At first glance, Llama NFTs may look like simple, static PFPs, but they represent much more than that. They serve two primary purposes:

1. Representing membership in Llama

2. Representing contributions to Llama over time

Membership in Llama

First, we wanted to create a fun, visual, on-chain representation of membership to Llama. We designed the Llama avatars to be playful and futuristic, echoing our optimism about the future of the work we’re doing at Llama. We wanted the artwork to be easily identifiable so that others could come to recognize our contributors as members of Llama.

Since the Llama community began, we’ve worked with some of the top protocol DAOs, including Aave, Uniswap, and Nouns. We’re proud of the work we’ve done with these protocols. The Llama NFT collection is an on-chain memorial of this work and belongs to those who made it all possible: our community.

Each Llama contributor will receive a llama NFT with a unique combination of fur color and eye style, which are selected by the contributor and remain a permanent part of their NFT. At launch, we have 48 possible combinations of fur color and eye style. As the community grows, we’ll add new traits and colors to increase the number of unique llamas.

As the NFT represents membership in Llama, it was important that the NFT remain tied to the contributor’s wallet. For this reason, we made the NFT a non-transferable (soulbound) token. However, we realized we also needed some flexibility to allow transfers in case a user needs to change their wallet in the future. The smart contracts governing the NFT allow the Llama multisig to approve transfers to new contributor addresses.

Contributions over time

One of the primary goals of the NFT project was to capture a llama’s contributions over time and reflect those contributions on-chain. For this reason, the NFT was made upgradeable, with different traits unlocked by different levels of contribution and experience at Llama.

We have a total of 5 upgradeable traits, all of which will represent different types or levels of contributions within Llama.

• Backgrounds

• Ears

• Glasses

• Mouth accessories

• Necklaces

The background trait, for example, represents a llama’s level of context and experience in the community. Our community design is composed of 4 levels and draws inspiration from llamas in the Andes: Travelers, Explorers, Mountaineers, and Ranchers. A new llama enters the community as a Traveler and, after spending more time contributing to Llama, they progress to Explorer, then Mountaineer, and finally Rancher. A llama’s journey is a lifelong one; we’re always at the beginning and we’re always climbing.

Travelers begin their Llama journey with 1 background option. As llamas earn new roles through their contributions, they’ll also unlock special backgrounds, like Shooting Stars or Sparkly.

At launch, the background trait is the sole upgradeable trait. Incentive systems are hard to get right, and we wanted to launch with a simple but meaningful trait that’s well established within our community.

A llama’s level is intended to be a holistic representation of a llama’s contributions over time. Some of these contributions are hard to incentivize on a granular level, such as answering questions in the Discord server or jumping on calls to help other llamas when they get stuck. These types of actions are incredibly valuable for the health of the community, but they’re hard to incentivize and record without running into Goodhart’s Law (when a measure becomes a target, it ceases to be a good measure).

Over time, we’ll use the remaining traits (ears, glasses, mouth, necklaces) to represent other accomplishments or contributions to Llama. These might be things such as:

• Deploying smart contracts

• Building data dashboards

• Authoring governance posts

• Number of seasons contributing to Llama

• $ earned contributing to Llama

• Traits unlocked by posting IRL photos with llamas

The Llama Platform

Llama contributors are able to mint their NFTs through the Llama platform. This platform is home to all of Llama’s onboarding guides, documentation, notes, and internal tools. As the needs of our community grow, we’ll be able to evolve with those needs and reflect them in the Llama platform.

These Llama avatars are an important part of the platform; over time, PFPs will not only be upgradeable, but will also be interactive. Other community members will easily be able to tell what different traits represent by hovering over someone else’s PFP.

In the future, these avatars may also allow us to permission-gate different internal tools. As Llama becomes a larger contributor DAO, many more llamas may join the community. If certain tools or features require higher context, experience, or trust, they could be gated based on the avatar attributes a Llama has earned.

Powered by Metagame

While Llama already had an internal platform on which to host the minting and updating functionality, Metagame provides the tooling, infrastructure, and smart contracts to map contributions and participation to traits. Working with the Metagame team allowed us to tailor the solution to the needs of our community - we needed a flexible starting point that allowed us to define other upgradeable attributes in the future, as well as the infrastructure to manage the minting, upgrading, and transferring process.

