SCRF Interviews | Resilient Treasury Management - Daniel Ospina and Darcy Allen (Ep. 6)


The third episode of our treasury management mini-series co-hosted with the Royal Melbourne Institute of Technology (RMIT) features a conversation between governance expert Daniel Ospina and RMIT Blockchain Innovation Hub senior research fellow Darcy Allen. Blockchain Innovation Hub faculty member Kelsie Nabben and SCRF’s Operations Lead Eugene Leventhal moderated the event.



Audio (Apple, Spreaker)


At issue:

  • How can we make a treasury more robust when web3 is such a chaotic, unforgiving environment?
  • How can tooling like “guardians,” courts and compliance committees help navigate this complexity?
  • What kind of attacks and forms of rent-seeking should DAOs be safeguarded against?

Takeaways from the discussion:

  • DAOs operate in a brutal, unpredictable environment, shipping early and often, and learning from mistakes quickly tends to be an effective evolutionary strategy.
  • Communities are constantly changing and governance mechanisms have to keep up.
  • There’s a tradeoff between guarding against insider and outsider costs. Don’t forget about social threats and the community.
  • Polycentricity–in effect, empowering subordinate groups within an organization–can harden a treasury against attack.
  • Use a sandbox approach instead of training wheels, create a minimal version and deploy it as soon as possible to learn from your mistakes and adapt to them.

When Daniel Ospina began working with DAO-platform provider Aragon (he became their Head of Governance in 2021) he took an existing design that had evolved from a series of decentralizations and recentralizations. “Historically Aragon was more of a network of organizations, rather than a direct-to-contributor DAO,” Ospina says. “I started designing a DAO where contributors could directly add value and then evolving to a teams-of-teams setup.”

One of the challenges he confronted was that there were a few stakeholders, some of whom were no longer actively involved, who owned huge amounts of tokens and could essentially sway the network. The new design had to empower contributors who had smaller stakes while remaining resilient against huge funds who might try to take over the platform by buying up governance tokens.

“My colleagues and I often think of governance as a constant minimization of two types of cost,” says Darcy Allen. “Insider costs, the cost that treasury might be misappropriated by the owners of its multisig, and then you also have outsider costs, the risk that an outsider might buy up all of the tokens and use the voting system to their advantage.”

The beauty of web3 is that it allows organizations to use new types of tools and governance structures to minimize those costs, but it’s still a tradeoff. Every step a treasury takes moves them closer to one cost or another. Vulnerabilities also change as an organization grows. “An early-stage organization might be threatened by outsiders and benefit from having a single owner protecting its treasury,” Allen says. “As the number of stakeholders grows, however, it would make sense to decentralize to minimize insider risk.”

One approach to protecting a treasury and the community it serves is to use a polycentric design, which means creating a system with many decision-making centers. “Many DAOs use a single decision-making system for their treasury and that can open a lot of attack vectors,” Allen says. “The benefit of having many centers of decision-making is that they can specialize in their areas. Of course this can look like a centralized system. But there’s value in having many systems, you get lots of value and learning about what governance can be.”

There’s always a trade-off between decentralization and centralization. “There’s the ethos of decentralization in the community to consider,” Ospina says. “There’s a lot of conversation about subDAOs in the community right now and some people do push back against the idea, stressing the importance of simplicity.”

He sees tremendous value in having subdivisions. It can help streamline the process of dealing with awarding small grants, for example (the entire community needn’t be involved in a $200 disbursement) so long as there’s plenty of opportunity for community feedback. Smaller groups can offer a chance for experimentation and faster iteration.

As with many startup environments, shipping fast and often – and learning from one’s mistakes – is the best way to deal with an environment as unpredictable and unforgiving as web3.0.


Are there any circumstances when shipping fast and often could be a mistake? It seems like you’d want to be careful choosing a legal framework before you set up a treasury, for example. Are there other considerations that might need planning ahead of time?

