REP Distribution and Inflation

State of REP Distribution

There is currently 1,236,855.47 REP held by 417 members.

980,000 REP was awarded during the reputation staking period about a year ago in 2019 (see section 5.3.1 of the original whitepaper). Members gained REP by locking ETH (80,000), locking approved ERC20’s (300,000), registering MGN which was earned by trading on the DutchX exchange (500,000), and also by purchasing REP with GEN (10,000). If I remember correctly, there were 399 addresses holding REP after the initial staking period. It is important to note there is no clear way to know how many people this represents and some parties may control more than one address.

DXdao reputation can be minted or destroyed through proposals. To my knowledge, REP has only been destroyed once, and that was when Loopring proposed to slash 20% (21,000) of their REP. The rest of the REP (267,715.91) has been minted since the staking period to members who pass proposals or otherwise have generated value for the DXdao. The majority of this has been awarded to workers helping build DXdao’s products. This represents a 27% inflation rate in REP over the past year.

REP Issuance Framework

Now that the DXdao has raised capital and deployed dxdao.eth, mesa.eth, and omen.eth, the time has come for the DAO to establish a framework for the issuance of REP. The following should be kept in mind as the DXdao establishes this framework:

  • It is essential for DXdao to increase its membership and improve reputation distribution in order to become more decentralized. This should be done by issuing new REP to those contributing value to the DXdao.
  • REP holders have significant responsibility (management of DXdao treasury and products) and it is important for REP holders to have proper incentivization. One way to do this would be to compensate REP holders with the portion of the revenues being used as operating capital. That is, revenue is sent to the DXtrust, 10% is deposited in the buyback reserve supporting the price of DXD, and the rest is sent to DXdao to be used as operating capital. Some of the operating capital should be used to incentivize REP holders. This budget item would be governance cost.
  • In order for DXdao to be decentralized, it is essential that no group gain a controlling stake in voting power. A balance of power must be maintained amongst DXdao’s constituent groups. While these groups are not well defined, they include but are not limited to software developers, business development professionals, DXD holders, liquidity providers, and product (Mesa, Omen, etc.) communities.

The Numbers

A REP Distribution Framework requires consensus on a max inflation rate and the breakdown amongst the constituent groups. It would also benefit from a valuation of REP, and a max amount of inflation for an individual address. To get the discussion started I will throw out a suggestion. I am only just coming up with this as part of this post so treat it as a starting point.

Current Yearly REP Inflation Maximum

  • 50%

REP Inflation Budget Breakdown

  • Technical Workers: 25%
  • Business Workers: 25%
  • Liquidity Providers: 20%
  • DXD Holders: 10%
  • Omen Community: 10%
  • Mesa Community: 10%

Individual Address Maximum

  • 4%

Valuation Model

  • 1 REP = (Treasury Value * 1.2 ^ 5) / (Total Number of REP * REP Inflation Rate ^ 5)

Some Notes

Decentralization
At the moment, the top 9 REP holding addresses have > 50% of the voting power. With an inflation rate of 50% for a year and an individual max of 4%, the addresses with the highest amount of REP will have their REP percentage reduced by one third. After 1 year of 50% REP inflation, only the top 4 addresses would have greater than the max and would represent ~20% of the voting power. If 8 more addresses had the max of 4% each, then the number of addresses to get above 50% would be up to 12. Over time, with REP inflation and an individual address max, the minimum number of addresses to get above 50% voting power becomes at least a number equal to 50 divided by the maximum, rounded up. So at 4%, it’s 50/4 = 12.5, rounded up to 13. If DXdao over time lowers this percentage, the number can get bigger. E.g. 50/2 = 25, etc.

REP Valuation
The model I used bases the valuation of REP on the value of the treasury, using a 5 year time horizon and assuming a 20% growth rate in the value of the treasury. If you plug in a 50% yearly REP inflation rate and a current treasury value of $7M, then you get:

1 REP = 7,000,000 * 1.2 ^ 5 / 1,236,855.47 * 1.5 ^ 5 = $1.855

If yearly REP inflation is instead 30%, this number is $3.79.

Budget Breakdown
Giving a reputation inflation budget to the product communities could be a great way to increase decentralization and increase engagement with the community. It should be noted though that there may be a decent amount of technical work required to enable this to be done.

6 Likes

I’m not sure we can value REP this way, as the value in the treasury should also be accounted for the valuation of the DXD.
Assuming a 20% treasury growth per year appears to also be highly speculative (treasury can lower and I think it will in the first few years).
On the other hand, we could also account for assets the dxDAO own (but those should also be accounted in the value of DXD).

I also think that REP should be given more to maximize good and inclusive governance in opposition to it being viewed as some sort of financial compensation.

6 Likes

I like the idea of a maximum REP inflation rate. I wonder if we should also speak of a target inflation rate. In other words, distributing REP with a desired annual inflation rate in mind.

One thing to be wary of when directly compensating REP holders with treasury funds is the potential dystopia of an eventual cabal of REP holders giving its members more and more REP and voting themselves more and more funds. So I think due caution ought to be exercised. Also, in most cases so far it seems that contributions that generate REP also already come with a financial compensation. It might be better to financially reward contributions that merit REP minting, rather than reward the simple holding of REP, in order to incentivize active contribution. DXD already exists as a ‘passive’ vehicle for dividends.

