Request for Discussion: Activate Intrinsic Productivity (IP) for DPI as a Standalone Product

I still think that the vote should show clearly an option for DPI + staking rather than “2 products”

Regarding the risks of being considered as a security if IP is added, would it be possible to request a formal legal opinion?

I am still not convinced that adding IP in a “neutral” way (by using the “default” yield option that each project proposes, without an active strategy) would increase significantly the risk of DPI being seen as a security. This risk is not null anyway, with or without IP.

If really the risk is as high as described in @oneski22 comment, then implementing a staking contract would probably be the most balanced solution. A second product, on the other hand, would be a bad idea, actually deserving DPI and not solving the initial problem (how to ensure that DPI holders can capture the whole value of the underlying assets).

1 Like

I join @oneski22 in being adamantly against the One Product Approach. It puts our US-based centralized exchange listing at risk. Intrinsic Productivity within the DPI token itself unequivocally pushes it towards the line of being both a security and a derivative under US law. That’s a reading that has been confirmed by several attorneys at large firms.

We won’t get that clarity under the current regulatory regime. The most anyone can say without an SEC enforcement action is “it gets us closer to the line.” That said, two things could happen that would make me revise my stance:

  • Major US exchanges say they think DPI is too much like a security regardless of IP. This would make the point moot. It’s unlikely based on preliminary conversations with counsel and exchanges and the number of DPI components listed on US-based CEXs.

  • The SEC releases guidance or clarity that changes the equation This would be slow and depends on what’s put out there. The closest thing we’ve got is this proposal. And that doesn’t speak directly to the DPI/IP question.

Until then, I think this would be a step back for the Coop.

5 Likes

@oneski @Metfanmike Thanks for your comments and feedbacks over the course of IIP-29

  1. IIP-29 is intended as an initial stepping stone. Yes 50 bps is small but we are only using ~10% of the assets in DPI and some of the bigger tokens like AAVE and SNX have higher APYs available. Once IIP-29 is passed we can do more auditing and research into the best strategies for those assets and increase far above 50 bps.
  2. At a smart contract level i don’t think staking strategy is any different from 1-product solution legally since farming would still be done with assets in collective DPI vault. The two layer system turns vanilla DPI into a productive asset with a theoretical promise of yield through staking anyway right? Where as two product approach is two distinct pools of assets that are managed individually with no intrinsic link between IP and nonIP tokens.
  3. All the lawyers have to say is a strongly worded “I dont know”. It’s entirely fear-based instead of on facts. It also comes from a stance that centralized platforms will be the only option available to retail and institutional money which is the opposite, self-hosted wallets are increasing from new comers and they are interacting directly with protocols. You also say institutions wont put money into it but look at all the theoretically illegal farms/securities that major trading firms participated in during DeFi summer with no repercussions.
  4. Using the yield farming strategies selected, I don’t believe IIP-29 will delay extrinsic productivity at all. xSUSHI is already listed on AAVE, xSUSHI is ahead of DPI for MakerDAO collateral, MakerDAO has yearn vaults (including yvYFI) whitelisted for oracle access because they rely on their vaults for DAI generation so much, yearn vaults have been used as collateral on money markets before e.g. yCRV, etc. This was part of my consideration for selecting only these 2 strategies for IIP-29 and will be part of the methodology for selecting future strategies as well
  5. Optimizing for US regulation is shooting where the puck is, not where it’s going. Ethereum is already on track to settle over $10T dollars in 2021. If you look bigger/longer term we are sacrificing a better product to a global audience for an inferior product that targets a somewhat dying/lagging audience. The unclarity in US law could mean it eventually works in our favor if they say yield farming doesn’t make it securities and we’d have lost all that potential revenue for nothing. The opportunity cost is too high in my opinion but other people believe US > ETH as global dominance.
5 Likes

To me, the following questions remain unanswered from those in favor of a non-@oneski22 approach:

  • Why isn’t @oneski22’s vault option - which seems to satisfy all the folks against the proposed IP options - while still allowing for IP - a perfect compromise?

  • Why is BDI failing if this is so attractive?

4 Likes

@Kiba I appreciate you taking the charge of this issue.

  1. I think overall IP is incredibly useful and makes sense on many of our products, I just happen to think DPI is different given the way it has been pitched and marketed. SMI, SYI, and more which are targeted to crypto native folks make better candidates for the initial exploration of IP. I also would argue that the IndexCoop could and should partner with experts in the space such as Yearn for strategies if we are to do IP. Look at Badger for example, partnering with Yearn for the byvBTC vault.

  2. Not a lawyer or a smart contract engineer so would love to hear from the #dev team. When using a “Yearn style vault” which is what the staking contract idea was, the assets would be transferred to the staking vault prior to being farmed via the associated strategy, so there is a separation.

