There has been some great discussion recently around adding new features to the DPI to generate additional yield or make the asset more “productive” for DPI holders and the Coop. Wanted to add onto some of the thoughts discussed and propose a framework to evaluate potential upgrades to the DPI
It appears the community agrees that a productive asset should:
- Offset the streaming fee for DPI holders (or even make it negative)
- Cover the opportunity cost for professionals to hold / trade / earn additional yield with DPI
- Add value to the Coop
What does productivity mean for the DPI?
- Generating yield directly in DPI. This means substituting component tokens for yield generating representations (e.g. SNX -> aSNX for 4% APY, YFI -> yVault, UNI -> cUNI etc). The additional yield natively would likely offset all of the annual streaming fee for DPI holders.
- Enable integration with lending / derivative protocols which enables shorting, leverage, borrow/carry trades. This is very interesting as it allows traders to hedge DeFi positions, create more opinionated market positions (long YFI, short DeFi), use DPI as collateral to borrow stablecoins to farm, or carry trade. E.g. Traders borrow DPI -> redeem DPI to underlying -> yield farm with underlying (stake in KNC, deposit in yVault, deposit in any random yield farm) -> borrow rate goes up -> lending APY goes up for DPI depositors. Depending on the spread between borrow and lending rates, the APY for DPI depositors in lending protocols may be pretty close to generating yield natively. Basically letting arbitrageurs doing most of the work to generate yield
- INDEX holders become a valuable participant in governance of underlying assets
Turns out there’s some good material already in traditional markets to draw some inspiration
- Many traditional indexes already lend out constituents to hedge funds to earn additional yield
- However, it is interesting to note that the SPY, the most liquid and well known index does not engage in lending activities, so definitely not a requirement to attract additional capital
- Additionally, index providers like BlackRock, Vanguard exert significant influence over its index constituents through their voting power. These companies have a special interest in shaping management decisions that can impact performance over the long run. And if components are lent out, that diminishes the voting power
So given that, these seem to be the biggest considerations around the DPI upgrade proposals:
- Risk in terms of getting more integrations. Considerations include how liquid are component tokens, compounding smart contract risk etc
- Yield profile. The additional yield that is unlocked from the upgrade
- Value accrual to the Coop. The additional value this upgrade brings to the Coop, in terms of fees, governance votes etc
- Cost of implementation. The time and cost of implementing these changes both from the technical perspective and operational.
Here’s an example of the framework being applied to several options discussed by the community previously. The end state may be implementing all these options, but this may guide us in the community to tackle the lowest hanging fruit first and go from there.
|Upgrade||Risk (for integrations / listings)||Yield||Value accrual to Coop||Implementation Cost|
|1) None (Keep DPI vanilla)||Low - Only Set contract risk. Low risk of being securities. Low risk of insufficient liquidity||Low (Medium - if used in a lending protocol).||Low - 70% of 0.95% streaming fee is revenue to INDEX holders||Low - Operational effort to get added to lending/derivative protocols|
|2) Add Yield tokens to DPI (e.g. aSNX, cUNI)||Medium - High - Compounding smart contract risk, unclear on difficulty to liquidate in lending protocols especially if utilization is close to 100%. Higher gas costs||High - yVaults etc||High - Streaming fees and % of yield farming fees to Index||High - Enable cTokens, aTokens, yVaults. Additional complexity for other DeFi protocol integrations|
|3) Launch new DeFi Accumulator Index||Medium - High - Same as above. But there will always be the vanilla DPI option||Medium||Medium - High - Potential cannibalization with the vanilla DPI||High - Same as above + operational cost|
|4) Enable non-staking governance (COMP, AAVE, UNI vs YFI, MKR)||Low - Same as 1||Low (Medium - if used in a lending protocol)||Medium - Same as 1, but given enough DPI AUV, INDEX holders can positively contribute to underlying component governance which indirectly benefits DPI holders.||Low - Medium - Governance module with COMP / UNI / AAVE wrappers + ongoing integration efforts|
Based on the framework above, my proposal is for the community to consider option 4 (enabling non staking governance) to be the first upgrade to DPI:
- DPI is already one of the largest holders of YFI (#14), AAVE (#16), MKR (#33), COMP (#62) and can already exert influence in voting for the benefit of the Coop. E.g. INDEX holders vote to include DPI as a collateral in AAVE or Maker -> more distribution -> more DPI demand -> more supply minted -> more governance power. This could kick off a governance flywheel
- It is still unclear the risks of adding staking to the DPI without a more formal economic analysis. e.g. if DPI grows to $100m and contains cTokens that are 100% lent out, how will external DeFi protocols that integrate the yield version liquidate the DPI? Does adding yield tokens present significant risk for larger exchanges to consider adding DPI
- Non staking governance does not introduce additional risk to external protocol integrations, and we can go full steam ahead on our integration efforts there
- DPI will become one of the largest holders of DeFi tokens (@overanalyser post). Smart decision making by INDEX holders in underlying protocols will drive additional value to the underlying protocols that eventually flows up to the DPI and Coop. This is longer term focused than immediate cashflows. Imagine the Coop, a long term oriented organization, holding significant influence in underlying governance tokens similar to Vanguard / Blackrock in their index constituents.
- From a technical and operation perspective, cost is low relative to implementing yield tokens. E.g. Forking Yearn’s or Curve’s governance contract etc
Happy to hear everyone’s thoughts on why or why not or if this is the right framework to evaluate the different options!
Here’s a poll
- Prioritize adding a governance module
- Do not prioritize adding a governance module