As to the point of giving DXD holders a more direct influence, perhaps the locking DXD for REP notion could be revisited. One concept that came to mind is the minting of REP upon DXD locking, and the slashing of REP upon DXD unlocking. This could be done with a timer on both ends. For example:

  1. user locks 100 DXD
  2. after 1 week, user receives 0.1% REP
  3. user chooses to unlock DXD
  4. the REP gained in (2.) is immediately slashed
  5. after 1 month, user’s DXD is unlocked

The REP slashing and timers involved in such a system might allay some of the concerns expressed in previous discussions.

5 Likes

I think it would be a mistake for the DXdao to assign a financial value to REP.

Reputation is voting power in the DXdao and explicitly tying REP to economic power threatens the long-term decentralization of DXdao. Of course, REP holders may value their REP in monetary terms - and rightfully so, given the significant amount of assets that the DXdao already manages, but that does not mean the DXdao should encourage the financialization of REP.

I thought this from Matan on the DXdao Epicenter interview was relevant (transcript available in DAOtalk thread)

So it’s true that making [REP] directly transferable opens up more problems and we don’t see the reason to do that. But it’s not the main point, the main point is that it is deductible…You probably can transfer…and can sell your reputation in a very very small scale like maybe to your brother but you will not be able to scalably have it like an auction marketplace where you can buy as much Reputations as you want or sell as much as you want. So it’s similar to [when] people talk about bribery in DAOs. Off chain bribery - I would not create an attack vector - [because] offchain bribery is something that is very hard to coordinate and scale up. But onchain bribery is super attackable. It’s a huge attack vector. So you want to make sure that these things are not scalable and onchain-able

In my opinion, there should be only two guiding principles for REP distribution:

  • Yearly REP inflation rate - how much REP should be issued each year. Whether a mathematical formula or a goal ratified onchain, this should lay out the total amount of REP distribution in a year without dictating to whom or how much REP is distributed.

  • Guidelines and stipulations for REP distribution to DXdao contributors - suggested REP amounts for various contributions and rules preventing the accumulation of REP in an individual or entity. The Worker Compensation Guidelines have started the process of establishing standards for full-time contributors and I think similar guidelines can be created for bounties, DAO partnerships and (potentially) DXD holders. Ingamar’s Newcomer Jumpstart Period thread is also relevant here.

Neither the REP inflation rate nor guidelines for REP distribution to contributors require a financial value to REP. Both can be calculated fairly, regardless of the financial status of the DXdao.

The discussion of REP value has come up because the current Worker Compensation Guidelines limit the amount of REP an entity/individual can hold, and those that have “maxed out” on REP are hoping to get monetary compensation in lieu of REP.

In my view, this is circular logic. A contributor would need to be compensated for the loss of additional REP, if and only if the REP he or she previously received was financial compensation from the DXdao, but to my knowledge, REP has not ever been given a financial value by the DXdao.

This discussion is separate from how much to pay contributors, and perhaps worker compensation guidelines need to better account for those that are ‘maxed out’ on REP. It’s important to pay contributors fairly, and if DXdao hopes to be a world-class organization, it will need to attract world-class talent.

The compensation guidelines were designed to attract the right type of worker, ie someone that is aligned with the long-term values of the DXdao, but the current discussion involves those who have already contributed significantly to the DXdao - as demonstrated by their previous REP distribution before there were any assets in the DXdao. I don’t know if these contributors still need to “prove” their worthiness to the DXdao. Perhaps, there should be separate compensation guidelines for entities/individuals with maxed out REP?

Additionally, DXD issuance for compensation seems like a far better alignment of interests - both for community contributors and the DXdao - than REP distribution.

Lastly (I promise!), valuing illiquid financial assets is very, very hard and I’m not sure this can be captured by a single equation, which also may be prone to manipulation if explicitly endorsed by the DXdao.

5 Likes

During the initial staking period, 10% of the REP distributed was sold in auctions for GEN token.

The current worker compensation guidelines allow for flexibility in asking for either more REP and less financial assets or vice versa. Establishing guidelines for this implies a financial value for REP. What is the proposed alternative?

What is the reasoning here?

Perhaps thinking of REP in dry financial terms or as “compensation” is not the way to go for DXdao. But I would want to answer a related, and perhaps more fundamental question . . . What are the incentives for REP holders to be responsible stakeholders in the DXdao?

4 Likes

This a great question with no public answer at the moment.

3 Likes

My first view is that more than 20% is unhealthy. But maybe we need to look from other hand. If dxDAO set up 120% inflation rate and provide tool to protect REP share for active voters. Let say if REP holder voting on no less 50% of boosted proposal during the month he is qualify for claim his (her) new mint REP. If half of the REP holders will protect REP share - unclaimed new mint REP will be the budget for new members. Additional to this 10% inflation rate must be set up once a year without protection.

Also dxDAO must provide voting gas costs compensation for REP holders. It can be ETH or DXD compensation. I know a few dxDAO members who dont vote because they dont want to spent ETH. Maybe airdrop 1 DXD for past period for each REP address is good step.

3 Likes

Ideally it is better for dxDAO when REP holders = dxDAO product users. It is much better compare to dxDAO REP holders = people who earn money from dxDAO

new REP mint budget from my view
Technical Workers - 10% (this number must be sync with workers compensation guide)
Business Workers - 10%
Liqudity providers Mesa - 10%
Liqudity providers Omen (markets creators) - 10%
DXD holders - 10%
Omen community - 20%
Mesa community - 20%
Reserve - 10%

With good inflation rate set up dxDAO dont need this limit

If dxDAO paying fair salaries for workers in liquid assets ANY number of REP as a bonus for workers is good step. If workers accept this dxDAO dont need USD REP valuation.

3 Likes