Water💧 Shared Liquidity Proposal

The routing through liquidity can be easily solved with a simple PR as far as I know. It also happened with COLD/HNY before, and we were able to solve it.

By the way, the main point that I might be against the proposal is the multisig as I believe we can also leverage our own governance tools. That is why I thought a more comprehensive and transparent discussion would be better for the long term. Also, at this point, I can say that transparent info from other parties to be involved in the multisig is missing. We can improve that.

1 Like

Great point… We should say exactly who is in the multisig… I will work on that… We do that for all Giveth Multisigs, and this one is no exception.

For now here is a link: Gnosis Safe

1 Like

There are fans of the idea of sharing liquidity among communities, and I’m also one of them as you can see in my first reply in this thread.

What I’m not fan of is about it’s implementation. I don’t trust the way it’s implemented and I don’t like the idea that DAOs can’t undo giving liquidity. Why should they support the impermanent loss, and on top of that lose 15% if they ragequit? Is it really ragequit if you must ask the WATER board to leave and they must accept it?

Also, as I was saying WATER is not backed up by anything. Monstruosity says that one community backs the other. This is similar to be the unpopular kid in the class, and invite everyone to your party telling them the most popular kids will come. It may be true if you succeed with your lie, but this does not make you popular, just the appearance of popularity.

I think WATER is doing something similar with the rest of communities. It is taking the cool communities to a party to share liquidity, which is great. The problem is that it wants to be as cool as the rest without deserving it, and what is worst, make the others pay for the drinks.

WATER is requesting at least $100k to each community in order to join the party, but it doesn’t put anything from its part (except for some seats in the board in which there is people who nobody voted and the community can not revoke).

Furthermore, normally when you have an index token, it can be redeemed by the underlying tokens, something that has been successful in Moloch DAOs or in Dandelion DAOs such as PieDAO. It is not the case for WATER, another way to see why it does not have any intrinsic value.

My question is clear: didn’t we lean anything as a community from the agave launch, the agave hack and the alvin hack? Do we really want to put more than the 8% of our treasury on an untested token without the supervision of the blockchain engineers of the community, and without receiving any guarantee that we will be able to get our tokens back?

I request @paul and the rest of the WATER supporters to cancel the proposal to take some time to demonstrate that my claims are wrong and more importantly that the proposed WATER model is safe and fair for the communities that join it, and then present the proposal again. If you continue with the same proposal without providing more details on how it will work, I will defend my points in front of Celeste.


I am not sure I see it this way. The analogy would be better to say each community wants a pool to swim in but by themselves they don’t have the pool nor enough water. $WATER is an empty pool and each community brings an equal amount of ‘water’ to pour into that pool so they all can swim in the pool. $WATER doesn’t need to contribute anything because it doesn’t gain anything it just acts as the holding for the communities. What is swimming in this analogy? It is deeper liquidity for their token. Does $WATER gain anything having deeper liquidity, no the benefit is to the communities. Rather than use $WATER would you be opposed to using $HNY for this experiment instead of $WATER? That approach didn’t seem to get much traction

… I won’t cover multisig concern which has already been brought up

1 Like

There are three two elements that do not fit in the swimming pool analogy:

  • Communities are not only bringing their tokens, they are mixing them with WATER, a token that does not bring any value per se, and extracts value from the communities it is mixed with. If at least WATER was redeemable for the underlying tokens in a trust-less manner, the incentives of the WATER holders and the other community token holders would be more aligned. As it is right now, the more community tokens sold for WATER, the better for WATER and worst for communities. In other words, WATER tokenholders are not aligned with communities, and what can be good for the price of WATER may be bad for the price of the rest of the tokens.
  • The swimming pool is not an open space shared by the communities, but a private farmhouse that has a certain rules for communities to enter and swim in the pool. I am totally OK with that, but if it is the case it should be treated as a DAO2DAO deal between WATER and each community. The current proposal treats it as communities giving a lot of money that will no longer control. In other words, WATER is an independent DAO, and it should be treated like that, not giving it away tokens for free.
  • Not all tokens are the same or hold the same value (they are not all the same kind of water). When people hold two tokens in an LP position, they deal with the risk of impermanent loss and take measures to reduce it, because they have skin in the game. It is different in the case of the WATER board which are managing huge amounts of tokens from the communities without having a direct negative impact in their lives if they do not remove an LP on time when a community fails badly. It also provides pernicious incentives to communities to externalize their own problems on others, creating a vicious cycle. In other words, WATER will centralize the management of the shared liquidity without having enough skin in the game.

