DXdao Proposal: Delegate Treasury Management to Karpatkey DAO

DXdao Proposal: Delegate Treasury Management to Karpatkey DAO


Below is a proposal to have Karpatkey DAO manage a small portion of the DXdao treasury (which is currently idle) to take advantage of a number of investment opportunities that will result in the growth of DXdao’s treasury, to fund its operations, development and growth.

Karpatkey DAO’s AUM are $550M+. They manage treasuries without holding custody of funds, using the Gnosis Safe multisignature wallet. Therefore, even though Karpatkey initiates transactions, every single one of them would need to be approved by a trusted DXdao signer for the funds to be actually transferred or deposited in a smart contract. Therefore, it is not technically possible for Karpatkey DAO to transfer DXdao’s funds, unless the transaction is authorized by DXdao. Nonetheless, DXdao will always be able to transfer its funds without Karpatkey DAO’s intervention.

Currently, the minimum treasury size Karpatkey DAO handles is of $100M, but a 6-month pilot test would be conducted exclusively for DXdao, handling $10M.


After the successful execution of the Treasury Diversification Authorization Program and the Treasury Diversification Authorization Proposal executed Jan 23, 2021, DXdao’s treasury currently has a significant level of diversification.

However, the vast majority of its funds are liquid and therefore not obtaining any yield. This results in a considerable cost of opportunity, since adequate treasury management could obtain a significant APY, achieving a sustainable treasury growth that would increase DXdao’s reach and breadth.

This proposal is aligned with the already executed Signal Proposal: Treasury Diversification Authorization Program - v2, and suggests that the management of a portion of DXdao’s treasury be delegated to Karpatkey DAO.

Karpatkey DAO specializes in professional treasury management, and has had a remarkable performance managing Gnosis DAO’s $550M+ treasury portfolio, sustainably increasing its size, profitability and diversification week after week since its inception in May 2021.

The benefit of such delegation would allow DXdao to focus on its core business while its treasury size increases steadily, increasing the DAO’s potential and capability.
The treasury goals are to fund the development of DXdao products, new ventures and boost DXD’s value proposition.

Karpatkey DAO will support the efficient achievement of these objectives by providing DXdao with a selection of market opportunities, and actively engaging with other protocols to achieve strategic partnerships through token swaps.

Karpatkey DAO’s goals are the following:

  • Design and implement an efficient and robust treasury management system
  • Identify the best economic opportunities for the treasury tokens
  • Facilitate DXD liquidity
  • Grow activity on Swapr

To align with DXdao’s vision and ethos, and increase transparency with DXD token holders, we propose that information about how investments are managed is made available to the DXdao community. Treasury management would be handled in a completely transparent way, positively impacting DXdao’s reputation and trustworthiness.

Past Performance

The inception of Karpatkey DAO as Gnosis’ treasury management consultant has resulted in a consistent revenue stream for Gnosis, increasing its treasury size by circa $1.5 M a week (YTD tracked since May 2021 includes but is not limited to: 71K GNO, 350K BAL & 20K FLX).

Karpatkey DAO has continuously increased the treasury holdings and the amount and complexity of positions while significantly reducing their size and associated risk.

This is illustrated in the chart below:

The number of positions was increased more than 3X, significantly reducing their size and associated risk. The improvement in capital management efficiency has entailed a substantial workload increase.

History of transactions for the main DEFI multisig wallet for the last year where a substantial increase in transaction volume can be seen as of May 2021.


As part of an initial 6 month pilot test, DXdao would Transfer to a Gnosis Safe multisignature wallet a total of $10M by sending equal parts (calculated in USD) of the following tokens: DAI, USDT, USDC, ETH and DXD.

The Gnosis Safe multisignature wallet will have a 3 out of X setup, so that Karpatkey will never hold custody of DXdao’s funds. Transactions will only be executed if DXdao’s trusted signer signs them. Also, the funds will be accessible to DXdao at all times, without any need of Karpatkey DAO’s intervention.