  3. The lawyers are not speaking from fear, they speak from analysis of the law, that is what they are paid to do. The reason we have not gotten more than a “We need to do more research but this is our thoughts” is that no lawyer will give a memo/opinion without paying them. Crypto lawyers are really expensive. Opinions can be upwards of $50k. The fact we have had DPI holders retain crypto lawyers and pay for advice at all, even if it wasn’t a formal opinion, should speak to the legitimate and serious concerns they hold. I do not disagree that self hosted wallets are going to be the future, but again 45M+ Coinbase users speaks to the current market we are actively trying to address. Major trading firms, like the major major ones (Goldman, JP Morgan, Morgan Stanley, etc) aren’t doing that form of degen trading yet. This also isn’t just US-focused for the law, we just happen to have US lawyers who are easily accessible. I think securities concerns are real in many jurisdictions. The Singapore opinion we currently have has a massive loophole “we assume none of the underlying are securities or derivatives” when there is a very valid argument that yvYFI and xSUSHI are derivatives. If the Coop wants to fund a multi-jurisdictional opinion which would be a very very large sum of money that is a possibility but I don’t think it is the best use of our resources.

  4. Can’t speak much on this, defer to others who worked on the AAVE/Maker proposals.

  5. This is bigger than US regulation. I agree with you that ETH will settle a lot and that it will be the global settlement layer. However, stepping back… for at least the next 12-18 months. Is the yield gained and the philosophical stand of having IP added to DPI worth all of the lost revenue from steaming fees that could be used to build the DAO? Is the yield gained and the philosophical stand worth missing out on the 45M+ users using centralized services? Do we ignore them and gate our products only to the DeFi native? I would argue no, as I feel like that is abandoning our mission of making things easy, safe and accessible. Should the global regulatory climate change we can change as well. Using the insights gained from SMI, SYI, and other products we won’t be far behind on knowledge and infrastructure. BUT for our flagship product that we are trying to get to everyone everywhere, I think adding IP would be a disservice to our real users.

3 Likes

On the opt-in and opt-out point I want to add another thing to consider. Users were sold on a methodology that we are implementing. Assuming they are responsible investors and did their research, explicitly in the methodology it states:

The token must be a bearer instrument. None of the following will be included in the index:

  • Wrapped tokens.
  • Tokenized derivatives.
  • Synthetic assets.
  • Tokens that are tied to physical assets.
  • Tokens that represent claims on other tokens.

IP would be a departure from that, which is within the rights of the Coop as we implement the methodology, but would be a really bad look to a holder, who bought something expecting a certain methodology, that was all of a sudden changed without them opting into the changes.

Any departures from the methodology should be opt-in in my opinion.

2 Likes

@Metfanmike I’m not sure I understand what you mean - what would you like me to clarify here?

image

On BDI’s Losing Market Share - I’m not sure the fact that it is losing market share to DPI means anything other than yield farming incentives can only go so far in attracting capital. From my perspective, DPI has legitimacy that BDI never will. Personally, I would not exchange my DPI for BDI even if it offered 100%+ yields. BDI’s anonymous developers, forking DPI methodology, and not using Set Protocol’s battle-tested infrastructure are all serious red flags. @oneski appeals to the value of maintaining trust as a reason to not implement IP via the one product approach:

1 Like

Not much to add on my part, I think this conversation has covered most of the concerns.

I’m personally against default IP implementation for DPI and we should probably focus on figuring out the way to implement it as an opt-in solution.

Although I don’t see much of a point in doing this for 50bps of yield. And I mean both solutions seem like too much work for 50bps. I think even if we used most of DPI’s assets, we’d get to maybe 2.5%-3% over the next 12 months which would be nice to have but certainly not mission critical.

5 Likes

This is definitely one of the most interesting forum threads I’ve seen! Really good discussion and expertise.

Adding my 2cents:
As much as I would like to see IP for DPI, reading all the risks associated with it combined with the effort and revenue potential, also raises concerns whether this is the right approach.
But as mentioned a few times above, I have high hopes for the “vault solution”. Isn’t this the silver bullet to our problem, as we don’t have to touch $DPI, but we can offer a way to activate IP?
Plus, since it’s an opt-in solution, making it easier to inform users about risks, we could probably also implement strategies that might have a higher risk/ longer lock-up periods (e.g. activating REN → see here).

5 Likes

First off, great work by @Thomas_Hepner and @Kiba putting in the groundwork and presenting these ideas. From what we already know and outlined again from @oneski22 's points is that there is a risk that implementing IP directly to DPI may jeopardize its investability leaving us with option 2 or some version of it.

I would just like to point out that there could well be a situation whereby institutions could lend their vanilla DPI through the defi market “legally” picking up a bit of yield. The defi native market could then borrow that and deposit it in the vault strategy to arb any additional yield. (Effectively a carry trade). We are probably some time away from this but this “regulatory arbitrage” is very common in tradfi. I believe we should be mindful of such future opportunities.

As @Lavi points out the vault strategy seems to be the silver bullet so keen to hear what the potential downsides are to this specific strategy.

5 Likes

Hi All,

I really like the idea of optimising how we utilise the tokens within DPI to generate additional returns. I am a big fan of this.