Metagame is looking for more communities to beta test their new Earnable Avatar Studio. Please reach out to @brennerspear (brenner@themetagame.xyz) if you and your community are interested.

View the full collection on OpenSea.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

Many thanks to grossehalbuer for the artwork for our NFTs and to @alexkayaian for the UI design!

The River, the Pond, and the Lake of Legends

Money stopped flowing through the gates and the ecosystem is changing. Those that perfected the catching net for the tumultuous rivers of 2020 might not be the ones that will master the placid lake, or better smelly pond, that is late 2022 DeFi.

DeFi liquidity, that almost reached $200b at the beginning of 2022, now lingers around $60b—source DefiLlama, and doesn’t show signs of improvement. While we wait for technological—the merge and the deployment of L2 scaling mainly, governance, and regulatory progress, the DeFi crowd is splitting in two: in one corner those focused on connecting new pockets of the world to the tech, and in the other the ones dedicated to morph and capture as much remaining liquidity as possible—while positioning themselves for the next bull run. Whatever will bring the bull back. Islands in the stream.

Recently, Dirt Roads has focused way more on the first bunch, and especially on TradiFi<>DeFi bridge builders and governance framework innovators. For good reasons: I care more about changing the world than stealing another piece of the pie. The degree of integration of blockchain-based finance with the outside world is all yet to be proven. Today, however, we shift to the second one: the liquidity catchers.

We can start by eyeballing the top 5 DeFi names by TVL—again DefiLlama. The list is a good representation of the ossified DeFi sector of our days, and gives a sense of the playing field. It includes a sovereign lender—MakerDAO, a liquid staking provider—Lido, a money market superapp—Aave, and two DEXs, although very different from each other.

In the pre-crypto reality, the financial sector has always been a derivative of the underlying economy, with a size that is a fraction of the net present value of the economy itself. Finance, ultimately, is an allocator and distributor of resources, and a financial sector folded onto itself is rarely a sustainable one. In aggregate, of course. Coherent with this framing, the current DeFi sector bases its value onto the intermediation of two economies: the Ethereum ecosystem—which could be compared to crypto’s sovereign industrial sector with measurable throughput, and the US economy. DeFi has pitched itself as a more efficient distribution layer for the allocation and transfer of resources, and it is entirely understandable that such layer has been massively dollarised over the years. People do indeed use DeFi to transfer and store their dollar wealth, and that’s a good thing. Web2 had a similar trend, with commercial and political capital flowing through the pipes of social media developed in the Silicon Valley. In my opinion, a self-referencing DeFi is an unhelpful illusion.

It is unclear what will be the next zero-to-one innovation that will help the industry making the jump towards the next phase of its evolution. In the meanwhile, while they wait for the next big inflow of liquidity, most OGs have started flexing to steal market share from others. A phase of great convergence has begun, starting around the most successful product to date: stablecoins.

The Currency Rush

But why are stablecoins, or more in general currencies, such an attractive product to develop? Let’s forget philosophy of money for a second and focus on the strictly financial characteristics of currencies. Currencies minted via a Collateralised Debt Position (or CDP) to be precise. Pragmatically, a borrower, who owns collateral, pledges such collateral to a counterparty, in exchange for a loan in stable currency form. The loan could be used for several purposes but that doesn’t interest us now. For the benefit of having the loan borrowers pay a Premium—in interests or fees, and give the ability to the lender to seize the collateral if it falls below a certain Liquidation Threshold. In financial engineering terms borrowing via a CDP is equivalent to packaging a structured financial product that involves some shorting and some option trading.

Phase 1: potential borrower owns the asset → At the onset, i.e. before taking on any loan, a potential borrower’s profits are linearly connected with the value of the collateral asset (e.g. $ETH) he owns; if value grows by $1 he will increase profits by $1.

Phase 2: borrower pledges his asset in exchange for a loan → By entering a loan contract, the borrower is de facto selling the asset to borrow another—we can ignore the value of the other asset here. By selling (shorting) the collateral asset he is netting out his straight long; trivially, his net payout is now indifferent to the price movement of the collateral asset.

Phase 3: borrower retains rights to access the asset → By pledging (and not selling) the borrower is retaining the right to swap back, at any given point in the future, the borrowed asset for the one pledged. He is, in other words, purchasing an American call option. If the asset value grows above the Pledged Value (i.e. strike price) his profit will grow linearly. The Premium paid by the borrower is equivalent to the option premium.