Karpatkey DAO’s professional treasury management carries out tailored strategies that vary twice a week, adjusting to the constantly changing DeFi market in a way that other investment alternatives such as Compound, Aave and Yearn can’t keep up with.

Karpatkey DAO’s success in handling DXdao’s treasury will be assessed 6 months after the beginning of their treasury management. If DXdao wishes to discontinue using Karpatkey DAO’s services, it will do so through its governance mechanism. If not, Karpatkey DAO will manage the rest of DXdao’s treasury, maximizing its growth.


  • Swapr: Karpatkey commits to making Swapr a core part of its treasury management strategy for DXdao across Gnosis Chain and Ethereum Mainnet Karpatkey will manage, among others, the following assets (DAI, USDT, USDC, ETH and DXD) and deploy them to the following pools, among others:
    • Swapr
    • Curve
    • Aave
    • Balancer
    • Uniswap v3


Karpatkey DAO will charge the following fees:

Management fee

A yearly management fee (in monthly installments) of 0.5% of the AUM would be charged to finance the services detailed above.

Performance fee

A monthly performance fee of 15% of the yield obtained during the last month.


DXdao may terminate Karpatkey DAO’s engagement under this agreement for any reason by way of Governance Mechanism. Kapatkey DAO may terminate this agreement upon four week’s’ notice posted as a new discussion thread in the DXdao forum. In case the termination is requested by DXdao, Karpatkey DAO will be granted an exit fee equal to the fees collected during the last 2 months, which will be paid in DAI. Regular fees will be collected until the day of termination.


Karpatkey DAO will provide a weekly treasury management report and support in identifying data elements upon request from the DXdao’s community. Also, Karpatkey DAO would review data reported and assist in performing analytics and quality reviews to confirm the accuracy of the data.

Custody of Funds

Karpatkey DAO is not a custodian of funds. The funds will be in possession of authorized REP holders of DXdao at all times.

Karpatkey DAO will sign transactions using two addresses TBD.

Likewise, Karpatkey DAO is not responsible for loss of funds caused by the existence, identification and/or exploitation of vulnerabilities through hacks, mining attacks (including double-spend attacks, majority mining power attacks and “selfish-mining” attacks), sophisticated cyber-attacks, distributed denials of service or other security breaches, attacks or deficiencies with smart contracts or protocols which are not owned by DXdao or Karpatkey DAO.

The plans outlined in this proposal are subject to discussion by DXdao and may need to be (re)structured to take account of legal, regulatory, or technical developments as well as governance considerations. This document should not be taken as the basis for making investment decisions, nor be construed as a recommendation to engage in any transactions.

Future Vision

While the initial setup will rely heavily on a trusted multisig, Karpatkey and DXdao agree to work towards building a long-term treasury management solution that would maintain DAO sovereignty over the treasury. This may look similar to the set of relayers that DXdao already uses for Swapr liquidity provisioning. With these relayers, treasury movements occur through on-chain proposals and the normal governance process.

Karpatkey and DXdao will build or partner with others to create similar smart contracts that can facilitate treasury movements. These ‘vaults’ would allow Karpatkey to manage treasury assets in pre-defined strategies, while maintaining DXdao governance’s control of the assets. DXdao will rely on its prior experience to aid Karpatkey in devising the most permissionless and trust-minimized system.

With a trust-minimized system that maintains DXdao’s sovereignty over treasury, DXdao could allocate more of its treasury to the management of Karpatkey.


Thanks for the proposal @karpatkey.

DXdao sits at the intersection of DeFi and DAOs; there is a huge opportunity to build decentralized governance tooling for treasury management that is tightly integrated with DXdao’s DeFi product suite.

DXdao has already pioneered trust-minimized treasury management through its GP Relayer and Swapr Liquidity Provisioning Relayer, which allow DXdao to conduct treasury operations while maintaining DAO sovereignty. Long-term, this is the model that all DAOs will use to manage (eventually) trillions of dollars on-chain.

I’m excited about the possibility to work towards that future with Karpatkey, which have the expertise to do more active treasury management.

