More equitable and effective distribution of RPL inflation

jort · 21 March 2022 22:19

Summary

The community is currently debating the best ways to use the funds granted from RPL inflation, particularly those allocated to the pDAO. There are reasonable requests for these funds to be used in ways that bring the greatest value to the protocol and are sensitive to the limits of these funds.

This proposal is aimed at creating greater equity in the allocation of inflation rewards, so that they can be proportioned to the most pressing demands for the Rocket Pool protocol and contribute to its growth and success.

I believe the most effective way to do this is to re-proportion, at least temporarily, part of the inflation rewards from the oDAO to the pDAO. I justify this with an historical look at how these the distribution of rewards came about and highlight the stark differences in RPL inflation between the various ‘pots’.

Disclaimer

This is not meant to be a critique of the usefulness and value of the oDAO. It is a wholly necessary component of the protocol, and the oDAO should (and I believe still will be) handsomely rewarded for its role in ensuring the security and success of the protocol.

I’m writing this on a plane, using my memory and rough figures. Please do suggest corrections for anything that is wrong.

History

RPL inflation currently 5% per year, which is ~900,000 RPL ($28.6MM). This was originally allocated between four identified pots within the protocol, which per year were:

Node Operator Rewards (70%; 585,000 RPL; $18.6MM)
Oracle DAO Rewards (10%; 90,000 RPL; $2.8MM)
Protocol DAO Rewards (10%; 90,000 RPL; $2.8MM)
Insurance Pool Rewards (10%; 90,000 RPL; $2.8MM)

This allocation prioritised NOs with the view that they were integral to the protocol and should receive the majority of rewards. Nearer to the launch, the team felt that the insurance pool was not, at least at the time, a necessary part of the protocol and the rewards from that pot were distributed equally to the pDAO and oDAO. Each would receive an extra 5% of inflation rewards, moving to the following distribution:

Node Operator Rewards (70%; 585,000 RPL; $18.6MM)
Oracle DAO Rewards (15%; 135,000 RPL; $4.3MM)
Protocol DAO Rewards (15%; 135,000 RPL; $4.3MM)

Distribution per ‘Capita’

I use quotes around capita as I am fully aware that these metrics, often based on ethereum addresses, are not representative of a single person. However, they are good approximations for the purposes of my argument. I respect that members of the oDAO are often institutions that represent many people and their value is beyond the sum of its parts – you can use your own multiples if you wish, but I’m not convinced any reasonable multiplier will change the thrust of the argument.

This is what the RPL inflation rewards look like per capita per year:

Node Operators:

585,000 RPL / 955 NOs = 612 RPL; ~$19,000

Oracle DAO:

135,000 RPL / 14 oDAOs = 9,642 RPL - ~9 ETH fees for duties @ 75 gwei = 8,761 RPL; ~$275,000

Protocol DAO:

135,000 RPL / 2574 rETH holders + 955 NOs + 14 oDAOs = 38 RPL; ~$1,200

This is crude, but I think it shows the inequity in the current allocation of inflation rewards. Many community members are making proposals for pDAO spending that are rightly scrutinised by the team and community for value for money, but I believe part of the constraint in spending is due to the disproportionate allocation of RPL inflation.

If we take the current NO RPL reward APR as fair and reasonable for their duties to the protocol (around 13% APR at present, but going down precipitously), I think the current rewards for the oDAO (550% APR on the 1750 RPL locked up) and over quarter of a million dollars per member is, frankly, better spent on bolstering the pDAO in the early stages of the protocol. I am unsure whether an argument could be made that this kind of compensation is proportionate to the needs of a well-funded pDAO during this critical early growth period. Further, I believe there is an assumption implicit in this equal allocation of rewards that the oDAO, as a part of the protocol, is of the same value as the many, many activities that come under the remit of the pDAO.

