# Rocket Pool 2.0 Tokenomics, Pt 1:
# ETH-Only Minipools (EOMs)

Wander, 2022/04/18 (vvander.eth)

###### tags: `rocket-pool` `research`

Rocket Pool has been live for about 6 months now, and in that time, we've seen quite a few suggestions on future protocol improvements. This series focuses on areas of research for Rocket Pool tokenomics and suggestions for the future. To start, let's discuss one of the most common suggestions: Eth-only minipools (EOMs).

## Naked EOMs

EOMs are a very common suggestion within the community, but this topic is significantly more complicated than it first appears. Many folks see the RP mechanics, notice that we're constrained by NO capacity, and think to themselves, "Maybe if we get rid of the RPL requirement, more people will want to stake!" If this is you, I'd like to first thank you for dedicating a little time and energy to Rocket Pool. The DAO needs your contributions, so please stick around.

As for the ETH-only minipool idea itself, this is an active area of research. Much of the community agrees that this could help increase our NO capacity, but there are some unsolved problems which remain, mostly with what I call "naked" EOMs: the idea of EOMs by themselves (i.e. no other changes to the protocol) usually using the minimum collateral of 1.6 ETH instead of RPL. There are variations of this which seem different at first glance, e.g. "mix'n'match collateral with both ETH and RPL as you like, but you only get RPL rewards if you use at least 10% RPL," -- don't let this fool you, though, because it's still naked. Any suggestion which allows for only using ETH as collateral instead of RPL with no other tokenomics changes attached is naked.

## RPL Devaluation from EOMs

The first problem with naked EOMs is that, ultimately, creating "naked" ETH-only pools means devaluing RPL, which is an existential issue for RP.

If naked EOMs exist, the only reason to stake RPL instead is to take advantage of a higher APY. One day soon, though, RPL APY will be equal to or less than ETH APY. When this occurs, if EOMs are possible, existing validators would be incentivized to exit, sell all their RPL, and re-enter to make EOMs instead.

Not only does this clog the queue and provide no true benefit, it also devalues RPL through sell pressure. This behavior would likely start in advance from pure anticipation that naked EOMs were coming at some point, which would continue as more people realized what was happening and sold their RPL, ultimately creating a cascading RPL devaluation which, in a worst-case scenario, could totally devalue RPL and remove virtually all rETH security.

RP does need a solution to this problem of RPL vs ETH APY regardless of EOMs (more on this in [Part 2](https://hackmd.io/@VVander/RPResearchpt2)), but if we introduce EOMs before solving the issue, it will exacerbate the situation. Remember, rETH security depends on the RPL/ETH ratio such that if the ratio drops, rETH security decreases. We can't change existing minipools, so we have to keep RPL value in mind when considering future protocol changes, and the consequences of naked EOMs on RPL's value in relation to ETH means that this isn't a viable path.

## Security Concerns with EOMs

The second issue surrounding naked EOMs relates to protocol security. Unless we are explicitly changing the security model (unlikely to get traction in the near term), all Rocket Pool nodes require 17.6 ETH collateral at minimum. However, current averages for collateral are higher: about 54% or ~7 "extra" ETH above the minimum collateral. Any proposal to change the protocol should consider the possibilities of how average collateralization could be affected and how that affects network security.

Naked EOM would certainly decrease network average collateralization since all EOMs would have the minimum collateral. There is no debate as to whether naked EOMs would decrease security through lower collateralization (they definitely would), and instead the debate tends to focus on whether existing levels of security are necessary. As of now, this is unresolved. To put it in Computer Sciencey terms, we need to study how lower average collateral affects security during protocol state edge cases.

This may end up as much ado about nothing, but you can't be too careful when it comes to protocol safety!

## The Social Inviability of EOMs

Yet another issue with naked EOMs is the resistance from RPL holders. Any weakening of the (admittedly articficial) utility that RPL has as required collateral for nodes will be met with skepticism. This is a byproduct of the same tokenization forces which are web3's superpower (more on this in a future installment).

Proponents of naked EOMs tend to see this idea as not competitive with the RPL token since, in their view, it's simply added functionality for the protocol, not a removal of anything. Under this view, RPL could continue to hold the same utility it does currently as collateral.

