# Rocket Pool Tokenomics 2.0, Pt 4: # Looking Forward -Wander, 2022/03/19 (vvander.eth) ###### tags: `rocket-pool` `research` --- In [part 2](https://hackmd.io/@VVander/RPResearchpt2) of this series, I discussed the active areas of research problems for Rocket Pool, and in [part 3](https://hackmd.io/@VVander/RPResearchpt3) I discussed existing proposals. In this final part, I'll first outline a philosophy and then a specific suggestion for future RPL tokenomics changes. ## Tokenization Philosophy Every protocol needs a way to fund development and participation. Tokens are a good way to use investor funds for this purpose until the protocol matures enough to generate revenue on its own, but for this to work, investors must be repaid by eventually directing value towards that token, either through valuation increases or exclusive inflation. This concept, called tokenization, is a unique feature of web3 which helps monetize public goods and acts as a powerful incentive for participation. I'm sure everyone can think of at least one project in this space which hasn't released a token yet and consequently receives endless "wen token" requests. ### The Role of Inflation Many protocols, including Rocket Pool currently, drive value to their investors via inflation, which mostly rewards protocol participants. These participants must also be owners/stakers of this token in order to prevent total loss of token value. That is, while the sell pressure from existing participants may exceed the buy pressure from new participants, a non-zero buy pressure keeps the token valuable enough to utilize for rewards. The easiest example of this mechanism is traditional fiat currency, for which nominal inflation acts as grease for keeping the economic gears moving (incentivizing usage instead of hoarding). The output of this engine keeps the currency valuable enough for the purposes of exchange and leads to a healthy and growing economic engine, so long as inflation remains low enough that the currency is devalued at a rate only noticable on the decade+ timeframe. Another, more crypto-specific example of this design can be seen in [the Pocket network](https://docs.pokt.network/home/v0/economics), which -- unlike Rocket Pool -- has chosen higher inflation rates to incentivize fast growth at the expense of its token price. The speed at which this high inflation devalued PKT led to instability in the system which required larger tokenomic changes to resolve. From the [Rocket Pool Explainer Series Pt 3](https://medium.com/rocket-pool/rocket-pool-staking-protocol-part-3-3029afb57d4c): > While some may find inflation controversial in a space surrounded by fixed-supply tokens, this is absolutely necessarily for any protocol that needs to reward its participants for working in the protocol’s best interests. **Without generating value, that value would need to be taken from somewhere else in the protocol**, creating an imbalance where some participants are rewarded at the expense of others — not a viable, long-term strategy for any successful dApp or protocol. The above quote from Rocket Pool's founder, Dave Rugendyke, succintly demonstrates the inflationary tokenomics issue at play: value must either be 1) generated by the protocol through growth and distributed via inflation or 2) shuffled around inside it to reward RPL holders at the expense of other protocol participants. While we can encourage value generation through incentives or other actions, this is generally the responsibility of the pDAO and sits outside the scope of a higher-level tokenomics analysis. As such, I'll focus on exploring the second option (value shuffling) in greater detail below. ### Protocol Value Shuffling First of all, there is currently only one fundamental value stream for Rocket Pool, rETH commission. All the value that the protocol generates for itself and its participants comes down to the revenue generated by taking a cut of staking returns on the ETH "lent" to the protocol by rETH holders. This commission is fully sent to NOs, who must hold RPL. NOs only receive some inflation rewards, not all, which creates a tax on NOs through which the protocol can redirect value to other participants (pDAO & oDAO). Any value shuffling which occurs inside this system will be a zero-sum operation, so the only reason to do so would be to reduce an overpayment. Let's examine each of these categories with that mindset. ### NO RPL Rewards NO capacity is a limiting factor for growth, so reducing value accrual there will lead to lower growth. While there may be a maximally efficient payment rate, during the growth phase of Rocket Pool (i.e. < 22% of staked ETH