GIP-145: Should Gnosis DAO run a 9‑month advisory futarchy pilot with $100k temporary liquidity?

[kelvin]

Should Gnosis DAO run a 9‑month advisory futarchy pilot with $100k temporary liquidity?

• In Favour
• Against

GIP: 145
title: Should Gnosis DAO run a 9‑month advisory futarchy pilot with $100k temporary liquidity?
author: kas.eth
status: Draft
type: Funding
created: (2025-12-04
duration: 9 months
funding: $100k

Category

Service Agreements

Summary

Futarchy Labs would like to propose a futarchy pilot to GnosisDAO.

• The pilot will estimate the impact of different proposals on the GNO token price.

• Based on the integration with Snapshot , a small widget will show up on each official proposal. See live example here. Voters can use this information, if they consider it useful.

• The following signals will be generated:

  • Snapshot will display the current spot conditional prices (YES and NO), and the impact defined as the percentage difference.
  • A 5-day TWAP may be calculated when sufficient liquidity is available, and a recommendation if absolute impact exceeds 1%.

• To provide minimal liquidity to such markets, we request 100k USD equivalent in GNO + WETH to a 2/3 multisig, to be used exclusively for providing liquidity, and to be returned within 9 months.

• Anybody will be able to trade to modify these estimates. Those in favor of the proposal can buy the “YES market”, or sell the “NO market”. See live market here on futarchy.fi. Alternative interfaces can be created: see this alternative interface created by @mkoeppelmann using CowSwap.

• These conditional markets will be implemented in Gnosis chain, using Swapr or some alternative agreed by the Gnosis representative, and are implemented using the Gnosis Conditional Token Framework (CTF).

• The pool created and tokens used are composable with CowSwap, and therefore users can use the Cow interface to place limit or TWAP orders, by adding manually the right conditional tokens. We plan on integrating CowSwap limit orders in the Futarchy interface during the pilot.

Justification

In doing so, GnosisDAO would join Kleros and Velora in officially adopting a futarchy experiment in partnership with Futarchy Labs.

Due to the foundational role of Gnosis in creating the prediction market ecosystem through the Conditional Token Framework (CTF), as well as Gnosis’ pioneering experiments with futarchy, we believe it would be fitting for Gnosis to become the first large DAO to adopt some pilot of futarchy governance, and to potentially play a major role in the development of futarchy in the broader Ethereum ecosystem.

There are some bugs on the interface, which is open source. We invite the Gnosis community to contribute to bug reports and issues, and to create alternative futarchy interfaces, as @mkoeppelmann has done. This pilot will allow us to learn how to make futarchy interfaces better and more intuitive.

Use of Funds

• The 100k USD equivalent may be used solely for the purpose of providing liquidity to conditional markets and associated prediction markets.

• Funds will be provided around equally in GNO + WETH, and the WETH amount may be converted to xDAI or sDAI for the purpose of creating markets denominated in these assets.

• The GNO will be split into YES_GNO + NO_GNO using the conditional token framework, and similarly for the currency token, for the event corresponding to the proposals being evaluated. The oracle used will be reality.eth, with Kleros used for any required arbitration.

• Funds will be used to provide liquidity on conditional pools such as YES_GNO / YES_sDAI, or NO_GNO / NO_sDAI.

• After the oracle resolves the event, the assets are redeemed to regular GNO tokens + currency tokens. The impermanent loss of the whole operation is equivalent to just providing liquidity in a regular GNO/currency liquidity pool.

• For deeper liquidity and more capital efficiency, we request the permission to use the available funds to create concentrated positions of up to 5x. Any rebalancing performed will be reported immediately.

A smaller amount of liquidity may also be provided for the following markets:

• Up to 20% of the allocated capital may be used to provide liquidity on the impact of accomplishing milestones, rather than the impact of approving proposals, on the price of GNO, as authorized by the Gnosis representative.

• Up to 5% of the total allocation of resources may be used for prediction markets associated with the proposals passing. This makes it easier to arbitrate between the spot and the conditional markets.

Pricing and payment

Futarchy Labs will charge no fee or payment, and will only request that funds be allocated to liquidity, as described above.

• Some fraction of the LP fees only (never principal) may be paid to Snapshot Labs as part of the partnership. Any remaining fees will be returned to the DAO.

Terms and Exit Strategy

The estimated impact estimates will be purely advisory, and do not in any way affect the regular governance of GnosisDAO. The DAO reserves the right to end the futarchy pilot at any time, and request the return of all remaining funds.

Team / Organization

This pilot is organized by Futarchy Labs (follow us on X).