Phase 4: borrower gives the lender rights to seize the asset at discount → The borrower, actually, is giving the right to the lender to seize the asset at a discount if the value falls below the Liquidation Threshold. This means that the inflection point of the borrower’s profit curve is not the Pledged Value, but rather the Liquidation Threshold, that acts as a backstop. The size of the discount helps the lending protocol to manage its balance sheet—by incentivising auction keepers to buy the collateral and sell it for a profit, but this is something that doesn’t concern us.

The charts above are quite simplistic, as they ignore the fluctuations in value of the borrowed asset—e.g. $DAI, and the fact that such asset can be swapped for more $ETH therefore increasing the sensitivity/ derivative of the payoff to $ETH prices. Those simplifications do not bother us, given that we are trying to prove how attractive is currency minting rather than borrowing. The charts stimulate the following reactions:

• Lender has an infinite amount of assets to lend—or calls to write: a stablecoin minter can mint as much stables as desired, assuming there is enough collateral to mint against. It can also do this for free, given that the all-in marginal cost of its liquidity is pretty much zero

• Good option writing is profitable: investment banks make a lot of money by writing and market-making options, and this is because they clip option premia from buyers—in the context of stablecoin minters liquidation penalties act as an extra sweetener

• Parameterisation is key: in order to run a profitable option writing business careful parameterisation (of Liquidation Threshold, Premium/ Fee, and Discount/ Penalty) becomes key

• Volatility is what matters: the Black-Scholes framework teaches us that the price of an option derivative is positively correlated with the volatility of the underlying asset, and volatility matters when picking parameterisation

• Volatility of the borrowed asset makes things very complicated: we have assumed until now that the borrowed asset (e.g. $DAI) is stable in value, introducing volatility considerations also on that front makes modelling strategies (and parameters) way more complicated—a highly volatile $DAI might attract a very different type of users compared to today’s

Aave and the $GHO Project

The list of the top 5 OGs (remaining) includes only one minter of currency: Maker.

If I were Maker, or a large $MKR whale, looking at the top 5 DeFi chart would give me vibes of terror. For one reason or another, all those protocols control a huge amount of financial value and there is nothing forbidding them from trying to mask some option selling in the form of a stablecoin and cross sell it to their users. It didn’t come as a surprise when protocol #3, Aave, opened the dances.

At the time of writing, Aave hosts c. $10b worth of assets provided across 7 networks— including Ethereum, Avalanche, Optimism, Fantom, Polygon, and Arbitrum. More than $6b out of the $10b have been provided on the Ethereum market, with c. $1.9b being borrowed and the rest sitting at the protocol level. $USDC is Aave’s largest market on Mainnet, with c. $1.5b supplied and c. $430m borrowed. Volumes are high and spreads are tight, with supplied capital clipping a variable 0.33% (nota bene way less than US treasuries with the 3-month at c. 2.9% today) and borrowed capital paying a variable 1.24%. There are structural reasons behind the bid-ask spread—we have debated it when introducing Morpho and its plan to port crypto money markets towards an order-book framework, as well as mundane ones like taxes and convenience.

Please meet $GHO → In early July, the Aave team introduced to the community $GHO, its native, dollar-pegged, stablecoin idea. As with others decentralised stables, the minting engine would accept crypto collateral (based on governance-approved collateralisation ratios) and provide to those wallets a certain amount of $GHO. The same amount of $GHO (subject to lack of protocol losses) would be burned following repayment or liquidation. All fees would accrue to the DAO treasury.

Permissioned permissionlessness → $GHO could be permissionlessly minted by Facilitators, which are protocols, entities, etc. authorised (read permissioned) to interact with the currency engine. Each Facilitator gets assigned a Bucket, which represents the maximum amount of currency that a Facilitator can mint. Each Facilitator can be theoretically completely different, with individual risk profiles, as well as liquidity and technical characteristics. Here on Dirt Roads we have advocated several time for the emergence of a lending sun of last resort that would provide super senior protocol-to-protocol liquidity across DeFi, we had just assumed that such a protocol would have been Maker, not Aave. As usual, our own emotions can blind us.

