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 (@overanalyserpost). 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!
I actually agree with @richard 's analysis. Seems like the Set team and others are making good progress on conversations to get added to lending markets and other distribution such as centralized exchanges. Not product improvements, but these initiatives are much more likely to bring in many more new customers than adding yield tokens itself which are giving <2% APY rn. cUNI is the highest at 3%, cCOMP is 1.5% and aSNX is 0.3% APY. Don’t want adding yield tokens too early that might ruin our growth efforts.
So option 4 or even staying with option 1 makes sense to me. Option 4 seems low risk, low cost and adds value to the INDEX token. We are still doing liquidity mining, so we still want a high INDEX price and high APY, so more capital flows in. It also makes DPI very easy to integrate externally as collateral.
I think another option is if we get added to a lending protocol, we can create a 2nd product that deposits DPI, borrows stablecoins to farm, and buys back DPI, and generate yield this way for customers. This is basically the yETH vault. I would imagine yield here to be higher than depositing the underlying
Thanks for the feedback @Sebastien@overanalyser . Interesting point brought up by @michael about a yVault like option that uses DPI as collateral and borrow stablecoins to farm. That’s something that could beat all the yields generated from just simply lending out the underlying. The stablecoin farms on Harvest and YFI are each earning over 10% APY atm. Beats 0.3% on aSNX right now. It keeps the original DPI as a “primitive” which is important for integrations, and doesn’t cannibalize with a second product
I fully hope that others will build vaults (yearn, harvest etc) that use DPI as collateral, as that certainly makes holding DPI more attractive as a long term hold for DPI users.
However, any yield that we capture on the underlying assets will be on top of what ever other build on top of us.
If you hold MKR, you get the benefit of the MKR burn which reduces supply. If we take MKR in PDI and we place it in AAVE we get an extra 0.1% MKR per year. If someone then places the DPI into a future yPDI vault earning 6% then the holder get the benefit of 5% MKR burn, some 0.01% DPI increase due to the aMKR within DPI, and then 6% additional DPI from yDPI.
That is the beauty of composability, what value we capture is added to that provided by others. It’s not an either or proposition.
P.S. YFI staking or yYFI is pretty sustainable at 10% so that adds good chunk to the potential overall DPI income.
Agreed. But I’m happy to let YFI build the vaults for the moment. then we can concentrate on others. When we get bigger, and the yearn fees start to look pain full we can circle back. At the moment yearn has the key feature of price oracle into Maker so it’s very unlikely to get liquidated.
I don’t see how Option 4 brings yield to the tokens if we aren’t staking to earn the cash flows. Don’t we automatically earn dividends on MKR and COMP just by holding them already and UNI doesn’t get dividends yet?
From a governance perspective, what’s our actual objective in voting? Increase long term prospectives of the protocols? Voting selfishly? A case stuff of Yearn and Curve might be helpful since that’s the only example I know of with one protocol having significant voting power over another major protocol. Maybe we partner with Gauntlet eventually.
If we are going to be activist investors we should be morally as well as financially activist. E.g. you could say YFI’s waifu memes goes against our own Code of Conduct (nearly unanimous approval) so are we violating our own Code by participating in their governance and not doing anything about that?
I think having two separate indexes DPI and DPIacc is a good idea but too complicated for now. Sticking to basic yield like yGov staking and putting a minor % of tokens into Aave/Compound seems reasonable until we figure out the operations, economics, and code around managing both at once. By that time hopefully the DPI market is mature enough that the two options actually brings in more capital by people that want higher yields and DPIacc being easier to integrate since DPI will already be established in other protocols.
If we do eventually do a DPI/DPIacc split it would be interesting to consider only holding Synthetix assets in the vanilla DPI since we don’t plan on using the assets for anything. We would have 0 slippage, possibly get SNX back as a rebate on trading fees with their volume program, and maybe it would make it easier to build derivatives of DPI since the underlying is more liquid and all backed by onchain ChainLink oracle prices.
To add to the narrative, productive assets have a huge premium over those that are not (whether they are natively productive or can be integrated on others).
For example, WBTC (Set is one of the initial DAO members) didn’t get any traction the first year it launched in 2019. But after yield farming began and WBTC was accepted as collateral on lending protocols and farming protocols, it exploded in desirability because it allowed those holding BTC to get yield and thus make their asset productive (where otherwise it would be sitting on a cold wallet). Today, WBTC is $1.2B in market cap.
In addition, we are seeing the same thing with DAI, a decentralized stablecoin. The desirability of DAI is in its widespread integrations, adoption, and places it can generate yield.