Sem… working on a more comprehensive launch doc which I’ll share here. I agree that there are important aspects of this launch that aren’t fully addressed yet.

Right now I want to cover the economic model of Water in more detail, because I think your comments come from a misunderstanding of either the proposal or of how uni v2 style decentralized exchanges work.

In the example below all 7 of the interested communities have joined Water. Here’s the spreadsheet which you can copy and play around with yourself. At equilibrium, the xDai values of each of the LPs are equal, and look like this:

But the individual token values for each community are different, so when looking at the unit values in each LP, it’ll reflect that. To make the model simpler I set all token prices to either $0.50, $1, or $2 to start, and included the HNY<>xDAI pair which is important in this economy as the top xDAI pair:

^ in this screenshot:

  • the only fixed token price is xDai @$1
  • the xDai value of HNY determines the Token Value of WATER, which then determines the Token Value of all other tokens
  • Total Liquidity = # of units in pool * Token Value

So what happens if the value of one of the tokens goes down? i.e. the token can be bought for a lower price somewhere else, but really this encompasses every imaginable reason a token goes down in value, which you can boil down to “people like owning that token less.”

Here is what the xDAI value LP chart looks like if the value of FOX has gone down:

Someone can now presumably buy FOX at a lower price somewhere else and sell it for $2 in the WATER ecosystem.

Assuming they want to cash it out for xDAI, the trade they’ll make is FOX > WATER > HNY > xDAI

This model assumes that price slippage will return the ecosystem to equilibrium when 0.1 $FOX is sold. Here’s what the unit value model looks like after that trade:

Damn! That was pretty cool. What happened?

  • All token prices have gone down
  • FOX’s price has gone down the most
  • HNY’s token price has gone down the least. Being the top xDAI pair dampens the price effect on HNY since it’s being bought in the HNY<>WATER pair
  • The price effect on WATER is the same as for all other tokens in the ecosystem
  • We’re back in equilibrium!

Now let’s look at the opposite, more pleasant event where FOX has gone up in value:

The arbitrage opportunity in this event is to buy FOX at $2 in the WATER ecosystem, since it can presumably be sold for more somewhere else. If the arbitrageur starts with xDAI, their trade will be xDAI > HNY > WATER > FOX. Again we’ll say equilibrium returns when 0.1 $FOX is bought:

Damn! That was even more cool! People are way happier now:

  • All token prices have gone up
  • FOX’s price has gone up the most
  • HNY’s price has gone up the least, since it’s being sold in the HNY<>WATER pair
  • The price effect on WATER is the same as for all other tokens in the ecosystem
  • We’re back in equilibrium!

With this I want to clear up some misunderstandings in your posts:

Because of arbitrage and price slippage, the price of WATER is a function of the prices of tokens it’s paired with.

There are no WATER tokenholders. 100% of circulating WATER at launch lives in liquidity pools. Unless you mean the arbitrageurs buying and selling WATER for their work, but they’re being compensated for the service of keeping the greater Honeyswap exchange in price equilibrium.

Anyone may choose to buy and hold WATER after launch if they like it as an index token, and that would benefit all the token communities in WATER. But investors and arbitrageurs aren’t able to sink WATER below its index token value because its entire initial supply lives in LPs.

The WATER board does not actively manage LPs, arbitrageurs play the role of active managers for WATER, and their incentives are well-aligned. The WATER board is only needed to remove liquidity when a community asks them to, or when a token is compromised. In the case of a hacked/failing token, WATER board members are personally motivated to remove that LP since it’ll negatively affect the price of the token’s they represent.

Communities are not providing tokens to WATER as an investment. They are paying for the valuable service of sustainable liquidity for their token.

This is the reasoning behind the 15% fee for RageQuitting. Participating communities should have the option to leave, but since WATER’s price could grow quite a bit, it might become tempting for those communities to treat it like an investment and try to add or remove WATER liquidity to maximize profit. This goes against the purpose of WATER which is to provide affordable, long-term liquidity.


