[SEP 24] karpatkey - SAFE DAO Treasury Management Core Unit

Thank you to the @karpatkey team for bringing this proposal proactively into the DAO. The way we will deploy the SafeDAO treasuries is indeed something that needs to be considered.

However, the proposal was posted 8 days ago and some questions, many of them raised by @AccelXR-1kx, are not yet sufficiently answered in @felix_greenfield and my opinion.

We believe that we need to first take a step back and define the intended purpose of the joint SafeDAO x GnosisDAO treasury in greater detail.

As reference, @lukas posted the following to the GnosisDAO forum in his Feb 22 GIP29 proposal to spin out Safe from Gnosis:

“Joint GnosisDAO <> SafeDAO Treasury

5% [of the SAFE token] will be allocated to a joint treasury, controlled by the GnosisDAO and SafeDAO in a 2-out-of-2 Multisig. This joint treasury will be used for rewards that are in the mutual interests of the Safe and Gnosis communities. For example, rewarding Gnosis Safe users on Gnosis Chain leading for increased adoption of both.”

Also, we would highlight that it should make a significant difference as to whether those treasuries consist of native Safe tokens or of assets in some other form. In the latter case active asset management could make sense, while in the former case we first need to decide upon either 1) diversifying out of the native Safe token for greater resilience against volatility in this one asset 2) spending the native Safe token on certain types of initiatives or 3) a mix of both.

We prefer re-opening discussions on the topic to debate SafeDAOs goals regarding treasury management and looking for a suitable service provider afterwards through a request for proposals.

The time reserved for discussions for this proposal was not sufficient to form an educated opinion on such an important topic, especially when 0.1% of the full token supply of a protocol is at stake each year in the form of a fee for the service provider. Additionally, the missing of a clear objective for treasury management and the lack of looking at possible competitors led us to reaching the following decision:

Greenfield will vote to make no changes on SEP24.


We recognise and appreciate the detailed response from @AccelXR-1kx and @Georg_Greenfield. This is an important proposal for Gnosis and SafeDAO, and this level of detailed feedback is crucial and welcomed.

There are a lot of shared questions around the following topics, so we aim to address them in a consolidated format below:

Treasury Management Options & Counterparty Assessment

From our perspective, karpatkey is the nearest to an in-house service provider. karpatkey conceived from the efforts of the Gnosis DAO with GIP [20] in 2021. We have an unwavering and long-standing commitment to the success of both the Gnosis and Safe ecosystems. Additionally, these ecosystems are foundational to our non-custodial treasury management services, and we have a deep knowledge of the tech behind the Safe product suite.

We caution against the notion that the DAO can spin up an in-house treasury management team from individual contributors. Without much examination, this may appear advantageous from an overhead cost perspective. Still, unfortunately, this is fraught with management risks that could lead to a net capital loss in the long run. karpatkey has teams dedicated to the active management of our treasuries who are supported by our risk and tech teams, allowing us to offer a full-ranging service. Despite multiple hacks in the industry and fluctuating markets, we have experienced no capital losses in any of our treasuries.

karpatkey’s understanding is that SafeDAO needs more than a treasury manager; it requires a reliable service provider who can tackle other needs. Some related to the $SAFE token, such as protocol integrations, and others like advisory work on investments aimed at strategically allocating SafeDAO’s resources to foster innovation, support community projects, and invest in opportunities that align with SafeDAO’s mission and objectives. karpatkey’s expertise in working with leading DeFi protocols like Aave, Balancer, Lido, and ENS, in addition to Gnosis DAO, will be of great value when addressing these initiatives.

Regarding counterparty risks, karpatkey leverages our on-chain non-custodial asset management infrastructure to mitigate these issues. On the one hand, the assets remain under the control of the DAO and, on the other hand, it is possible to swiftly remove them from any DeFi protocol they may be deployed, thanks to our solution, which builds upon Safe.