Kelvin Santos (follow me as kas.eth) is the CEO of Futarchy Labs.

Robin Hanson (follow him here) is our Chief Scientist.

Success Criteria:

• One measure of success is whether the conditional markets produced sensible, non-random signals. We will evaluate accuracy by computing the 5-day TWAPs for YES, NO and the GNO spot, and comparing these to the 24-hour TWAP of the actual GNO price after the proposal outcome (simple T+24h benchmark). This helps measure whether YES/NO conditional prices anticipated the direction of the actual price movement.

• Another measure is whether DAO members found those signals helpful or at least thought-provoking in making decisions, and if the DAO is interested in continuing or expanding the use of futarchy; for example, moving from advisory to more formal integration for certain proposal types (potentially a future discussion akin to Kleros’ approach, but only if the data supports it).

• Naturally, a third measure of success is whether we have significant user activity in the conditional markets (e.g. >10 unique users per market, >20k volume per market). But overall we expect this product to be more appealing to sophisticated investors, rather than retail speculators.

Conclusion

Near the end of the pilot, Futarchy Labs will deliver a comprehensive report to GnosisDAO, which can then decide on next steps.

[staworth]

Thank you for your proposal Kelvin! We’re very excited to support this, and get involved with this futarchy pilot programme. For anyone else looking for more background, we thought your talk at DappCon this year was an excellent intro to futarchy.fi, its ambitions and its deep ties with Gnosis. We’re excited to see the impact that this may have.

[Netrim]

On a personal level I am interested in seeing how Futarchy can function in this case.

From an operational perspective if the WETH is to be sold to xDAI/sDAI, wouldn’t it make more sense to just request GNO and xDAI?

[mkoeppelmann]

Excited about this proposal.
Futarchy has always been an interesting concept - over the years we regularly tried to make it happen already (even started an implementation for the infamous “TheDAO”).

That being said - now it seems FINALLY the time is right with prediction markets getting significant traction and the “Conditional token framework” playing an important role in it (Polymarket) it seems like a good next step to try those slightly more complex markets. For GnosisDAO it is a low risk proposal that could ideally lead to significantly better governance.

[john_szczepaniak]

Thanks for sharing your proposal @kelvin . When you’re ready to move this proposal to phase-2, please follow the instructions here in the DAO Docs. If you have any questions, please don’t hesitate to ask.

[refri]

Seems to be a great experiment…have to dive a bit deeper but most likely I will vote for. Thanks for this proposal!

[kelvin]

We appreciate the support.

@Netrim, the snapshot widget currently shows USD value, so it is already being converted from sDAI. However the yield is not significant as proposal markets only last for ~2 weeks, so we are considering just using xDAI instead, and that would remove the need to convert prices to Snapshot.

At the same time, some traders in favor or against proposals want to exchange ETH for GNO rather than stables for GNO. We haven’t decided yet which way to go so it would be good if we could convert later depending on which currency is used.

Now regarding the plugin. As it is purely advisory, Snapshot has the opportunity to enable it already, even before GnosisDAO vote, with third-parties providing liquidity. We have experimented with this, and the plugin was showing in the current vote for a few hours.

I think that would be interesting, as it would enable the Gnosis community to see what advisory futarchy looks like in practice. We would like to hear what you all think about this.

[NolanV]

Let me start by saying that I love the concept of futarchy; I was devouring the blog posts written by Gnosis years ago. However, I don’t understand what would create an incentive to use it.

Let’s take the example of ending the contract with kpk.
Imagine I think that kpk is a net negative for the value of GNO (this is hypothetical, of course ), but for one reason or another, I don’t think the vote will pass.
So I should take a position by selling NO, is that correct? If so, why would I take on this liquidity risk, smart contract risk, and uncertainty about my allocation, for a gain of <+0.5%?
For there to be an incentive to bet, there would need to be an imbalance to generate larger gains, but I don’t understand who would bet at a loss. Maybe, one would have to believe in the absolute efficiency of markets, which, especially in crypto, seems tenuous to me. We’ve already seen protocols gain a few percentage points while in the middle of being hacked.

Do you have any articles that quantitatively analyze the value of this type of market?

[0xAlex]

At Kleros Cooperative, we strongly support this proposal and will vote FOR.

Gnosis is foundational in the history of prediction markets and we believe it’s time to start experimenting with Futarchy at the DAO level. The non-bidding advisory Futarchy approach outlined here is an excellent way to do so progressively and build on real learnings.

We have been running Futarchy.fi for several months at Kleros — both as an advisory tool and as a safeguard mechanism for sensitive proposals.