market share), the protocol is best served by growing as quickly as possible. Therefore, the best strategy here is to keep NO payments as high as possible until a maximum healthy market share is reached, then reduce RPL rewards for NOs until equilibrium is found. ### pDAO RPL Allocation pDAO contributions are currently minimal, but this may also turn out to be a crucial value stream for the protocol's success. For one, the pDAO is likely the only way to generate liquidity incentives so that rETH (or RPL) can see wider adoption, but the pDAO funds will also likely be used for other important but less-critical expenditures like marketing or ecosystem grants. Realistically, there is a maximally efficient expenditure rate for each of the pDAO's initiatives, so we must be careful not to over- or under-allocate resources. The best strategy here may be the opposite of NO reward adjustments: start small and ramp up slowly to study the effects and find an efficient equilibrium. ### oDAO RPL Payments I believe the final value stream, oDAO payments, is the most reasonable target for strategic reassessment in the short term. To see why, let's examine the three justifications for oDAO payments typically used so far. 1) Currently, the oDAO is a technical necessity and must be compensated effectively 2) Higher oDAO payments mean members are more incentivized to actively support Rocket Pool 3) oDAO payments provides ecosystem funding through prominent member payments The first of these is inarguable for the moment, unfortunately, but on its own, this doesn't justify more than minimal compensation. As for the second argument, while some postulated before Rocket Pool's launch that oDAO members would be incentivized to promote and contribute to Rocket Pool, it seems clear that, with the notable exceptions of the development team, @superphiz, @Butta from beaconcha.in, and @Yorick from Cryptomanufaktur, few oDAO members are actively and significantly contributing to the protocol. In fact, there is no governance and community engagement from at least half of the oDAO members. In some cases, oDAO members have even remained neutral enough to continue promoting the protocol's competitors, which certainly isn't a promising sign of the oDAO payments fulfilling this secondary goal. Ultimately, even if these payments are somewhat effective in incentivizing active support, we must also ask: is the oDAO an appropriate mechanism for this? It may be better to create more specific grants for this. Finally, while it's true that critical ecosystem projects are oDAO members (e.g. beacon chain client teams), many current oDAO members are not public goods. Consensys, for example, is a for-profit and private company. What's more, the permissioned nature of the oDAO presents challenges in fulfilling this vision. Becoming an oDAO member requires approval of existing oDAO members, and we shouldn't be conflating oracle duties with other responsibilities if possible. oDAO members should not be evaluating specific public goods projects to decide which are deserving of funding, as this is ultimately the wider ecosystem's responsibility. #### oDAO Payment Analysis After settling arguments 2 and 3 above, let's focus again on argument 1. The oDAO is a necessary service to the protocol, but its service is binary (it either works or doesn't), so the effectiveness of this service doesn't scale with increased compensation. Since the oDAO has not had significant impact on protocol beyond simply sustaining its existence, we should therefore treat it as a fundamental protocol expense by the DAO, meaning we must aim to eliminate or diminish the role of the oDAO as much as possible. As discussed in [@jort's excellent inflation analysis](https://dao.rocketpool.net/t/more-equitable-and-effective-distribution-of-rpl-inflation/420), 135,000 of the approximately 900,000 RPL annual inflation is allocated towards the 14 oDAO members. Using a relatively high network activity average of 75 gwei gas prices, this provides a profit of 8,761 RPL per year per oDAO node. Even during today's bear market prices of $20/RPL, that's ~175k DAI per year in pure profit for each member. Is this total oDAO service payment of 135k RPL an appropriate amount? Would the protocol be more efficiently served by treating oDAO payments as an expense which should be lowered? If so, RPL inflation could go to other protocol actors like NOs and the pDAO which can make more efficient use of these funds. #### oDAO Membership Analysis It's worth noting that the more passive oDAO members are all organizations. The active contributors, @superphiz and the dev team members, are individuals. Is it