AGP discussion: Aragon Association acquisition of ANJ for security purposes

Wanted to open a discussion about this AGP for the Aragon Association to acquire ANJ and hold it over the course of the first few months that the Aragon Court will be operating, until it’s proven and seamlessly working.

I created this proposal based on @jorge’s insights on Aragon Court needing some handholding in the case something breaks and we need some ANJ to interfere in a dispute under force majeaure conditions. He thinks 1m ANT would be a reasonable amount to lock.

However, I’m not familiar with the ANJ boding curve and its parameters, so I’d love for people familiar with the matter to chime in and think about the possible implications of this AGP. Some questions:

  • Can the AA, and therefore the project, lose money with this operation?
  • Is it sensible to lock so much ANT even before the Court has been working for some time, and there are valuable disputes and economic activity worth protecting?
  • Does the AA having so much ANJ create any issues in terms of DAOs using Aragon Court, since they would be somehow depending on the AA storing keys safely and not tampering with disputes?

Looking forward to feedback on those. Open to withdrawing the AGP until a later point in time if something doesn’t make sense.
Luis

6 Likes

This is an interesting proposal. Im not knowledgeable enough about the curve or its paramters to understand the implications but some other thoughts

How would the AA go about this? If it locks up 1,000,000 ANT, this will naturally have a affect on the price of ANJ. Depending on how much ANT is locked up this could be quite big. If this is only for a few months it may have an adverse affect on those who decide to buy after the presale.

Given that there is not likely to be a great demand for cases in the very short term, it would not be a good situation if early adopters of ANJ loose money. This would be a problem because we want to have as deep a pool of jurors as possible

It’s also worth noting, it’s not just the economic activity we are protecting, it’s the reputation of the court it’s self. If the AN is successfully attacked, this would be an absolute desaster. Regardless if the economic impact to a DAO was small

I don’t think AA’s key management is a issue. It’s been managing the networks funds for a long time and there have been no concerns. The issue will come down to trust. Do DAOs trust the AA to not intervene.

I don’t have any answers my self just a few extra points to think about

2 Likes

Defnititely. I assume AA would buy that ANTduring the presale, so it wouldn’t affect the price initially and would get a better deal. But the problem arises when the AA decides to redeem. Look at these graphic, where X-axis represents total ANJ supply and Y-axis represents price:

  • If after selling the total supply is below the initial supply reached during the presale, all tokens below that mark (E in the figure above) would be sold at a lower price than the presale one. Essentially, if ANJ generated after the presale is quite lower than ANJ bought by the AA during the presale, then the AA would lose money. We must take into account that tokens sold over that mark would have a gain. I can come up later with a formula to calculate exactly how much AA would gain or lose depending on those 2 parameters (ANJ bought by the presale, ANJ generated afterwards), but looking at the figure you can have a good intuition: selling to the right of E gives you profit (the height from the horizontal line by E to the curve) while selling to the left of E makes you lose (the height from the curve to the horizontal line by E).
  • For that reason, the AA could also make money on that operation. This would look like weird to me.
  • Another option is not to sell ever. But then users of the Court should have to trust that the AA would not intervene if not necessary. And how we define necessary?

About the other 2 question I pretty much agree with Aaron.

4 Likes

Btw, about this phrase in the AGP:

there may be 51% attacks to the Court itself, which could put the whole Court in jeopardy

The Court already has a mechamism to mitigate that: jurors who vote yes on a final round get their tokens locked so the rest can exit the bonding curve before them.

1 Like

Thanks @bingen! Interesting, if this is the case, then I don’t see the need for this AGP.

Also, I agree with your points: if the AA can both lose money or earn money, it should also be looked from the investment perspective and analyzed as such.

Would love inputs from @facuspagnuolo @jorge @lkngtn on this as well!

1 Like

Good point. This AGP would put enormous sell pressure on ANJ for the first few months. This would definitely be an issue for anyone who plans to buy ANJ, knowing there will be such a huge sell pressure.