There are two things that this proposal would introduce for the first time to the DXdao community that are worth discussing further:

  1. A trusted multi-sig for a portion of the treasury. DXdao prides itself on having its entire treasury controlled through governance. While the vast majority of the treasury will remain this way, this proposal could put up to 25% of the treasury into a trusted multi-sig. @dlabs is working on setting up a new Treasury multi-sig and there should be an additional proposal authorizing its uses and members.
  2. DXdao-supplied liquidity for DXD. While DXdao has close to $5m in liquidity on Swapr, treasury funds have not been used to provide liquidity on DXD. There was a debate all the way back in October 2020 about providing liquidity, but the community ultimately rejected that due to low price of DXD. But now with the DXD buyback program entering its 13th month and the recently passed extension clarifying that the 3.6k DXD purchased by the general treasury “could be used for liquidity provisioning”, this seems less of a concern. Nevertheless, this would be a major step and should be discussed by the community.

Exploration of questions and ideas around Karpatkey DAO’s proposal:

I am in favor of exploring how DXdao can do better treasury management as a proper DAO.

This proposal is a great way to ignite that conversation.

Long-term, I believe that DAOs and, in fact, most (all) capital management by 3rd parties will eventually be done using mechanisms that allow for entities to keep full custody of their funds.

Not having to give over custody of funds to 3rd parties allows for a much more robust and safe financial system.

Mechanics of this MultiSig
(Note: This conversation likely should be isolated into its own thread as @Powers mentions, but it is also important in this context because the proposal is highly dependent on that exact workings of this MS)

The proposal states: “The Gnosis Safe multisignature wallet will have a 3 out of X setup, so that Karpatkey will never hold custody of DXdao’s funds.”

However, DXdao will not hold custody of the funds either. This is a common misunderstanding often in the DAO space, that a few people from the community holding custody of something means the DAO has custody.

We have many examples of DXdao holding custody of assets in the DeFi ecosystem. Yes, we do have examples of DXdao trusting a smaller trusted set of REP holders with limited amounts of capital for short periods of time.

Up to now, the largest amount of value DXdao has trusted to a smaller trusted set of REP holders is less than $1.5m at any time.

So DXdao would need to decide how much value would it trust with a very small trusted set of REP holders (in combination with Karpatkey wallets). Would this be only 1 trusted REP holder that could take action in tandem with Karpatkey holders? Or 2?

If DXdao gets comfortable with giving up custody of funds to this set-up, what would this amount be? In this case, $10m dollars seem very high. $1m to experiment with these new ideas seems more reasonable. At the size, it would be more to see how things work than the yield being seeked. Plus, we could start to learn what tools and practices we actually need to achieve the goals we desire.

This whole situation raises other questions, more for DXdao community than for 3rd parties, around Multisigs in general:

  • How would this type of Multisig be set-up?
  • Who would be on it?
  • What are their responsibilities?
  • What are the risks and rewards to being on this MS?
  • Is there a safety risk?
  • Should these people ever be in the same place together?
  • Is this role a “full-time” job?
  • At what value held in MS do these questions become more important?

There is likely a need for a separate thread to dig into details of how DXdao uses Multisigs.

Size of Yield and Alternatives

What type of yield are we seeking? And what do we compare this against?

If the target yield is 5-9% that is decent.

Currently, via ETH staking, I believe yield can be 4-6%? If so, we are talking about a 1-3% improvement before fees, in addition to giving up custody, and also bringing in potentially many new risks like smart contract risk and other defi-associated risks.

Is that a smart risk-reward ratio?

Active Yield Seeking versus More Passive Yield Seeking

How do we measure the value of performing more active yield seeking versus slower paced yield seeking?

DXdao already has the ability to do yield seeking activities. So far, DXdao has done this conservatively and slowly. This pace and risk can easily be picked up as a DAO.

What are the key benefits to jumping to active yield seeking with a 3rd party?

Other Important Questions

How does the use of DXD fit into this plan/proposal? (Note: Another question that needs debate among DXdao community)

Currently, DXdao does not use DXD in its own products (or others). How critical is this element to the proposal?

