Given the solution was discussed and outlined in the green paper, it seems like this PEP really just focuses on sizing up the impact the DAO is targeting. If that is the case…
TLDR - there is no exact answer as to sizing it perfectly unless we have existing data points on what impact any allocation would have. The proposal then becomes more about addressing all the relevant issues of how we allocate and bootstrap the wPOKT program comprehensively. If you skip to the bottom, I think this amount is sufficient should it check all the right boxes.
To do a deep dive on the topic and have complete precision as to set an allocation, we’d need to know:
- what is the exact demand for staking on the Pocket Network with a newly available token for new markets
- what the probability is for getting additional DAO support later on (to either a POKT specific LBP or general AMM) in case 5m POKT isn’t enough
- that we could forecast what the actual impermanent loss would be
These may not be knowable right now, but we could answer them after this initial launch. Re: #1 the launch acquires those first data points in a sort of price discovery mechanism, re: #2 any decision here doesn’t explicitly preclude us from asking the DAO for additional support later on so I don’t think the probability is 0%, and re: #3 impermanent loss is still largely misunderstood given that, like @JackALaing suggested, it could be recouped with block rewards, and it is also a proxy for more enthusiasm in ETH which actually adds to the flywheel that creates more interest in the Pocket Network.
Furthermore, @Garandor adds some great points re:
I would also STRONGLY recommend to adapt a (nonlinear) bonding curve AMM model (see Nexus Mutual NXM for example) for wPOKT instead of dumping liquidity on uniswap.
In that case the 5M initial liquidity might even be enough, a small allocation would simply limit the growth speed of the network compared to a larger liquidity allocation.
and
Instead of creating a custom AMM bonded with staked POKT, an alternative could be a 1INCH pool with a very high price-impact fee that would
a) disincentivise speculative big buying/selling (thus keeping the community OTC in business)
As far as the AMM dynamics and allocation size pointed out above, it looks like they are properly addressed:
- if the DAO wants a bonding curve, it should be covered in both Uniswap and 1 inch AMM code so we should be covered from that angle
- the proposed token reward system on the AMPL Geyser V1 6 smart contract would actually achieve a similar goal of incentivizing long-term holding, so additional curve changes may not be needed
- going from 20% to >20% allocation of the treasury is certainly possible whereas it is more complicated/controversial to go the opposite direction
- 1inch could be a fine farming program to use as it has the optionality to support 1 month pools. If the DAO is left wanting more liquidity, the DAO can extend with additional pools and we incrementally move into larger programs
Given the insights above and the stated motivation, it seems like the 10M POKT ought test out what this bootstrapping can do for the Pocket Network. Some of the stated boxes to check would be:
drive measurable growth in relays from the funded Apps
create a normalized set of data points on price-discovery method where POKT would fit into the wider Ethereum-Pocket ecosystem (without relying on OTC)
gain exposure (organic marketing) of being visible on a large exchange
test whether we can recover any of the impermanent loss through additional block rewards the DAO gets rewarded with
create operational lessons as to how a LBP just for POKT is used for a distribution event and an LP on an AMM is used for ongoing liquidity
There is no perfect allocation that ensures any a priori knowledge so it really about whether the proposed amount is sufficient. It stands to reason that 10m POKT is a manageable amount to initially test out support of a wPOKT program and measure against metrics for each of the check boxes above. The beauty to starting at 20% is that the DAO can adjust along the way depending on the measured outcomes.