GIP-9: Launch Liquidity Mining program for Prediction Markets using Gnosis Conditional Tokens

Should the GNO Liquidity Mining program for Prediction Markets using Gnosis CT be launched?

  • Yes, let’s launch this Liquidity Mining Program
  • Make no changes

0 voters

GIP: 9
title: Launch Liquidity Mining program for Prediction Markets using Gnosis Conditional Tokens
author: Jimmy Ragosa
status: Phase 3
type: Meta
created: 2021-03-22

The original phase-1 post about setting up a Gnosis CT Liquidity Mining program was well-received and the requests for clarification were mostly focused on Gnosis CT value capture mechanisms for GNO holders which can be designed and proposed in other dedicated threads. The proposal is thus now moved to phase 2.


GnosisDAO will launch a Liquidity Mining program for prediction markets using Gnosis Conditional Tokens that would reward liquidity providers with GNO to incentivize a curated set of markets and thus boost the use of Gnosis Conditional Tokens.


The GnosisDAO will allocate 120,000 GNO from the treasury for the first 12 months of this liquidity incentivization initiative. It will dynamically manage the GNO rewards distribution by assigning “allocation points” to specific markets that the GnosisDAO votes on in order to stimulate the markets with a higher participation potential or that would generate quality predictive data for the GnosisDAO. Currently, it means incentivizing mainly liquidity providers for Omen markets on xDai but the program would be open to any solution using Conditional Tokens (for example, it could also incentivize ERC20 wrapped conditional tokens on exchanges such as Uniswap, Balancer, SushiSwap, and 1inch).


In the current environment full of yield farming opportunities, providing capital to any protocol incurs a high opportunity cost. Moreover, prediction markets are known to be more attractive to users and more potent at predicting events when the markets are liquid and efficient enough to allow for high-scale arbitrage (cf. Vitalik Buterin Prediction Markets: Tales from the Election blog post).

Anyone can add liquidity to prediction markets using Gnosis Conditional Tokens by minting them for existing markets or even creating their own markets. Liquidity attracts traders, trading generates prediction market fees, and ultimately this profitability attracts more liquidity. This is a flywheel effect that has to be incentivized within the Conditional Tokens ecosystem in order to position ourselves in the prediction market scene.

Regardless of the venue, early liquidity providers take on more risk and opportunity costs: including contract risk, low initial profitability, etc. Moreover, conditional Tokens users should get to participate early on in deciding how the GnosisDAO evolves and plans to build new versions of the Conditional Tokens.

Toward this end, we propose implementing a Liquidity Mining program for prediction markets using Gnosis Conditional Tokens rewarding liquidity providers with GNO tokens. These types of initiatives are proven to attract significant amounts of capital into products and to kickstart a virtuous cycle of additional liquidity (see the examples of Balancer, Sushiswap, and 1inch) and thus improved user experience and efficiency which will, in turn, bring more usage to Conditional Tokens.

The indirect benefits of this program for the GnosisDAO are multiple:

  • Fostering the use of one of Gnosis core products (CT framework).
  • Upgrading the distribution of GNO by rewarding it to valuable community members.
  • Improving the quality of data generated by prediction markets using CT.

The potential direct benefits to the GnosisDAO are just waiting to be unlocked by the governance. In the future, it could take the form of the implementation of a redemption fee mechanism embedded into the Conditional Tokens framework, or of a trading fee in exchanges controlled by the GnosisDAO. Legacy prediction markets like PredictIt charge significant fees (10% of the profit on a market and 5% of withdrawals), blockchain prediction markets like Augur or Cover charge a smaller redemption fee (~1% for Augur and 0.1% for Cover). As long as the GnosisDAO provides continuous support and development for Conditional Token, a small fee will probably be accepted by the users.

It would also be possible for the GnosisDAO to find novel revenue sources around Conditional Tokens. For example, frontends controlled by the GnosisDAO could also allow users to burn some GNO in order to have their markets displayed in more attractive positions of the UI.