Given that platforms like https://www.mimic.fi/ and https://www.dhedge.org/ exist, is there a way to already leverage them today?

If not, what needs to be done? What needs to be built? And how can we incentivize people to build the things needed? :carrot:

Long-term Thinking

Working with a top-notch entity like Karpatkey (who is also part of the greater Gnosis “Friends and Family” ecosystem) is a great idea. If we can all work together to accomplish the shared goals and build the tools necessary to reach those goals, that is a win.

Map and Timeline

One way to get everyone in alignment could be to construct a very clear outline of what the end goals are and what is needed to get there. Without this map and timeline, it could be hard to make sure we are all working towards the same goals.

What is the end goal?

Looking forward to Discussing

I look forward to discussing these questions and ideas with the DXdao community and Karpatkey.


Hi @Powers. Thanks for your comment.
As we mentioned during today’s governance call, we’re all for decentralization and would be happy to collaborate with you in the creation and usage of adequate tools for treasury management integrated with your product suite.

For the time being, we can work with the trusted multi-sig you mention in point 1.
Regarding point 2, we’ll go with whatever you decide, but we strongly recommend allowing us to get the most out of your DXD, increasing its value, using it as collateral to get leveraged loans in stablecoins, and making it a token of choice for DeFi users.

1 Like

Hi @sky. Thank you for having taken the time to put together such a comprehensive reply! This is the most thorough we ever got. We’ll reply under quotes, and try not to leave anything out:

It’s important for us to clarify DXdao doesn’t give up custody of the funds. DXdao would continue being able to use its treasury at any time. Karpatkey holds non controlling keys and doesn’t hold custody of funds. The trusted signers from DXdao would be able to control the part of the treasury that is sent to the multisignature wallet, without Karpatkey’s intervention.

We have been handling the $550M+ treasury of Gnosis for over a year now, with very good results.
We understand your caution when trusting your treasury to a third party, but the allocation of small amounts will reduce the investment possibilities and limit the complexity of the strategies we can apply, defeating the purpose of outsourcing to professional treasury managers. We’d actually suggest not to delegate the treasury management unless you can allocate at least $30M to it. We’ll work hard to provide you with an outstanding performance with a very low risk. We’d lose you as a client if you’re not happy with our results, and we do this for a living.

Treasury management is a full time job, and you practically start from scratch every day, reviewing the figures of all positions, checking if the asset allocation could be improved, looking for new risks or threats to determine if a position needs to be dissolved, getting a sense of the market sentiment in social media, studying the papers and audits of new protocols, coming up with composability strategies, it’s endless.

It would be set up in a Gnosis Safe, with a 3 out of X configuration. Karpatkey would only have 2 signers, and DXDao could have as many signers as you wish. They will need to sign each and every transaction we initiate and confirm.

2 Karpatkey signers and X DXdao signers.

The signers will be responsible for all transactions executed, they will have to double check amounts, parameters and addresses of each transaction, to ensure that there are no mistakes, and strictly follow Karpatkey’s audited security protocol, which has failsafes that have allowed us to operate with zero errors.

The main risk is that Karpatkey and the trusted signer/s come to an agreement to steal the tokens from DXdao, which would go public soon after, making Karpatkey go out of business.
The reward of being on this Multisig is that you get tailored real-time execution, which doesn’t require for a previous strategy disclosure that would expose you to massive front-running, that would fully impact the yield, diminishing it.
Also, the execution team validates that all the parameters and underlying reasons behind the transaction are still valid. Deciding on the spot whether the transaction is still beneficial for DxDAO or if it should be changed, always getting the most out of your treasury. This cannot be done with any automated solutions nor with pre-disclosed investment strategies subject to community vote.
Regarding individual rewards for the trusted DXdao signers, it could be a part-time contribution fee.

