 Likes