The GnosisDAO liquidity mining program for prediction markets using Conditional Tokens will be inspired by Balancer and SushiSwap liquidity mining programs.

For the first year 120,000 GNO tokens, or 10,000 per month, will be distributed to liquidity providers staking their conditional tokens liquidity into the GnosisDAO liquidity mining contracts. For networks other than Ethereum mainnet (in particular xDAI), those GNO tokens would be wrapped on the particular network.

The amount of GNO distributed to a particular liquidity pool will be determined by an allocation point system. The GNO will be distributed between pools in proportion to their allocation points and within pools in proportion to participant liquidity.

For example let’s assume there are 3 pools A, B, and C with 100, 400, and 500 allocation points, Alice has 200 ETH of liquidity in pool A which has a total of 500 ETH of liquidity. Alice would receive 10000*(200/500)*(100/1000)=400 GNO per month.

In order to be able to rapidly incentivize new markets when opportunities come up, setting up allocation points will be done through a fast-track procedure on Snapshot which does not require going through the complete governance procedure.

Any account with at least 1000 GNO (from a snapshot taken at the beginning of the month) will be able to make an allocation point proposal each month for each 1000 GNO it holds (so an address with 3000 GNO would be able to make 3 allocation proposals per month). Those would be put to vote instantly with a 1 week voting period and without quorum requirements.

An allocation proposal would contain:

  • The conditional token liquidity pool to be incentivized.
  • The number of allocation points for this pool.
  • The start and finish date of the incentivization.

Allocation proposals can provide allocation points to new pools or change the allocation of a current pool (including setting it to 0).


The plan outlined in this proposal is subject to discussion and change. They may also need to be restructured to take account of legal, regulatory, or technical developments as well as governance considerations.


This would be a separate snapshot space?

Not especially, it can be done in the same snapshot space, using a different prefix dedicated to this initiative allocation points proposal (GAPP for GnosisDAO Allocation Points Proposal).

(Example: GAPP-1: Allocate 800 points to Omen 2020 Olympic Games Winning nation market or GAPP-2: August 2020 Prediction Markets Mining Allocation Points proposal )

But if it were in the same snapshot space, we wouldn’t be able implement the voting strategy that you specified. It believe it would need to be a different space to have a different strategy.

The “minimum 1000 GNO held to make an allocation point proposal” is a requirement to make a proposal, not a voting strategy, and could be enforced like the current GIP proposal process is enforced.

Would the “No Quorum needed” requirement necessarily require another space? I’m not sure.

Right, but there is nothing stopping an account with less than 1000 GNO from creating a proposal on the Snapshot space (you currently need >=1 GNO to create a proposal). We would need a separate space with a different strategy to enforce >=1000 GNO.

Quorum is not actually a feature of Snapshot, it is enforced on oracle side.

You could, of course, enforce the 1000 GNO requirement on the oracle side too. But I’d worry a little about there being lots of invalid proposals (noise) that end up being rejected by the oracle because they weren’t created by an account with >=1000 GNO.

After thinking more about it, I agree that a different snapshot space dedicated to this Liquidity Mining program (or all LM programs) with the =>1000 GNO requirement would be the best way to keep the proposal process open and resistant to spamming.

For reference:

  • Balancer is moving to a system where a limited committee of individuals has the power to make Liquidity Mining proposals in a dedicated snapshot space separated from the main governance one.

  • Sushiswap still uses the same snapshot space but the proposal can only be made by their core team

1 Like

The proposal is now live on Gnosis DAO Snapshot and voting starts tomorrow.


The proposal is currently at 26 000 / 60 000 GNO of the quorum and will finish April 29. If you’ve followed the proposal and the discussions don’t hesitate to vote.

Hey - first of all - I really appreciate the effort made for this proposal.