Using this analogy, it is critical for DPI to grow widely it would need to be a productive asset either natively or through its integrations and usefulness in other protocols.
And the more productive we can make it, the better.
For me governance rights are a soft power. They can help build (or destroy) our reputation.
If you look at the recent UNI vote. They were within ~1 M votes of passing the resolution to change the “quorum” (net votes for) and proposal thresholds. DPI contained ~500,000 UNI. If we had voted that proposal, it may have pushed it over the edge. [In reality we couldn’t vote it as we didn’t have time to organised coop governance, set up the smart contracts, and votes had to be delegated / self delegated before the proposal went live…].
My road to 1% suggested that we delegate votes for the moment while we focus internally. Then we start looking at how we can use our goverance power to help others.
Yes we can use our power selfishly, current CURVE voters are disincentive from allowing others to be whitelisted. https://twitter.com/bantg/status/1318862473819205635?s=20 . However, I would tend to argue that long term, anything that benefits the members of DPI benefits INDEX. TO be seen to act selfishly when involved in others governance impacts our reputation.
If you look at Lesher, his proposal for UNI delegation rights builds a reputation as a partner in DeFi and wanting others to succeed. Will he vote against his own interests? I don’t know. But if he votes in a way thay people don’t like they will remove delegation rights. Leshner will protect Uniswap - Delegation Pitch - Uniswap Governance
DPI will become a super delegate in many protocols. 1% of circulating tokens is a massive chunk. If we abuse it, then some DPI holders will prefer to redeem.
Synthetic products already exist and grow. But 100% ownership of the underlying tokens (or yielding versions) is a different proposition. They have different risk profiles and uses.
I don’t fully understand SNX and synthetics. I get the principle, but the execution doesn’t speak to me. My SNX is sitting in xSNXa for someone else to mange.
Another question is how would a INDEXcoop index containing synthetic assets differ from one generated by SNX? Or is it simply a white label deployment? (which is not bad thing).
Synths only track the price of the asset. So none of the good stuff of owning the underlying tokens, lending yields, using it natively in the protocol, governance etc. You trade against the oracle price (ChainLink in this case) vs the market. SNX also has a DeFi index product here
Yea there are definitely tradeoffs like we wouldn’t have governance power with the synthetics but we could also theoretically have more money in a synthetic DPI than the summed marketcaps of all the underlying tokens since we aren’t constrained by token supply which means more fees to the treasury. We could supply infinite demand for the index without having to worry about the market being able to handle our movements and we still have the DPIacc that holds real tokens for governance power. It’s still possible to earn yields on synths if we wanted to e.g. sETH/ETH Uni pool and sBTC Curve pool as the most prominent examples.
Yea I agree that governance has a lot of power. My main concern is that if we don’t have an explicit goal/ideology around participating in protocol governance then we will waste a ton of time and energy in our own forums debating about votes in other protocols which doesn’t seem very productive. This is especially so when we have to contend with thousands of DPI holders who want their voice heard who don’t even hold INDEX. Sounds like a shit show imo and I’d rather avoid that headache until we have a clear framework.
I’m still undecided on this. I think my main questions on prioritizing governance are:
What are we prioritizing it over?
How many hours of work will it take to implement for all the protocols mentioned (UNI, COMP, and AAVE)?
What is our political agenda in governing these other protocols?
I don’t think all the additional overhead that comes after implementing this (governing how we govern) is worth distracting us from activities with higher growth returns.
Could be interesting, but synths are actually constrained by SNX market cap, since SNX stakers mint the debt at 600% collateral ratio. Limits the size of a synthetic to current market cap of 400m / 6.
Agreed. I see product improvements (upgrading to yield tokens, governance) as separate to growth initiatives (getting on lending protocols, derivatives, exchanges etc), and onboarding new methodologists which can happen concurrently.
Taking a stab at your questions:
Over options 1, 2, 3 as we get more community consensus around how to pursue yield, and clarity from external integration partners. The mint/redemption fee may be higher priority, but that’s outside the scope of this thread
Very low technical cost. 2-3 days for smart contract work. Then there’s sending it to audits etc
Up to INDEX holders (DPI holders have no input). Imagine passive indexes would be very long term oriented. Could even delegate votes to an external party first. Short term can even be a marketing play to strengthen the Index brand
True forgot about that. At the moment its limited to SNX but I think they’ve already started allowing ETH to mint synths too so it’d be at least SNX + ETH marketcap but still constrained so I guess that takes away the main selling point in my argument