Yes, there always are, but we go great lengths to minimize them to the point they’re negligible from an operation standpoint.
From a structural standpoint, the risks are the same that all DeFi users are exposed to: smart contract exploits and general protocol hacks. That’s why we only use reputable battle-tested protocols which have been thoroughly audited by professional third-parties.

No, they have to be in different physical locations, in private rooms, with their hardware wallet, and they must participate in a videoconference with their camera on and state out loud every transaction before executing it. If they’re being coerced, they must use the panic phrase and we’ll terminate the call casually.

No, it will require one or two weekly sessions of between 1 and 2 hours.

For us, the value is indistinct with relation to these questions. I mean, they’re equally important regardless of the amount of tokens handled.

We don’t target a specific yield because it varies depending on the market conditions. We can tell you however that the yield we have obtained for Gnosis DAO has been greater than the yield obtained by all the rest of the DAOs put together (source).

We don’t know of any other player in the ecosystem who applies strategies similar to ours to which we could compare in fairness, so we’d compare against GnosisDAO’s treasury performance.

Via ETH staking, you could get a yield of circa 4%. If we manage your treasury, not only you could get more yield, but also you would be exposed to significantly less risk, since we would diversify the token allocation above and beyond any other solution in the market. This would result in minimized exposure to smart contract and DeFi-associated risks, since a hack would only compromise a single position, which would represent a fraction of the whole. On the other hand, if you just stake all your ETH, a hack to that position would drastically reduce your treasury size.
It’s worth noting that in the hypothetical case in which you were investing $73M of your liquid treasury, a simple 1% would yield $730,000. This is not meant to state the obvious, but to adequately dimension the order of magnitude of the operation.

We don’t doubt DXdao has the ability to do yield seeking activities, but we disagree that the pace and risk could be easily picked up. Our strategy team is constantly monitoring the market for investment opportunities, and checking the soundness of the positions we’re already in, and looking for composability alternatives to get the most out of the positions.
Our development team keeps our anti-liquidation bots sharp so that we can maximize capital efficiency and get stablecoin loans that allow us to get yields and increase treasury asset diversification without negatively impacting the price of DXD. They also maintain our early alerts system to detect behaviors that allow us to anticipate market reactions in situations when minutes count.
We also need to constantly monitor social media, Discord and Telegram to be able to get the pulse of the market sentiment and keep up to date with the current trends.
We have early alert systems and an emergency protocol (revised monthly) which allows us to take immediate action in case of events such as a drastic drop in token prices or potential depegging of Stablecoins. This protocol has allowed Gnosis DAO to suffer zero losses in the market crisis of May 12. DAOs currently don’t have the necessary agility to react quickly and would be exposed to a full blown impact in case of a fast market free fall.
This is just a very summarized detail of what we do, and every day, we’re back to square one. These tasks have to be done systematically and constantly to keep up with the market.
We also advocate for our clients to have their tokens accepted as collateral in major protocols, we negotiate on their behalf to conduct token swaps and improve the positioning of their DAO and token.
The benefit of outsourcing active yield seeking is not only that you consistently get the most out of your treasury, but also that you can forget about it and just focus on the core business of DXdao, which has nothing to do with coping with the dullness and endless demand of finance. You can focus your efforts on what you do best, while your treasury grows and provides you with funds to do it.

We plan to increase the value and availability of DXD across the entire ecosystem. We’d initially use it in token swaps and OTC deals, we’d also work to make it acceptable collateral and then take loans of stable coins against it, further diversifying the treasury and investing the stablecoins to leverage the loan and grow the treasury.
Using DXD is not critical to the proposal, as we could use any other tokens you hold, but we strongly recommend you allow us to get the most out of it as described above.

No other alternative in the market offers services with a similar level of diversification and complexity, and Karpatkey’s operation is not exposed to dapp failure or hack risk, since it’s currently manual. Therefore, our solution offers not only greater yield, but also less risk.
However, we’re working on dapps that will allow us to subject investment strategies to the governance mechanisms of DAOs (such as Snapshot), and require no signature from trusted members of the DAO. We’re actually using the dapps for certain Gnosis operations, and providing feedback to the development team twice a week.
Also, we’re working on a tool to automate the execution of the investment strategy, which will allow us to scale our operation.
Before we start using these tools in live environments, we’ll run thorough tests in test environments and then start using them with small amounts for several months, until we feel confident enough to have them pentested and audited by third parties. Only at that point will we deploy and announce them.

