I'll preface this post: I am genuinely interested in this coin and what it can offer and like most of the ideas being proposed. I am in the IRC channel and intend to be active in it. I am not here to FUD or bash this coin. But...
I have a concern that the developers are actually able to take a MUCH larger amount than the 10% subsidy just by staking their premine coins. Remember that 30% of all minted coins are given as PoS rewards.
On day 1 the developers will have 4% of all coins, but actually 50% of all coins in existence. If they stake these, then they will get 50% of the 30% proof of stake coins, a further 15% of ALL minted coins. This percentage will taper off over time as their % of all minted coins becomes smaller, but the numbers are still scary...
I have a concern that the developers are actually able to take a MUCH larger amount than the 10% subsidy just by staking their premine coins. Remember that 30% of all minted coins are given as PoS rewards.
On day 1 the developers will have 4% of all coins, but actually 50% of all coins in existence. If they stake these, then they will get 50% of the 30% proof of stake coins, a further 15% of ALL minted coins. This percentage will taper off over time as their % of all minted coins becomes smaller, but the numbers are still scary...
Thank you for bringing this up on IRC, it's critically important to understand how the subsidy system works and where the funds are going. The base reasoning is correct, but the inferences require detailed explanation inline below. The calculations referenced in this explanation are contained in a spreadsheet, linked here.
- The premine will split 8% equally between two groups: (A) c0/devs and (B) the airdrop participants. This means that 50% of all the decred in existence at launch will be owned by c0/devs. The other 50% will be owned by the larger community.
- At the time of launch, the development organization (dev org) will have 0 coins. As blocks are mined on mainnet, the dev org will begin to receive its 10% subsidy from each block.
- Since approximately 3.02M coins will come from block subsidies, i.e. not premine, this means the dev org will collect 302K coins. The dev org will fund ongoing work, so it is expected to spend most of the coins it receives and not accumulate large amounts of coins. The average amount of stake held by the dev org should be relatively low as a function of the org not carrying a large balance.
- Starting with 50% of the coins at launch and 4% of the total does not entitle c0/devs to 15% of all minted coins. The reason is that 60% of the subsidy goes to PoW miners, which c0/devs have no overlap with. The plan is to have a few GPUs mining Decred after launch, but 60% of all the coins end up with PoW miners. When these coins are held by PoW miners, or whoever they sell it to, they can be used to mine PoS themselves. If coins are voted in proportion ('in proportion' explained below) to the amount held, the 4% premine turns into 7.32% of the total at the end of year 5, which is assuming none of the premine coins are spent. This means that c0/devs have 4% turning into 7.32% in 5 years, not 4% turning into 19%.
After 1 year, it is estimated that 4.7m coins will have been freshly minted, which is approximately 22% of the 21m coins. If the developers did not sell their premine coins, they would still hold their 4%, PLUS 2.2% for their developer subsidy (10% of the 22% freshly minted), PLUS whatever they have earned from staking for 12 months (roughly another 10% of all coins minted in the first year / 2.2% total, as they will be taking 15% of PoS coins on day 1 and around 6-7% of PoS coins after 1 year). So after 1 year the devs would have 4% + 2.2% + ~2.2% = 8.4% of all coins (~1.764m) which is a staggering 27.6% of all coins minted to date by then (1.764m out of 4.7m + 1.68m premine).
This also means that at the start of year 2 they could still be getting 8.3% of all newly minted coins through staking (27.6% of the 30% PoS) on top of the 10% dev subsidy.
I can understand the concept behind the 10% developer subsidy AND the premine (though I have a question mark over whether they need both), but I am seriously concerned that after 1 year the developers could still be raking > 18% of all new coins if they just sit and stake their initial premine.
This also means that at the start of year 2 they could still be getting 8.3% of all newly minted coins through staking (27.6% of the 30% PoS) on top of the 10% dev subsidy.
I can understand the concept behind the 10% developer subsidy AND the premine (though I have a question mark over whether they need both), but I am seriously concerned that after 1 year the developers could still be raking > 18% of all new coins if they just sit and stake their initial premine.
- Per above, the concerns about the dev org mining lots of stake are addressed. The plan is that most of these coins will be paid to a diverse group of developers in the form of contracts who complete deliverables for the project on an ongoing basis.
- There are two "bounding" scenarios to consider: (1) where everyone mines PoS and (2) where nobody mines PoS besides c0/devs. The spreadsheet linked to in this post contains the projection for the "everyone mines" scenario. In that case, the 4% premine, if entirely unspent, grows to approximately 5.38% after year 1, 6.16% after year 2, and 6.67% after year 3. The percentage of total coins held by the 4% drops from 50% at launch to 24.4% after year 1, 18.0% after year 2, and 15.1% after year 3. The "nobody mines" scenario is a failure mode of Decred and is unlikely to happen.
- The hope is that a large number of the coins are used to mine PoS since you not only want users to share in the gains from PoS mining, a strong participation in PoS is needed to secure the network.
- In the event that there is a low participation rate in PoS mining, the plan is to throttle c0/devs' PoS mining so that it does not hoover up all the coins. The goal with PoS is not to hoard the coins, but to create a strong incentive structure for users to participate in securing the network with their voting. c0/devs will not mine more PoS than projected in the "everyone mines" scenario, as it would go against the spirit of the project.
Am I missing something in all of this?
Could there be a mechanism to prevent the developers from staking their premine coins, until they are sold on or until much later(year 3 or more) when their % of minted coins to date becomes much less?
Could there be a mechanism to prevent the developers from staking their premine coins, until they are sold on or until much later(year 3 or more) when their % of minted coins to date becomes much less?
The threat you describe, "How do we moderate the amount of PoS mining power for our 4% premine?", was also discussed prior to your bringing it up. The plan was to start mining PoS at a moderate pace so that people besides the 4% premine can get stake tickets without too much effort. After you brought this up, the other possible options were discussed that have been mentioned by you and others in the community.
The complexity of the constraint for what would be considered good behavior on c0/devs' part makes implementing such a constraint difficult. The belief is that the proper constraint would be "c0/devs can mine some PoS, but not too much PoS, and no pump and dumps" or similar. Other projects have CLTV'ed coins to make them unusable until a certain block number, but this would mean c0/devs cannot use these coins to secure the network by mining PoS. A notable distinction here is that Decred never took external funding for the bringup work, and, as such, has no pre-existing social contract with the userbase that would merit a promise to behave a certain way. Although, it must be said that inappropriate behaviour by c0/devs removes any prospect of sustainable development, which goes against c0/devs' history of development over the years. This is because there is strong incentive to reproduce this level of activity for Decred as one of the development stakeholders in the ecosystem. The reasoning for airdropping 50% of the premine was based on the logic that Decred will be most vulnerable at mainnet launch and therefore require a steward to ensure it is not essentially dead-on-arrival. Airdropping more than 50% of the premine was considered, but doing so would be a risk to the security of the chain since bad actors could scoop up a large percentage of the initial coins and wreak havoc with PoS mining. A 50% airdrop in the premine was the largest safe size for the airdrop that could be determined, and it can be viewed as a set of training wheels that are taken off as time passes after launch.
This is presented as an "intention of stewardship" and will be judged against deliverables, instead of taking people's money and making lots of potentially-empty promises. There's no claim it's a perfect system, it's an alternative model that will explore new areas in both technology and governance, but hopefully this post sheds light on the stewardship process and issues raised.