Terms of Engagement and Responsibilities

We structured the terms of our proposal so that SafeDAO and its token holders have the ultimate authority to end our service provider agreement at any time. If our performance is deemed unsatisfactory or any other arbitrary reason, they can remove us from the treasury safes via the governance mechanism. This is also a significant concession that gives the DAO and its token holders power over the proposed agreement. SafeDAO and GnosisDAO retain complete control over the funds at all times, with the joint treasury being moved to a 3/6 Safe; karpatkey will only have two signers, with the remaining four signers to be appointed by the Safe Foundation and GnosisDAO. SafeDAO and GnosisDAO also control the Safe through the Safe Module installed on the joint treasury.

The mention that the agreement is open-ended minimises the above concession and implies that we will have diminished oversight. To be clear, we expect that the SafeDAO, its token holders and investors will play an active role in treasury-related discussions.

Our non-custodial treasury management method is public and open and inherently lends to an active relationships with the DAOs we serve. Our active engagement is demonstrated as a reality throughout our existing service provider agreements, where we have worked with community requests, and by authoring proposals which were ratified via established governance processes.

To adddress our responsibilities, in our proposal we state that karpatkey’s responsibilities will be contributing to the following:

  • Treasury management
  • Protocol integrations for $SAFE
  • Risk management
  • Research

With the passing of this proposal, the signers responsible for this joint treasury are intended to prioritise and develop actions that correspond to these responsibilities. The initial will likely goals will be focused on token utility and liquidity provisions.

Fees and Payments

Karpatkey has proposed SAFE, ETH, and Stables as payment options for fees. However, we’d note that being long-term aligned partners is crucial for any of our engagements; therefore, we would not sell SAFE simply to cover our fees and would be happy to accept SAFE directly as payment—as outlined in the proposal. The payment options are simply to provide a set of options for when the treasury is more diversified.

With the SAFE token currently in a non-transferrable state, with no price and no liquidity, uncertainty makes it difficult to predict the optimal fee structure; we feel that the fee structure outlined in the original proposal is appropriate currently. As previously stated, the DAO can change the agreement at any time, and we are willing to be proactive in addressing new fee structures, such as moving to more performance-based fees as the treasury develops.

Alternative Processes

An RFP process in the SafeDAO is challenging because this framework does not exist. The only established process we know of is the Outcomes-Based Resource Allocation (OBRA) framework. Also, the SAFE token transferability has unanimous support, which imposes a more immediate need for treasury management.

The immediate assets to be managed under this proposal are those of the joint treasury between SafeDAO and GnosisDAO. Given that karpatkey is already the treasury manager of GnosisDAO, thereby already holding half the delegation of the joint treasury, we believe karpatkey is a natural fit for the delegation. Having two separate treasury managers for the joint treasury would likely result in more friction and inefficiencies.


Already being voted on? This SEP process (posting to voting) moved way faster than I thought it would.

Three days after this SEP was posted Safe Guardians and delegates with sufficient voting power began to agree that this SEP was ready to be sent to Snapshot for a vote.

What happened to OBRA?

The transferability of the $SAFE token took what ~18 months of discussion? This SEP is being rushed through with minimal discussion.

Not one person knows how much this SEP will cost SAFE DAO per month. No one knows how much $SAFE will trade for, so I’m not sure how anyone can make an informed decision on voting yes or no on this SEP. I don’t think this SEP should have been sent to Snapshot when $SAFE still can’t be transferred.


is there a policy that regulates the time between posting a proposal to the forum and it going to snapshot?

because i share the sentiment here with others that sep24 requires more discussion.

up until now, i have really enjoyed steps within this doa being taken slowly but deliberately. not sure why we are not seeing that here now


Hey @0xSafeUser. Yes, this proposal has moved to Phase 2 Voting on Snapshot.

This proposal proposes delegating management of the joint treasury between SafeDAO and GnosisDAO, which does not fall under the scope of OBRA (read more here).

1 Like

Hey @gosuto

All proposals must be on the forum for review for a minimum of 1 week and can only move to vote with the signaling of at least 3 Safe Guardians/Delegates with sufficient voting power.