I have to retract of one of the arguments I did before and I edited and stroked out. The reality is that…


  • The WATER token supply do not ever increase (which means that the WATER board do not ever mint new WATER).
  • The WATER tokens removed from liquidity pools with excommunicated or rage quitted tokens are never ever used (aka it is burned o sent to zero address).
  • The LPs are keep safe (aka the WATER board does not perform what is popularly known as “rug pull”, which would be unlikely).


  • The WATER token is still an empty-of-value token that acts as a value conducting vessel for the rest of the tokens paired with it.
  • It DO NOT extract value from the communities in any case. The incentives are well aligned in this particular case because all the WATER tokens created should be inside the initial LPs, so there is no way they WATER token price can go up and most of the rest of the tokens paired do not.

Although I still do not like the method in which WATER wants to be implemented, I apologize for the bad argument, and I hope this post serves as a good correction.


Thanks for the examples @paul, I think they they help understand WATER much more.

I still think there are lots of concerns that are not amended in the proposal and I’d like to propose two WATER alternatives that give more control to the communities:

Alternative 1

The WATER board could receive the funds from the different communities, put all of them in a Symmetric Finance pool (Balancer friendly fork on Gnosis Chain), and send them back to the communities.

We all obtain an index token and a huge amount of liquidity in Symmetric in just one step, and without most of the concerns we have been discussing about, because in this case the communities would have control their liquidity with the rest of the communities.

If a token had to be removed, the WATER board could suggest a new distribution, and a contract could be written that would automatically withdraw the liquidity form the old pool and add it to the new pool without the removed token, all in one transaction, without the rest of the tokens even noticing a scratch in their liquidity.

Alternative 2

WATER board could serve as an OTC market used when a community wants to exchange tokens with others in order to add liquidity to their token. Instead of buying the tokens in honeyswap, they could have a better deal doing an order such as 50% HNY, 30% AGVE, and 20% BRIGHT in exchange of any amount of their native token.

The WATER board would try to fill all the orders (maybe taking a small fee for the work it is doing), and sending the tokens to the communities.

Each community, who had already chosen which tokens are more convenient for them, would be able to hold them in their treasury or pair them in an LP. As in alternative 1, they would have entire control over the tokens they bought with their tokens and could decide when to stop providing liquidity with communities they suddenly trust less than before.


Both Alternative 1 and Alternative 2 send the tokens back to their communities, so we do not have the concern of a multisig holding the funds from multiple communities anymore. This also allow communities to decide what to do with their treasury.

The difference between them is that in Alternative 1 there is a recommendation made from the WATER board who chooses which token distribution should have the Symmetric LP based on the amounts provided by each community, whereas in Alternative 2 each community chooses which tokens to buy in order to build it’s liquidity treasury.


Hi @paul, I really like the concept you are working on with $WATER. Last week I was brainstorming a similar concept on the ShapeShift discord and a couple members referred me here. The ability for DAO’s to contribute to a multi-asset pool, as a method of diversifying DAO treasuries and creating a vested interest in other DAO projects, is vital to the DAO ecosystem. The solution you are proposing seems solid, but I think there are opportunities to simplify it so that it can scale and reduce friction for completing the pools more efficiently. I think Alternative 1 by @sem is even closer to a concept I was drawing up. I put together a summary write-up that might have some relevance to what you are working on with WATER. I’ve linked it here, and would appreciate any thoughts on it: D2D Multi-Asset Pool

I realize this might distract from your specific proposal here, so if there is better place to continue the discussion, let me know.


Here are the details of who is in the Gnosis Safe:

I think the plan is to also create a second safe that will be able to mint water, an 8/12 safe with the same people.


Remember that one of the conditions to effectively not extract value from the communities is precisely do not mint new WATER and to burn all the WATER that is obtained from retiring liquidity (in ragequits or excommunications).

I understand that in order to refill WATER for project that want to add more liquidity you need a reserve of water, but having some reserve of WATER for that purpose and have the ability to create as much WATER as you want are very different things.

WATER board is already trusted on holding the liquidity pool tokens, but granting powers to mint new WATER becomes problematic if the token becomes popular among the rest of the people and it’s used outside the LPs. Is there any reason why WATER total supply is not capped?

1 Like

Thanks for sharing Gunnar - lots of overlap, but I think a D2D Multi-Asset Pool is solving different problems than Water. The main goal of Water is to provide sustainable, affordable liquidity for the projects that join, and exposure to a diversified pool of assets is sort of just a nice byproduct (although critical to its utility).