I shortly want to share my view on this. Personally, I am neutral towards this proposal and will not participate in voting - in case it reaches the required I will of course put in my best effort to make it successful but in its current form I would not “campaign” for the proposal.

Here is my reasoning:
The shift of focus:
At the beginning of the year, Gnosis attempted to increase partnership with dxDAO and to run Omen together. Those negotiations for now failed. As a result, we at Gnosis put our focus away from conditional token/ prediction markets and fully concentrated on: GPv2, Gnosis Auction, The Safe and Open Ethereum. While we see potential overlap with CT and GPv2 we still decided to first cover regular token trading with GPv2 before we build extra specializations for conditional tokens.
Given that CT are thus currently not our focus this proposal is asking for a quite high amount (120k GNO - at current market price >$20m (!))

To personally get behind this proposal I would like to see at least 2 changes:

  1. lower the amount of GNO. Or at least have some form of success metric that defines and makes the total amount depending on the program being successful.
  2. A clear connection to a Gnosis Trading product. Eventually, that could be GPv2, short-term I think Gnosis Auction would make sense.
    E.g. a form to create liquidity for new prediction markets would be to start them with a large(er) auction. E.g. a specific amount of GNO is given to everyone who takes DAI and creates yes/no outcome tokens and is then putting the yes tokens into a Gnosis Auction selling for DAI.
    With this scheme, we could have very large-scale events (selling e.g >$1M for one event) that could really attract larger players to participate in the initial price finding for new events.

E.g. I could see that there would be a market for “will ETH1/ETH2 happen till the end of the year” - with sufficient GNO liquidity we could certainly get $1m into the auction which then would create a large incentive to participate in this auction and thus create a very strong signal for the actual likelihood of this event. This auction would not ensure that after the auction took place there is continues liquidity on this question but as time passes this question also becomes less relevent. So simply getting a good estimate today already creates a lot of value.

  1. Some agreement/ coordination with Omen. Despite the suggestion above I still think this should be coordinated with Omen and also - as a large-scale liquidity mining program would bring a lot of value to Omen - I think GnosisDAO should negotiate a stake in the Omen project.

Hi Martin,

Thanks for the well argumented post.

I think this proposal, would give Gnosis a lot of power in potential negotiations:

  • In the current environment, incentivizing liquidity is a major part of the value creation. For example Sushi and Balancer respectively allocated 90% and 65% of their token to people providing liquidity. Some people at the dxDAO (note that it’s just a feeling from the forum as no vote has been conducted yet) seemed concerned that there was not much people working on conditional in Gnosis anymore (note that I don’t share this concern as the tech behind Omen was mainly built by Gnosis). Here if most of the liquidity is brought by Gnosis, I would expect the dxDAO to be way more inclined to work with Gnosis to keep this liquidity.
  • The proposal is not Omen specific and could be used to incentivize any conditional token pool (like a Uniswap pool, a Sushiswap combined reward, some pool of wrapped tokens on Matic). So it leaves power to the DAO to assign the allocation points where it believes they may be the most effective.

The absolute amount in fiat is, but the relative amount is actually quite low (8% of circulating supply and 1.2% of the total supply).
Since the product is already well tested and Gnosis is a known brand, I would expect significant liquidity in respect of the GNO given and some APR between 20 and 40%. So this should bring 50 to 100M$ of open interest while general competitors are at lower than 1M$ (and application specific are at around 10M$). That would make Gnosis the main actor in prediction markets and very hard to dethrone.
This would also like spin out some second layer projects (think yearn finance for prediction markets where they would manage liquidity of users, withdrawing before the event is known, re-balance between markets, etc) which would also grow the ecosystem.
Moreover, the extra attention and proof that prediction markets work would probably lead to more people taking positions in GNO. The question is would this effect be greater than the dilution? I think the answer is Yes (but yeah, on that part it would have been better if we still had impact estimate via prediction markets).