From what I understand, the ability for the AA to free up its locked token is going to depend on the court’s traction right @bingen @jorge? From our current calculation at the AA we have liabilities of up to ANT 3M (potentially paid to the Nest program as well as past and current teams). This leaves us approx. 4.4M ANT to play with.

Locking/spending 1M ANT is 20%+ of available treasury and has a non negligible opportunity cost. Can this amount be minimized?

2 Likes

Not sure if I understand the question correctly, but the AA would always be able to sell the ANJ to get the ANT back, what would vary would be the price.

1 Like

Hey eagles!

Quick feedback after landing back from the Far East and drafting our proposal that is centered around helping teams figure out the intricacies of a Fundraising campaign.

Here you can see the Fundraising app’s presale configuration basic parameters

And the pricing formula for the bonding curve:

The DAO can launch a campaign with a certain “Initial Tokens Offered %”. In other words the DAO deploying the Fundraising campaign would be able to keep ANJ minted for itself. The AA (or another DAO) could thus get hold of minted tokens without a purchase in the presale which it could subsequently sell against a portion of the underlying collateral during a future vested period.
This creates a tradeoff space between how much the DAO wants to reserve for itself for Court security and treasury investment purposes and guaranteeing a stabler price for presale participants.
One radical solution mentioned by Bingen to acquire this security buffer while preventing fears of a selloff by other participants would be to have the AA take a substantial portion of the token supply and commit to supporting a voting app proposal to blacklist its address from trading.
Theoretically AA could always act adversarially and move these tokens to another DAO/address or reverse the blacklisting decision but the vote would take time and give participants extra time to exit.

To achieve balance the Network’s position could be acquired through a balance between ANJ minted and ANJ purchased.

The vesting schedule and cliff would determine the rate at which minted tokens and purchased tokens can be sold after the curve is opened to trading.
In effect this creates a link between the liquidity of the tokens minted or purchased during the presale phase and the resulting selling pressure during the trading phase.

Regarding valuation models of bonding curves we are still in a theorisation and testing phase. However my hunch is that bonding curve tokens value could be compared to a stock’s traditional DCF valuation where the value of a stock is supposed to be composed on the terminal value of present and future dividends discounted by the risk-free return rate.
For a startup the value of the token is driven mostly by growth expectations, for a stock such as an utility company or coca cola where growth expectations are lower the price is driven mainly by profit ratio/revenue size and thus it’s dividend ratio.

In a bonding curve instead of dividend payments the price of the tokens is dependent on the amount of collateral in the curve, wherever it comes from stakers or revenue flowing in. It could then possible to compute the expected value of the token by using forecasts of collateral size and simulating how revenue and token demand for service use influence on speculative token demand on each other.

In the Court’s particular case I assume there would be no revenue buybacks meaning the price would be driven purely by the expected demand for ANJ which itself is a proxy of the Court’s fee model and thus it’s “economy”. It the expectation from the Network is that stakers will be very hyped then it makes more sense to lock a high amount of ANT while there is not activity yet. If however the Network believes stakers will “wake up” to the value of the Court as it starts operating more sizeably it should lean on the side of waiting.

Bingen already cited one preventative mechanism:

Additionally to prevent 51% attacks on the court you could also have the bonding curve token holders vote not via the classical voting app but via the upcoming conviction voting app. This would make a hostile takeover longer to execute and give the necessary visibility and time for stakers to leave the bonding curve compared to a potentially immediate purchase of 50%+1 of the ANJ token supply.

4 Likes

I don’t think the AA should participate in the Aragon Court.

Aragon Court needing some handholding in the case something breaks and we need some ANJ to interfere in a dispute under force majeaure conditions

I think setting the expectation that there is a centralized entity controlling a decisive stake would be counter-productive and undermine the actual utility and purpose of the court.