possible that the oDAO rewards are high enough to be meaningful incentives for individuals, but not for organizations? Would an _increased_ oDAO reward structure more appropriately incentivize these organizations? Perhaps, but I think it's best that contributions should be incentivized through other mechanisms like grants. However valuable high-level strategic alignment may be, it is unfortunately too nebulous of a concept for which a public good protocol can budget. ## 2.0 Tokenomics Proposal: Inflation Adjustments As we saw in part 2, sell pressure from RPL inflation is being outpaced by growth. This is great for the long-term prospects of RPL value, but it's problematic from a growth efficiency standpoint. Therefore, while we must wait for a post-withdrawal world for truly accurate data, we should be ready to begin experimenting with inflation rates as soon as withdrawals are implemented. To do so, we need to begin designing these experiments today. With that in mind, I informally propose that we make the following changes to RPL tokenomics. ### Variable RPL Rewards for NOs First, I propose a **variable RPL rewards rate for NOs which is dependent upon the amount of new effective RPL staked in the last rewards period (28 days)**, using this equation: `rplRewards = newRplStaked/expectedGrowth * expectedYearlyInflation/daysInYear * daysInRewardsPeriod` `rplRewards = newRplStaked/125000 * 900,000/365 * 28` `rplRewards = 0.552 * newRplStaked` With this inflation rate, we can establish a rate that remains broadly similar to existing conditions, where ~125k new effective staked RPL in each rewards period leads to the typical 5% RPL annual inflation. On the other hand, if 0 RPL is staked in a rewards period, 0 new RPL rewards are distributed. If ~250k RPL is staked, the rewards would double, etc. This issuance strategy provides the benefit of directly incentivizing network growth while keeping inflation minimal in times of low growth so as not to unnecessarily dilute RPL value. This variable payment should not affect pDAO or oDAO payments, which can be re-evaluated separately. ### pDAO Inflation Solidification Since by the time withdrawals are implemented, the pDAO will have existed long enough to gather appropriate data, we should aim to determine maximally efficient pDAO funding rates for liquidity incentives, marketing, etc. Like with the oDAO, this minimizes overpayment for services. Once these rates have been determined, we can solidify the pDAO inflation rate and either redirect "extra" pDAO RPL to NOs instead or keep it in the treasury for future use. ### oDAO RPL Rewards Adjustment As mentioned above, we should **reform oDAO inflation to minimize overpayment**. I propose a variable compensation package which will repay gas costs required to perform oDAO duties as calculated retroactively for the previous rewards period and provide a lowest-necessary additional incentive. Under this framework, oDAO members are required to pay gas costs on their own and then later repayed by the protocol at a rewards checkpoint, plus 2% as compensation for the protocol "borrowing" these gas fees for 28 days. For the additional incentive, we can calculate compensation for oDAO member server administration through the maximum "likely" sysadmin salary range ([~300k DAI/yr according to Glassdoor](https://www.glassdoor.com/Salaries/sysadmin-salary-SRCH_KO0,8.htm)). Assuming an extremely generous 5 hours per week of administration time, that's about 37,500 DAI/yr per year, or ~2,900 DAI every 28 days. At today's RPL prices, this is roughly a 6,875 RPL reduction in oDAO payments per member, or 96,250 RPL less inflation per year, which represents about a 10% reduction in inflation. This previously excessive inflation can instead be used to fuel protocol growth via increased NO RPL rewards, sent to the pDAO treasury, or simply eliminated for efficiency's sake. ## Conclusion As we've discussed previously, Rocket Pool has a number of interesting tokenomics challenges ahead of it based on the current design. The tokenomics changes outlined above would position the protocol for a more sustainable future where inflation is managed effectively regardless of network or macro conditions, but this is, of course, just one of many possibilities for tokenomics reform. Hopefully this sparks some important conversations among the community about the future of Rocket Pool tokenomics. What are our overall goals for tokenomics? How do we improve the fairness and efficiency of RPL tokenomics? What about more creative tokenomics reforms like RPL burns or additional tokens? These questions, of course, are left as an exercise for the reader.