It’s an improvement on founder-provided liquidity, liquidity rewards, and 1-to-1 DAO shared liquidity which are the most common ways tokens in our ecosystem have built their LPs.

I really like the interface in your proposal for creating a pool framework with terms that other DAOs can fill. Maybe the 2 ideas could be integrated where DAOs creating a pool could have the option to add shared liquidity between the projects as well?


@paul Yes I think that is an important piece that I was missing. One of the pool terms could include the % of tokens required to be contributed to a shared LP. I can see situations where some may want the shared liquidity, while others might want to use it as collateral, or sell the tokens for stables. Some DAO’s could use it for trading, selling in bull markets and buying back in bear markets, without needing to transact on their DAO tokens.

The shared LP pool would probably need to be under the custody of the marketplace DAO. Early redemption by participating DAOs would trigger a withdrawal of their portion of the shared LP, less the 15% charge which would be redistributed to remaining participants. Participants should hodl the marketplace DAO tokens to maintain some governance of the shared LP.




1. Handling token price fluctuation between proposals passing and Water launching.

Each community is requesting $100k to pair with $WATER, but by the time Water is minted token prices will have changed. In our direct shared liquidity proposals we set aside excess tokens for funding partnership initiatives - for example FOX from the 1Hive<>Shapeshift shared liquidity prop was just used to launch the first Regen Farm on Giveth.

At launch $WATER will be paired with all tokens at $1 and can be paired to match the current value of the tokens participating, with excess either burned or saved to add to LPs later. However with multiple tokens being paired with $WATER, the main effect of price fluctuation will be the balance of tokens weighted against Water as an index token. This will happen naturally after launch anyways as token prices change, but it probably makes sense to set limits on how much price fluctuation can happen before launch.

We can hedge against this effect based on whether tokens go up or down in value:

  • For tokens that go down, the community can either send more tokens before launch or the multisig sends all tokens back to the community.
  • For tokens that go up, the multisig will send the excess tokens back to the community.

A nice number for a price fluctuation cutoff is 50%, since that represents a reasonably substantial change. But if there isn’t consensus on that then we can find another value the community feels better about.

At what value of token price depreciation from $100k should we require communities to send more tokens?

  • Less than $50k
  • $50k
  • More than $50k

0 voters

At what value of price appreciation from $100k should we send excess tokens back to communities?

  • Less than $150k
  • $150k
  • More than $150k

0 voters

2. Should the Water multisig be able to mint more $WATER?

@sem pointed out that the ability for Water to mint more tokens after launch is a substantial power. The protection for this in the proposal is to mint $WATER in a separate safe with 8/12 signers needed, vs. the 5/12 multisig to add/remove LPs. The alternative is to cap the supply at launch and not allow any more $WATER to be minted. This would ensure the security of Water, but would shut out the possibility of building a governance system for this version of Water that can be expanded to new communities.

This is a super tough question that I’m personally on the fence about. On one hand I have no fear that an 8 of 12 multisig for minting $WATER would be compromised, and think that capping supply at launch does shut down the chance of a really powerful, ecosystem-wide liquidity solution being built out of this version of Water.

On the other hand, given that this is an experimental token design that wasn’t built for long-term governance, treating it only as a 1-time shared liquidity event would eliminate the risk of any issues related to long-term governance being implemented that hurt any of the communities involved.

SO, how do you feel about letting the 8/12 multisig mint more $WATER after launch?

  • HELL NO! Cap the supply at launch and throw away the keys!
  • HELL YEAH! Give this thing a chance to grow!

0 voters


Prop #2 is up in the Honey Pot: Gardens

This one is requesting $44,600 xDAI equivalent in HNY based on the current value of HNY in the LP Safe, targeting $100k total:

Based on the poll results we’ll do $50k-$150k as the range for token price fluctuation, and we’ll do a 8/12 multisig safe that will be capable of minting more $WATER.

Note that if Panvala joins (and all other tokens join as well), there will be a 13th signer, making the LP safe 5 of 13 and Minting Safe 8 of 13 signers.