1 Like

It was certainly fascinating to watch four professionals discussing real-world treasuries that they all freely admit no one fully understands, no one knows the best way to design, and that no one can fully control—partly because no one wants a central point of control that would make them more vulnerable.

And yet these treasuries will continue to be put in place and evolve, and they will continue to be fueled by real people’s money in the real world. Until now, these “experiments” have allowed that real money to be lost for various reasons, often by unethical people running off with funds contributed by ethical believers.

Obviously, these continuing DAO experiments need to be protected, and the concept of “sandboxing as opposed to training wheels” motivated me to watch this podcast in the first place.

I first understood the idea of “sandboxing” from the days when operating systems started running applications “in a sandbox” to prevent a crashing app from crashing the entire computer.

Is that analogous to how the term sandboxing is being used here? Is there a protective layer or wrapper that can be installed around a DAO that would prevent real-world misfortunes—from accidental code bugs to deliberate human badness—from taking down a DAO treasury and everybody’s investment in it?


When I first learned about DAO’s, one of my first questions was, how does a decentralized group manage a treasury efficiently and effectively while aligning with the goals of the stakeholders? This group does an excellent job of discussing the issues pertaining to this enormous subject. A few of the most interesting topics to me were the idea of communities having the mindset of experimentation and polycentricity.

@rlombreglia Mentions the idea of sandboxing and I think this is a great way to summarize the ending of the podcast’s focus on experimentation. In my opinion, in the early stages of DAO development, it is important to emphasize experimentation. This allows people to take risks in the name of innovation. @rlombreglia I’m not sure how to best protect users and allow experimentation. However, we run into an issue of human nature, is a person acting in self interest or truly for the good of the community? On top of this, in experimentation, we need to find the points of failure, it is easy here to place the blame on a person or group. I believe it is important to remember that experiments will fail and we ought to have almost a sense of grace and believe that these things were done with experimentation/innovation in mind.

Regarding the notion of poly-centric decision making the speakers point out that having committees vote on different issues. They point out the issues of voter apathy and community size; it may not be a bad idea early in a DAO to have the founders utilize a multi-sig wallet to ensure the project stays on course. This discussion of poly-centricity gave me an idea, what if DAO’s acted primarily as a board of directors that controlled a company which could be more or less centralized as defined by the DAO board? DAO board members could do the work of helping define the trajectory of a company and they could also hire for more traditional roles, C suite, managers, ect that align with the companies values. The DAO board could also help create specific policies, procedures, and so on that create an environment where members of the company are treated fairly and power doesn’t become too centralized. I believe this would also allow the company simply to be a machine of execution, it would be able to implement operations and track execution. I understand that this borders on centralization but in the spirit of experimentation, I am curious to hear what people think about this, please reply and point out any thoughts you have.


Hi @jmcgirk.
100% agree on legal settings as a barrier of uncertainty (and potentially liability) for DAO operators.
Kain Warwick from Synthetix emphasizes the importance of token distribution (who governs) and capital allocation (how spend) as essential initial settings in a “DAO First” approach (my notes on this in context here)

1 Like

I don’t disagree with your emphasis on experimentation, and your idea of coming at these experiments with “a sense of grace” strikes me as beautiful — in the abstract.

But even a quick glance at the realities of fraudulent behavior reinforces (for me, anyway) the necessity of having these experiments contained in some kind of protective wrapper (i.e., sandbox) that leaves people a way out that’s less drastic than a hard fork.


Are best practices available for this kind of decision? I’m also curious how important it is to build a legal framework into these plans? Would it make more sense to have an LLC and a designated agent in control of the treasury in the beginning? Seems like most platforms have some kind of holding company for their tokens and for paying operational expenses in fiat, is there a standard way of arranging that? I’ve heard investment clubs suggested before, but would that limit US involvement to accredited investors? How do VC-backed DAOs arrange their treasuries?