I have been pondering how to do this for while now and I keep coming back to the thought - Why pass on IP benefits to holders with our left hand and then takes fees from users with our right hand. If I am reading this right DPI generates 95bps fee revenue and with this proposal, DPI would give to holders 50bps. Holders essentially pay 45bps fee whilst DPI still generates 95bps.

I believe @LemonadeAlpha has done some interviews/research that indicates our holders are no particularly sensitive to the stream fee. With that in mind, I would be reluctant to think passing this IP onto holders is going to make any difference.

50bps in DeFi isn’t going to get people excited. It is nice, but really I don’t think it is sufficiently large enough to move the dial in terms of generating more units of supply.

I would however do IP and put the proceeds in our own back pocket. We can make some extra income on top of the streaming fees, that would be great for INDEX holders. I would rather see Index Coop implement some level of IP and think of it as a way to boost our income and the profitability of the product. This would enhance our margins, profitability and I don’t think it will impede any of the extrinsic use cases being looked at.

I think we can explore IP and think of it as a way to bolster our profit margins on the product.

3 Likes

So the problem with the vault model AFAIK with the way Set Protocol works is that the assets still come from the main DPI vault so there are no “user funds” it’s all DPI, so all DPI holders are exposed to risk you’re just giving rewards to the “opt-in” stakers but then everyone is incentivized to stake so there is no vanilla DPI or vanilla DPI holders have disproportionate risk/reward.

If you want to remove the assets in the vault from DPI then that’s a two product solution because we’re taking AUM from DPI and need another rebalancing operation.

3 Likes

@oneski22 @Metfanmike @Lavi @MrMadila Please see Kiba’s comment for why the vault model is not a feasible third option.

@richard @puniaviision Do you agree with @Kiba from a technical/implementation standpoint about the vault model?

1 Like

adding DPI to the iDPI vault (which is what I am temp calling it) would be a redeem operation [or the vault could hold a reserve of DPI, and unwrap prior to using strategies], moving the underlying from DPI to the iDPI vault), a withdraw remints DPI (swapping if needed to get the balance right) [or pulls from reserves].

This creates a firewall between the two. iDPI holders bear all risk of IP. DPI is protected. Legally and additional smart contract risk due to IP strats.

As for @Matthew_Graham’s point for getting revenue from this, we could again take the yearn model of 2% management fee and 20% performance fee for the vault.

4 Likes

@oneski22 @Metfanmike @Lavi @MrMadila @Matthew_Graham @fallow8 @trx314 @Kiba

Thank you all for your thoughtful commentary - the community sentiment polls are now live!

Edit for Clarification Purposes: The “vault model” described in the comments would be a subset of the two product approach (options 1a and 2a.). The vault model is intentionally not included in the voted as it needs to be vetted among other technical implementation options from an engineering perspective.

1 Like

Re a middle way vault style model.

In the past we have discussed the idea of allowing users to stake DPI, in return for receiving income from IP. here and here.

The staking makes it purely opt in, and allows only the tokens underlying the staked DPI to be farmed (but we can farm 100% of those), So it does solve a number of problems.

However, there are downsides:

  1. DPI holders will need to pay gas to stake / unstake - we are making holder become more active
  2. DPI holders will need to stake (likely time blocks > 30 days) or with a cool down period.
  3. Eng resources will be required to do the IP, staking contracts, possibly rebalances, calculate the share of income.
  4. Extrinsic productivity (Cream, AAVE, Alpha) isn’t available for staked DPI.
  5. IP on DPI just isn’t that large - how long do you need to stake to cover gas?

PieDAO DeFi+L share their IP data for a similar product:

To me, such approach is a lot of work, and isn’t likely to move the needle for coop AUM. So, the discusisons with @Thomas_Hepner and @Kiba came down to a single product with IP, or two separate products. (as the original post above)

4 Likes

Would be nice to have a voting option for those who still think we should not be moving forward with IP for DPI.

3 Likes

would have been nice to include, seems the question is more “if we are going to do IP, which method should we do”

Agreed. Without asking that additional question, one would likely conclude that the results indicate clear community support for moving forward now with IP and that it’s more a question of which proposed IP path is preferred. I selected the two-product approach because it’s the lesser of (what I see as) two problematic paths, not because I believe we need to launch a solution today.

That said, if the thought process here is to gauge community sentiment on the preferred IP approach and then put that to a separate vote - “Do we do this IP approach today or should we hold off for now?” - then this might be an unwarranted concern.

Is that the thinking, @Thomas_Hepner?

@Metfanmike @oneski22 Apologies for the delayed response to your comments - I was out sick the whole weekend.

The intention of this governance post was to gauge community sentiment for which version of IP should be implemented (one product vs. two product approach), if/when IP for DPI is prioritized. I do not think there is clear community support to move forward with IP now ahead of all other priorities.

Per @puniaviision , prioritization is determined by a Work Team Analysis after DG1 vote before moving to a DG2 vote.

@Metfanmike @oneski22 Does this address the concerns you both raised in your most recent comments?