This is incorrect, however. To see why, let's examine the fundamental (not speculative) RPL value propositions and how they might change with EOMs:

1) Governance
With naked EOMs, those NOs can't participate in governance, so a lower percentage of NOs will be represented, but otherwise the value proposition doesn't change here. If you want to participate in governance, you have to own RPL.
2) RPL Rewards
RPL rewards act as an incentive for NOs to increase their RPL collateral, accomplished via inflation. If EOMs exist, average collateralization will decrease as discussed above, but otherwise the existing value proposition in regards to RPL rewards does not change.
3) Requirement for Node Collateral
Currently, if one wishes to be a NO, one *must* purchase RPL. If naked EOMs are allowed, the situation changes so that to be a NO, one *may* purchase RPL. This value proposition completely disappears under EOMs. Note this is an artificial value proposition, as this requirement is forced by the protocol and not a necessity for protocol operations or security -- watch for more on this later in this series of essays -- but it is still a value proposition. In the 6 months since the protocol launch, the RPL/ETH ratio has remained remarkably steady, which is evidence that this value proposition has a sizable effect on real market prices.

In summary, while the first two fundamental value propositions are unchanged, the third value proposition completely disappears with naked EOMs. While it's debateable as to the effect this value proposition has on the overall token value, it is undeniably a reduction in value. Therefore, no rational RPL holder today would ever want to see naked EOMs implemented. This, of course, comes as no surprise to those involved in the RP community, though outsiders tend to miss it, likely because EOM proponents are those who don't own any RPL and have a hard time imagining the mindset of those who do.

While this issue is extremely similar to the first one (RPL/ETH ratio devaluation), I list it separately because it's a problem with many EOM suggestions, not just the naked ones.

## Non-naked EOMs

As I've discussed above, there is at least one incentive model showstopper, one area of concern for security, and also a social impossibility for naked EOMs. But what about other flavors of EOMs? There have been a lot of suggestions, like *EOMs but the commission is lower than RPL-collateralized pools*, for example. Many of these suggestions help incentivize RPL usage to some extent but still don't solve the earlier problems entirely.

Although there aren't any serious non-naked EOM suggestions at the moment, here's what one might look like:

- Assume 5% APR on staked ETH
- EOMs are allowed but forgo 2/3 of commission (10 of the 15%), which is taken by the protocol as a fee and used to purchase RPL on the open market and burn it.
- If `x`% of the current ~5,400 minipools were changed to this model:
- 5400 * x% minipools; 16 * 5400x = 86400 * x% ETH
- 4320 * x% ETH of rewards annually (86400 * x% * .05)
- 216 * x% ETH for their owners at 5% commission (4320 * x% * .05)
- The other 10% of commission, 432 * x% ETH, is used to purchase and burn RPL.
- Inflation is currently ~9000 ETH worth of RPL per year, which would outpace EOM revenue significantly, so we must also eliminate RPL inflation completely.
- Currently, x% of the network represents ~5mx% staked RPL (~50,000x% ETH @ 0.01 RPL/ETH) which would be sold if changed to EOMs.
- Finally, we calculate a "time to profitability" for the network of `y` years: the resulting `sell pressure = burn rate` equation becomes `50000 * x = 432 * x * y` -> `y = 115 years`

How can we get that time to breakeven lower? By lowering the RPL/ETH ratio, of course. If the ratio were 100x lower, the breakeven time would be 100x lower, so just 1.15 years. So all we have to do to allow EOMs in Rocket Pool is completely destroy the current value of RPL. This isn't a great path forward, either, of course.

## Conclusion

Obviously, EOMs in Rocket Pool are untenable as of now, but there's a much easier solution which is more likely. Forking Rocket Pool and removing the RPL token completely at genesis is, by far, the fastest way to accomplish EOMs. Of course, this hypothetical fork also loses all the benefits of tokenization, so it's questionable whether this is attractive enough to develop in the near future.

## Part 2

This is the first in a multi-part series. [The second part](https://hackmd.io/@VVander/RPResearchpt2) focuses on the current problems facing Rocket Pool tokenomics researchers.