Proposal

I think the previous equal distribution of the insurance pool rewards between the oDAO and pDAO was inequitable, and the pDAO could make better use of these funds. For greater equity, protocol growth and effective use of funds I would propose the following:

Node Operator Rewards (70%; 585,000 RPL; $18.6MM)
Oracle DAO Rewards (10%; 90,000 RPL; $2.8MM)
Protocol DAO Rewards (20%; 180,000 RPL; $5.7MM)

This would change the oDAO per member rewards to 5547 RPL per year or ~$175,000 (after ETH fees incurred), which I believe still hugely respects the role of the oDAO in the protocol, but also allows us to allocate limited funds more effectively to the multiple use-cases of the pDAO.

Curious what community thinks and if different proportions sit better with them.

Wander · 21 March 2022 22:55

Interesting proposal! A few questions:

Why did you choose 10% to the oDAO? Seems arbitrary.
Can we view the oDAO duties like Ethereum views its security budget? I.e. least possible payment
What about expanding the oDAO? That would decrease the amount that individual members receive while keeping the same “security” budget.

cyberhorse · 21 March 2022 23:00

I too am concerned about the respective allocation of funds between the oDAO and pDAO. For a start we now have around 5 months of operation of the pDAO and I am unaware of a single project it has funded. How long is that denial of function to the pDAO going to continue and who/what is going to change it? I do not see the bulk of RPL holders having any say in this.

Secondly I am not even sure who the operational oDAO members are and what specifically are the duties they are contracted to provide to RP. THere are 14 listed on the RP web site but I understand that 4 oDAO slots are actually held by Rocketpool Pty Ltd which performs the dev function. So there are several oDAO operators listed on the web site that are not actually doing anything.

Of those that are listed arguably the client teams for Nimbus, Prysm, Lighthouse (but not Teku) are valid for oDAO support but I don’t understand how the oDAO payment is of benefit to them. Does it go the an organisation or to specific devs on those teams? There are other recipients whose contribution is a mystery such as Staked, Blockchain Capital and Blockdaemon. Who/what is CryptoManufaktr? Bankless has already been the subject of controversy due to their support for Lido and Consenys is already a 10% shareholder in Rocketpool Pty Ltd as well as being an oDAO member.

Originally I understood that oDAO members were going to be able to overcome any NO shortage by being able to create unfunded minipools. That hasn’t happened. If their initial benefit was to demonstrate that RP had the support of the Ethereum community then that objective is now well and truly in the past. Whatever they have earned is more like a one time appearance fee and I don’t believe it should continue forever at the expense of highly targeted projects that must be funded from the pDAO.

jort · 22 March 2022 00:17

The 10% figure is based on the historical % that it once was. You are right that it is arbitrary, however, my argument is more about the relative % to the pDAO. Having the oDAO at 50% of the pDAO rewards seems much more sensible, but it is still arbitrary.
I like this idea of least possible payment, although I have no ideas about how to go about this. The difficulty with this idea is how unlikely the oDAO are to be voting for minimal payment.
Expanding the oDAO is an urgent priority, but does not solve the problem of the greater need for the pDAO and stimulating protocol growth. The oDAO has 0% allocated for protocol growth, it is better placed invested in other functions at the present time. We can think about paying the oDAO more when this is a greater need.

jort · 22 March 2022 00:24

Here is an old medium article on the responsibilities of the oDAO: Rocket Pool — Staking Protocol Part 2 | by David Rugendyke | Rocket Pool | Medium

As I mentioned, they are a really important part of the protocol and I do not mean to imply they have no value, rather a lesser relative value to the pDAO. The breakdown of rewards in my post does not consider the costs of running an oDAO node, which can involve considerable ETH, and this would reduce the ‘profit’ on the headline numbers I’ve presented. I’ll see if I can edit this to be more accurate when I can.

I believe the unfunded pools could no longer be part of the oDAO possibilities due to the issues around the pre-launch exploit and these changes would need further audits. This might change in the future, and you raise the possibility of other avenues of funding for the oDAO outside of RPL inflation that can come from fees.

jort · 22 March 2022 00:52

Just edited the oDAO RPL reward breakdown to include the fees they incur as part of their duties. These are not insubstantial, but are very small in the scheme of things. I’ve opted to use a generous average of 75 gwei for this. Credit to @knoshua for the comment and information around this.