Advisory Futarchy in practice

• Example of a controversial, non-consensual proposal where token holders were clearly split: the accompanying Futarchy market showed a bullish outlook for PNK (Kleros governance and decentralized court token) with expected positive price impact. This gave uncertain voters additional signal if their goal was to maximize token value.
  → See the proposal and the market referenced in this tweet.
• Conversely, another proposal was rejected by the DAO where Futarchy markets forecasted negative PNK price impact — some voters even explicitly cited the market signal in their rationale.

Safeguard mechanism

We passed a governance change that adds an extra condition for any proposal requesting PNK minting: it must win the regular Snapshot vote and the associated Futarchy market must show a positive expected price impact. This acts as an additional protection for token holder value — something Gnosis DAO could also experiment with in the future for controversial proposal involving GNO minting or large spending expenditures from the Treasury.

Because we strongly believe price-based Futarchy is a superior decision-making mechanism that better preserves and grows value for GNO holders, we will delegate a substantial amount of voting power that will vote strictly according to the Futarchy signal (using a 5-day TWAP window):

• If Yes price > No price → vote FOR
• If No price > Yes price → vote AGAINST
• Unlike what happened in the recent GIP-143 Should GnosisDAO Terminate karpatkey Treasury Management Services?, we will not ABSTAIN even if the price signal is low — we want to encourage adoption and market participation.

This will be executed manually at first (via my personal account) and could be automated later. Importantly, anyone trading on the Futarchy markets will effectively influence how this delegated VP votes — making the experiment even more powerful.

[kelvin]

Hi @NolanV. Let me answer your questions!

In your hypothetical, yes you could (and maybe should!) take a position by selling NO.

This would indeed be only for a gain of <0.5%, but only if you think the proposal is a net positive but only slightly. There are arguably many proposals where the impact could be much higher, >10% or more!

Very recently, we had liquid markets on Kleros where the absolute impact of a proposal was frequently trading at >4%, (see post by @0xAlex) and I think we’ll eventually see that on some controversial (and important) proposals for GnosisDAO as well.

Regarding the risks. Even though we could do a better job at making this very obvious and easy to verify, the main smart contract risk should be relatively low.

All assets are stored in the GnosisCTF contract, which is the same contract used by Seer, Polymarket and most other prediction market platforms, for the last 8 years. I believe the risk of a problem in the main custody of funds is minimal.

The assets are split, merged and redeemed using the Seer FutarchyRouter contract, so there is also some risk there, as these require allowances. Even though these contracts are very simple, and don’t maintain ownership of any funds, they haven’t been heavily audited yet.

There is the risk of splitting and trading the “wrong” assets – although it can be done, it is currently a bit hard to verify yourself whether the YES and NO assets you are getting really correspond to the question/event you are supposed to trade on, so most users are relying on our front-end for that. If this was done maliciously, funds could be locked.

Finally, there is the possibility of Oracle risk, if you don’t trust Kleros or reality.eth to correctly decide the winning branch based on what actually happened. This risk, however, is limited, as long as you trust the oracle to at least resolve quickly, because in futarchy markets, compared to prediction markets, a malicious oracle cannot really steal the principal (e.g. if you’re trading GNO for sDAI, a malicious oracle can at most decide whether you get the GNO or the sDAI amount, but cannot really take both from you).

Futarchy markets for proposals have one advantage: they are likely to end very soon (in a week or so!). So the liquidity risk is limited. So my hope is that it would indeed be worthwhile to trade, even for a 0.5% gain! Would also be interested to hear what could make futarchy more appealing to trade on.

I thank @0xAlex and the Kleros Cooperative for their support. Kleros has been not only a strong partner as our resolution oracle (alongside reality.eth), and as a technical partner (through Seer), but also as a very bold adopter of futarchy on its own governance.

We applaud Kleros for their courage in giving veto power to markets even at this early stage.

[Chim9]

GIP-145 is live on Snapshot : Snapshot

[kelvin]

There are two more days for voting on the snapshot proposal. We haven’t reached quorum yet, so please vote!

As we mentiond on the snapshot proposal, we also created a market on the conditional price of GNO is the proposal is approved vs rejected – go ahead and trade here!

This also included a 5-day TWAP period before the voting started, and the TWAP of the YES market over 5 days was ~2.5% higher than the TWAP of the NO market.

This should be seen more as an example of how these signals will work, due to limited participation for now. But anyone can trade and sell GNO at a small premium to spot, conditional on the proposal passing, or buying GNO at a small discount conditional on the proposal failing.

If the proposal is accepted, such markets will show up in snapshot, see example for a Kleros proposal.