How the governance cycle works:

Each governance cycle runs in 4 week sprints. Each cycle consists of an optional Phase 0 Discussion (1 week), required Phase 1 Draft (1 week), and Phase 2 Voting period (13 days).

  • Phase 0 of pre-discussion is optional, although recommended for the reason you’ve mentioned.
  • Therefore the maximum time a proposal has for review from the community is 2 weeks and the minimum amount of time is one week (under Phase 1: Draft). In this case, karpatkey skipped Phase 0 and submitted on the first day of Phase 1.
  • There was a community call to review all proposals in Phase 1, including karpatkey’s.
  • An additional checkpoint is the signal from at least 3 Safe Guardians/Delegates with sufficient voting power to signal if the proposal is ready to move to a vote. (Gentle reminder, this is only a signal to mark ready to move to voting. Signaling does not equal endorsement) In the case of this proposal, it was sufficiently reached on the day of the deadline.
  • Authors can update/incorporate the proposals with feedback, but the proposal will be frozen on the deadline of Phase 1 / when it is submitted for voting on Snapshot.
  • FYI: Any proposals that are submitted after the Phase 1 Draft are only available for the next Sprint (e.g. next sprint starts on April 15th and earliest can go to a vote is April 29th)
  • You can subscribe to the governance google calendar here or view the breakdown of dates here.

This framework was introduced to make voting and review periods more predictable (collate proposals to one specific timeframe). If we were to impose a review period policy per proposal, SafeDAO would have to vote at irregular times when each proposal is ready for review.

That being said, happy to take additional feedback (on a different thread or DM) on the structure (e.g. notification or review periods). Hope this helps!


Thank you for the additional detail - I continue to believe Karpatkey is uniquely and positioned to manage this joint treasury. Additionally, the ability for the SafeDao to vote to change or cancel the agreement seems reasonable as we embark on setting up the arrangement.

1 Like

This proposal should still not be up for vote.: because (i) the concerns shared in this thread have not been properly addressed in an amendment of proposal and (ii) the SAFE DAO requiremenfor treasury management have not been clarified yet.

The idea of establishing a core unit with broader competences beyond the joint Gnosis DAO <> Safe DAO treasury could severe consequences for the DAO. It raises the risk of increased opacity and loss of management control, which are critical issues that need careful consideration and deliberation. Many of the previous comments, esp. from 1kx, in this thread should be taken into account before allowing for a governance event or supporting this proposal.

The most critical points in my opinion are:

  1. Lack of treasury management KPIs or processes within the DAO
  2. Mixing different roles for the DAO in one single core unit
  3. Mandating a sole, global asset manager instead of specialized asset managers
  4. Inadequate pricing model
  5. Operational complexity, risking a delay in the implementation as happened with ENS DAO

1. Lack of treasury management KPIs or requirements
Before mandating a treasury manager, the DAO needs to establish its requirements for the general treasury management KPIs (e.g. runway needs from the treasury), the legal/risk frameworks, asset manager selection criteria as well as controlling procedures. Just like with the grants council, SAFE DAO could appoint experts within the DAO or outside of the DAO as with the grants program. Considering the grants program is managing a budget of $1M, a similar procedure would be appropriate given the DAO treasury is multiple sizes larger than this.

2. Mixing different roles for the DAO in one single core unit
From my experience in other DAOs, mixing the role of a treasury manager, protocol relations, and native token ops creates a conflict of interest and represents a thread of power struggles in a DAO. Having various collectives contributing to protocol relations can negatively impact the coordination and alignment within the protocol relations group, representing a thread of contributors running in different directions, losing ownership, accountability, and time to market.

3. Mandating a sole, generalist asset manager specialized ones
The treasury holdings require dedicated strategies and expertise for the various asset types from native token, crypto assets and stablecoins. In sophisticated portfolio management strategies there would e.g. a dedicated strategic decision for the split between SAA and TAA (long-term strategic asset allocation vs short-term tactical asset allocation).
Ideally, a top-level portfolio strategy should be decided on by the DAO and executed by specialists for the various topics: yield farming, staking, money markets, equities, venture investment, credit or lending including the portfolio-specific risk management with hedging, delta neutral strategies, and asset liability management.
Native token management including liquidity management, sensitivity analysis for token sales, and token rewards programs is of strategic relevance and should be mandated and monitored individually.