First its worth noting, as far as I can tell that this position would also require the ANJ be actively participating in all disputes they get drafted on. In order to have a decisive position, this would mean being involved in the bulk of all disputes and appeals, and crowding out other participants. It would not be possible without continuously loosing stake and generally making the court function improperly, to have the AA actively stake and not participate in disputes outside of Emergencies.

Second, when the court launches it will not have established a reputation or credibility, and it will be absolutely important that the jurors who participate early work to firmly establish and earn that reputation. There isn’t a shortcut here, there is risk, and there is upside for joining the court as an early juror and establishing that reputation. While we don’t have much in place to prevent any single party from taking a large decisive stake in the court early on, we should discourage this as it puts legitimacy of the court at risk and makes the appeals process less effective. People participating as jurors should know that taking a large, decisive position is counter-productive because a single entity having a decisive stake in resolving disputes will decrease the legitimacy and utility of the court as a dispute resolution mechanism for potential subscribers.

One of the main short-term goals of the court is to start to take over the role of the Association in Aragon Network governance, but rather than the AA being an active participant in the process initially and gradually exiting that role, it feels prudent to instead gradually increase the responsibility of the Court’s authority over time. (For example we could start to allocate funds to the independent and autonomous Aragon Network DAO on a predictable schedule), gradually transitioning more governance authority from the AA to the Aragon Network DAO as the process stabilizes and the court establishes its credibility, reputation, and the legitimacy of the ANJ distribution.

3 Likes

Wow Daniel, what a thorough explanation! I’m bookmarking it for future reference. :wink:
Just a quick comment: the version of the fundraising app (+presale) that we are using for ANJ is quite simplified: it doesn’t have things like tap mechanism or vesting.

2 Likes

I agree with Luke on everything except one thing:

Although quite inconvenient, the AA could only stake the tokens and activate them just before an emergency final round, to deactivate them right afterwards. (It could be drafted for other disputes being drafted during the same term though)

2 Likes

As it is currently written this proposal would represent 35% of the quarterly budget.

However, since this particular proposal is a re-allocation of the Associations assets its not entirely clear how this should be handled in the budget. My interpretation is as long as this is in the finance track it will be debited against the quarterly budget, but a case could be made for proposals which impact the “Portfolio Allocation” to either be moved to their own track, or moved to the Association Policy track as AGP-11 delegates treasury management to the AA. Would appreciate @light as well as other’s input on this interpretation.

2 Likes

Very interesting thread and great comments on this proposal. I think the easiest way to handle this would be for the AA to use its discretionary powers (and the treasury management responsibility explicitly given to them in AGP-11, as you point out) to make this allocation outside of the AGP process. If the AA would like to receive the approval of ANT holders before making this allocation, then it could pass either an Association or Proclamation track AGP (I think either could work, depending how it’s worded) stating the proposal and a detailed plan about what they will do with the ANJ and then ANT holders can vote to approve the proposal or not.

If the Association/Proclamation AGP is approved, then it would have no impact on the budget until the exchange is actually made – long after other approved Finance track AGPs in this vote are processed, since the curve doesn’t even go live until 2020. (Edit: Unless we count the ANT allocated in this proposal as a liability - in which case it would deduct exactly that much from the treasury balance at the moment the proposal is approved, but have no effect on the budget otherwise.) Whenever the next Finance proposal is approved in a future ANV and we have to calculate the budget again then we’ll have to figure out how to give a fair market value to the ANJ held by the Association. This (valuing ANJ) is an interesting topic but probably outside the scope of this thread, since as mentioned I don’t think there are any practical implications for this AGP or this upcoming vote/ budget period.

Edit: I started a separate thread to discuss the general topic of how to handle Finance AGPs that re-allocate or re-balance the treasury, using this proposal as an example were it to actually be presented as a Finance AGP. Happy to further discuss this more general topic on that thread, and reserve this thread for discussing this specific AGP.

2 Likes

This makes a lot of sense to me