For us, the end goal is to consistently grow DXdao’s treasury and allow it to finance its entire operation with the yield we’ll get from it. We could put together a document detailing the stages we’ll go through to get there, but we think it might be better to post it in another proposal after you’ve seen the results of our work.


This was discussed on yesterday’s Governance discussion starting at 11:17


Hi team. Thanks for all the answers.

If we want to achieve some collaboration while maintaining certain key principles of DXdao, I’ve been thinking more about how we can do this. We want to design some set-up where DXdao and Karpateky DAO can work together.

Karpatkey DAO has a desire to use a Gnosis Safe to manage actions. It is what Karpatkey is comfortable with and a proven, safe, flexible tool.

DXdao has very big questions/concerns (and no experience) with turning over control of funds to a single (of X) reputable DXdao REP holder.

So let’s explore the use of a Gnosis Safe more.

See Lukas’s recent thread. Safe is already designing for this:

“3) Treasury Manager: Third-party that is able to manage assets on a Safe but restricted to certain operations (e.g. pre-defined functions on specific DeFi protocols).”

Safes have the ability to have that single additional signer be a DAO and not a person. In the past, I have raised the need to look into this further with the Zodiac/Safe teams to see how DXdao or a new scheme can be that Safe Signer/owner. Now is a great time to do it.

Combine this with new capabilities of DXvote - something like a 6 hour or 12 hour start-to-finish proposal scheme specific to this new set-up.

We now have a relatively quick trustless DAO-owned treasury management system tied in with the experts at Karpatkey.

The period for how fast these proposals should be passed needs to be discussed further.
If the person (or people) as a signer or signers are sleeping or busy, even regular people could take 6 or 12 hours to “audit” and sign the MS txn. In this case, the DAO is the final signer.

All this tech ALREADY exists. It just needs to be assembled properly.
Yes, we are trusting Gnosis Safe smart contracts, but the Safe is controlled/owned by DXdao.

Safes have the ability for DXdao to sign messages on Gnosis Safe: A signature can be made using an onchain transaction, meaning that any smart contract could in theory be a signer in a Safe.

We have all the pieces we need.

Would Karpatkey be open to exploring this path?

1 Like

Hi @sky, thanks for getting back to us.

We currently are working with the Zodiac module and providing feedback to their development team twice a week to iron out some wrinkles. The solution is great, it’s supported by very talented developers, and will be a standard in the future; but we’d rather further battle test it before we consider it as reliable as human signers.

The next step in our roadmap is to offer a fully decentralized solution for DAOs, that will allow them to vote on strategies as you describe. We are working towards this not only because it’s aligned with the ethos of the industry, but also because it’s THE best way to scale our operation: allowing us to provide trustless services to DAOs that will be able to use their standard governance mechanisms to collectively decide which strategies they will allow us to carry out for them.

Given the cost of opportunity of waiting for the necessary dapps and smart contracts to be mature enough, we are proposing an interim solution for DXdao that will allow you to get the most out of your treasury in the meantime.

We understand your concern, and we are willing to walk you through the entire process, operating transparently in execution sessions where you can invite as many trusted members as you feel comfortable with, to oversee how we manage your treasury. Your trusted signers would have to authorize every single transaction, and you could rotate them frequently if you wish.

We would like DXdao to be a success case for Karpatkey, so we’d do our best at managing your treasury, for which we have an investment strategy outlined that we believe you’ll be very pleased with. We suggest to start operating with trusted signers from your end, and to switch to a trustless model driven by DXvote as soon as the module is mature enough to be considered trustworthy.

That way, when the moment of switching to a trustless operation arrives, your treasury will be bigger, more diversified and safer than it would be applying other alternatives.

How does it sound?