This would still allow for karpatkey to win all mandates when competing against other service providers.

4. Inadequate pricing model
Operational activities should not be compensated with a percentage of the SAFE AuM. And portfolio management activities should not be based before a strategy has been deployed and performed well. I strongly support the concerns raised by 1kx above.

5. Operational complexity, risking a delay in the implementation as happened with ENS DAO
For the execution of the legal framework and technical deployment, the DAO should define an appropriate process depending on its needs as a pre-requisite, before deciding on a mandate and “buying” a setup that might not be robust or matching the conviction of the DAO for future regulatory and technical resilience.


Dear All,

We represent RockawayX, an investor within SafeDAO, and so far we focused on evaluating proposals and participating in voting stages only. However, the current proposal to allocate treasury management to Karpatkey necessitates us to voice our opposition.

We agree with 1kx’s analysis of the key issues and fully align with their approach. While we appreciate Karpatkey’s proactive stance, we believe that passing this proposal in its current state is premature and could potentially place the DAO in an unfavorable position.

Our disagreement revolves around four primary topics:
The Process: We believe that for an institution of this size, there should be a comprehensive tender process in place when selecting a treasury manager. This process would allow the DAO to assess various solutions while understanding the market-determined costs associated with such management.

The Cost: As highlighted by 1kx and others, the fee structure outlined is non-standard and disproportionate to traditional asset management fees. This could pose a significant financial burden over time. Furthermore, without a proper selection process, the DAO cannot ascertain the market rate for these services.

The Team: From our understanding, the fact that Karpatkey has yet to assemble a team responsible for treasury management underscores the lack of adequate preparation for a vote of this magnitude.

The Termination: There are concerns regarding the ease of terminating the agreement. Changes to the structure of delegated stake and voting power, which cannot be vetoed by token holders, concentrate a considerable amount of power among community members in favor of this proposal despite voiced concerns.

In its current form, Karpatkey’s proposal appears extractive to the DAO, lacking provisions for termination and posing several structural shortcomings. The proposal is established for an indefinite period with minimal risk on Karpatkey’s part (e.g., not assembling a dedicated team for evaluation purposes) for potentially substantial rewards.


As of April 15, 2024, this proposal has been ratified.

We are immensely grateful for the community’s strong support and trust in this proposal. The next steps will involve detailed planning and collaboration with the DAO to ensure a seamless process.

We are dedicated to maintaining transparency and efficiency as we move forward. Throughout the upcoming phases we will be available on community calls to address any further questions and provide additional clarifications as needed.

To follow-up on the concerns initially stated by @AccelXR-1kx and supported by @lianefiligrane and @ROXPM we have a previous reply where we address several topics including:

  • treasury management options & counterparty assessment;
  • terms of engagement and responsibilities;
  • fees and payments; and
  • alternative processes.

There were a couple of comments that we feel are incorrect, and we wish to address them below:

We serve as treasury managers for significant DAOs and protocols in the Ethereum and DeFi ecosystems with over 40 full-time contributors. The assertion that there is no team or there was a lack of preparation for a vote of this magnitude is not accurate.

To provide context on our operations, we establish a core unit for each DAO we work with. Given that the proposal was only ratified a few days ago, we are still finalizing the composition of the team exclusively dedicated to SafeDAO. However, it should be clear that we have ample resources from our existing team to meet SafeDAO’s needs. If further needs arise, we are prepared to recruit more full-time contributors to focus solely on SafeDAO.

No unexpected delays occurred with our implementation in the ENS DAO due to our operational complexity. On the contrary, our treasury team’s feedback from stewards has been overwhelmingly positive over the past year. We continue to work with the stewards and community to achieve a stable and self-sustaining endowment.