Based on those updates here is the Launch Plan for $WATER:

Water Launch Plan

Water Gnosis Safe Details:

Participating Token Deadline: April 27, 2021

Tokens must be in the LP safe by this date. At this point, the Water LP Safe will send back any governance tokens if:

  • A community is unable to send all $100k xDAI value of their token
  • A token’s value depreciates below $50k xDAI value and the community doesn’t send additional tokens in time.
  • A token’s value appreciates above $150k xDAI value - excess tokens will be sent back to the community.

At the Participating Token Deadline:

  • If any of the Interested Communities have not added their token yet, their representatives are removed from Water LP Safe.
  • 2nd Gnosis Safe with 75% required signers is set up (“Minting Safe”)
  • Multisig agrees on an undisclosed date and time for token launch.

Token Launch:

  1. xDai value of all tokens in multisig are added together.
  2. Minting Safe mints 1.5x this value in units of Water token and sends to Water LP Safe (extra Water is minted to cover any price appreciation during launch. Leftover tokens will be returned to Minting Safe).
  3. Beginning with tokens whose primary liquidity is on Honeyswap, or who have 0 liquidity on Honeyswap (HNY, GIV, WORK):
  • 100 $WATER is paired with $100 xDai value of tokens for each community and added to liquidity pools.
  1. Then for tokens with some Honeyswap liquidity, but whose primary liquidity is on another exchange (BRIGHT, FOX, TEC, PAN):
  • Find xDai value of community’s token on its most liquid exchange.
  • Using tokens in the Water LP Safe, buy or sell the token on Honeyswap until its xDai value on Honeyswap is within 1% of the xDai value on most liquid exchange.
  • 100 $WATER is paired with $100 xDai value of tokens for each community and added to liquidity pools.
  1. For all tokens, the remainder of each community’s token supply in Water LP Safe is paired with $WATER and added to LP at the current exchange rate on Honeyswap.
  2. When all LPs are added, leftover $WATER tokens are returned to Minting Safe.

Water launched Friday, April 29 and is working as expected :slightly_smiling_face:


An initial mint of 790,511.512 WATER was used to add to the LPs at $1 per WATER.

After LPs were added, 268,854.253 WATER was leftover and sent back to the minting safe.

Safe addresses:

Next Steps:

  • Marketing for Water token. Private buyers of the Water token (who swap for Water with a token that isn’t paired with Water) get an index token with extra upside. We want to get the word out about this since people buying Water creates upward price pressure for all paired tokens.
  • Long term governance. Start work on building a more secure, decentralized, and community-oriented governance system for Water.

Adding a GNO<>WATER LP

The treasury management team for GnosisDAO - Karpatkey - has expressed interest in adding a GNO<>WATER LP to our shared liquidity :slightly_smiling_face:

They proposed $30k in GNO, which would be paired with $30k in Water from the 8 of 12 signer Water Minting Safe, and then LP tokens would be held in the 5 of 12 signer Water LP Safe with the other LPs.

Karpatkey would also add its own signer to both Gnosis Safes, so the Minting Safe would become an 8 of 13 signer safe and the LP Safe would become a 5 of 13 signer safe.

The current signers are pretty excited about this - myself included - because GNO is really well-aligned with all of our communities and because Karpatkey is doing super impressive treasury management work. They’re an ideal partner in helping build more scalable shared liquidity through Water.

If anyone has questions or hesitations please post them here. Otherwise I’ll plan on approving these transactions when they come up - and I’ll keep using this forum for updates as they go through.



The TEC community has voted exit the WATER partnership on Gnosis chain should the community vote to migrate to Optimism. That migration vote just passed and the migration plan is underway.

This request, on behalf of the TEC, is to exit WATER and pay the 15% penalty fee.

Currently, the amount of TEC is ~48K.

All funds, minus the fee should be sent to the TEC’s Common Pool: 0x01d9c9ca040e90feb47c7513d9a3574f6e1317bd

The main POC for this should be @griffgreen, who is also on the WATER multisigs.

Please let us know if this is not going to be possible before Friday, Dec. 15th.

1 Like

TEC’s exit from the Water pool is complete!

The Water LP Safe removed all of the TEC<>WATER LP and sent 39,170.5069 $TEC to the TEC Common Pool:


Since TEC’s LP grew in proportion to the average LP in Water, the fee was based on the initial # of tokens added to the pool which was 46.082 $TEC added in 2 separate transactions:


Thanks for this update, @paul, and executing so quickly.

1 Like