I think if the amount of GNO is too low, we may not be able to fully benefit from the self-reinforcing effects of liquidity mining:

  1. GNO rewards attract capital/liquidity.
  2. Capital/liquidity attracts traders.
  3. Protocol use attracts people taking positions in GNO.

It goes back to step 1. but with more value.

If the % is too low liquidity mining and its resulting use would not consist of a significant part of GNO price which may lower this virtuous circle effect.

On the success metric, even within this proposal it’s perfectly possible for the DAO to change the issuance (vote to put all the allocation points to 0 if the program is unsuccessful or make a new one to change the allocation).

I think using non-AMM exchanges would be interesting, we’d need to determine how tokens are allocated in a way it can’t be gamed.

For GPv2, we could reward limit orders depending of how close they are of the settlement price (we could use the equation e^(-k*distance_to_clearing_price)² used by balancer).

For Gnosis Auction, I don’t think it’s currently ready for liquidity mining:

  1. There is no frontend for the sale side.
  2. You can fake liquidity by making an auction and buying it yourself. It doesn’t matter if you overpay because you overpay yourself, for example for a Yes/No token where the winning outcome redeems to 1 ETH you could have auctions of both and buy both of them at 1 ETH, no one would overbid you as those token can only be worth at most 1 ETH.

But we could make it work by having auctions where multiple can join on the seller side (so you never trade only with yourself) and incentivize those.

Yeah, creating some initial auction would be very good to allow initial pricing (because currently LP have to guess what prices would be and can suffer from IL if they get it wrong).

Yeah, I also think that the dxDAO and Gnosis should share Omen as the project is way more likely to succeed if it has more teams backing it (smaller part of a bigger cake).
But would this not work, the Gnosis DAO could easily redirect liquidity mining to other platforms.

Overall if the program as it is written could be improved, I definitely think that it would be way better than nothing/waiting. Gnosis had quite a head start compared to most competitors (the tech is there , the brand is recognized as serious and its applications as safe, the funds are there) but we are already seeing competitors eating its launch (see Cover which is basically an application specific prediction market which has around 10M$ of liquidity mainly due to liquidity mining). If the prediction market segment (which is supposed to be the core of Gnosis) is abandoned other projects will come and take it.

Yeah, you are exactly right - this is what I had in mind. One large pooled sale where a fixed number of GNO is distributed to all that participate in the sale.
While it is true that sellers can also act as buyers they still would need to decide on a buy price. If they set it too high they risk buying from other sellers at a too high price. If they set it too low they will not buy at all.

Just yesterday we released a UI to start auction (A safe-app). Given the scale of this program, we certainly could quickly develop a slightly modified version that allows pooled sales.
cc @auction_master @cmagan

Hi, pulpmachina here, working with DXdao on Omen’s upcoming token launch and liquidity mining program.

Agree that the proposed numbers are quite high in $ terms, but not so much in protocol voting power terms. When looked at it as a way to decentralize the protocol, 1.2% of total supply is very conservative. Still an initial 30M allocation is high. Half that amount may be a more appropriate place to start, and the DAO can extend the amounts upon program success.

Regarding Omen’s role in all of this,

  • DXdao is launching the OMN token & planning to give away 20% of the protocol to the community through liquidity mining and airdrops in the first year. More details will be discussed soon on DXdao’s forum.
  • OMN token liquidity mining could sync nicely with a GNO program. DXdao could help do the heavy lifting by automating GNO rewards and integrate it seamlessly in the existing UI rewards.

There’s lots of areas for collaboration between DXdao and Gnosis on Omen. Integrating GPv2 into Omen is beneficial for the conditional token framework, Omen, and Gnosis protocol. A DXdao / Gnosis partnership should be discussed for these longer-term goals.


I think I made the vote in time…not sure…but will definitely be participating in much more!

Interesting. Looking into these proposals now.

The proposal, despite getting 82% of the votes failed to reach the quorum. I think we should work on a new proposal.