Interestingly, and somehow naturally, $GHO’s first proposed Facilitator is the Aave Protocol itself. Governance would assign the Aave Protocol its dedicated Bucket to bootstrap the $GHO currency. Based on what has been shared on forums, the Aave DAO will rule upon $GHO’s parameters—including interest rates and Bucket assigned to itself, and the risk parameterisation and execution of its own market. I personally dislike governance commingling, but I guess we should wait for more details around the implementation and watch.

The chart below, published on the Aave forum, included also real-world assets (i.e. RWA) as potential Facilitator. We have debated at length on the complexity of integrating opaque (and heavily regulated) dumb contracts with the blockchain, and we are looking forward to read more about the implementation strategy when the first Facilitator of this kind will come online. The same can be said of delta neutral positions and market makers.

GHOsting Aave

How would the Aave Protocol Facilitator work within the $GHO context? The integration goes hand-in-hand with the launch of Aave V3, and benefits from several of its features.

Isolation Mode → On V3, isolated markets can allow borrowers to post a specific asset as collateral, giving them the ability to borrow stablecoins from the protocol based on defined collateralisation parameters. The standard V3 approach aims at isolating (often novel) collateral volatility risk as a support for stablecoin suppliers, but it doesn’t change the core of the Aave protocol—the stablecoins lent out have been initially provided by other peers. With the integration of $GHO into V3, Aave becomes instead a conduit of freshly minted $GHO expanding Aave’s balance sheet in a way that, yet again, reminds a lot of the way Maker mints and distributes $DAI.

High Efficiency Mode (eMode) → eMode is another new feature brought forward by V3. Individual Aave markets can be categorised, based on certain selected features, and assigned a collateralisation factor and liquidity penalty. Categories, typically, wrap tokens of similar characteristics and highly correlated with each other—i.e. all stables, $ETH and derivatives, $BTC and derivatives, etc. Asset suppliers or borrowers, in theory, can benefit from commingling those currencies to diversify some of the idiosyncratic risks out, and ultimately get better rates and parameters. If you have perceived some Curve-y vibes you are not alone. By incorporating the ability to borrow $GHO against each category, Aave intends to internalise demand for both stable and volatile assets in bear and bull markets. Borrowers, the theory says, would borrow in $GHO against (e.g.) $ETH to extend their long position, or deposit stables to borrow more $GHO (with a close to one-for-one ratio) to bring their leverage down on the $ETH eMode when prices start dropping. V3’s Portal would allow liquidity to flow across markets and chains.

Flywheels and bootstraps → Aave protocol supporters would have preferential treatment in borrowing in $GHO by providing evidence of staked $AAVE tokens. The amount of $GHO funding available at discount (for each $stkAAVE provided) and the discount itself will have to be decided by governance. There are currently c. $280m staked in the protocol’s Safety Module, subject to a risk of slashing of 30% in order to protect the protocol against losses, against a c. 9% APR. Given the flimsy connection between protocol revenues and valuations, there is nothing bad in trying to use vibes to accrue financial value for governance tokens.

Risk of cannibalisation → There definitely is some risk of cannibalisation involved here, with demand of $GHO potentially causing utilisation and borrowing rates on deposited stables to go down further. However, it is fair to assume that the blended interest margin on $GHO is way larger than that retained on the money market side. My back-of-the-envelope calculations show that the margin retained by Aave on deposited $USDC is 1-2 bps only—i.e. 0.01-0.02%. We don’t know much yet about the launch of $GHO, and I guess the team is in no particular rush given the lack of borrowing demand in the crypto space nowadays, but the spread would probably be 50-100x larger than this.

The Next Dance

Aave might not be the most obvious candidate among the OGs for spinning its own stablecoin. Yes, it controls a lot of deposited value. Yes, it has the intention to become some sort of retail-oriented DeFi superapp and you need a stablecoin in your arsenal for that. Yes, the cannibalisation costs are moderate. Yet my top pick would have been another one: Lido. The runner-up? Curve.

Curve → And so it goes. At the end of July, Curve soft-announced they were joining the stablecoin war. Curve controls vast amounts of dollars in homogenous liquidity pools— $stETH+$ETH and $DAI+$USDC+$USDT are the largest ones. Currently, depositing into those pools spits out an LP token that has been used as the main weapon of the Curve bribing ecosystem. With the sustainability of the Curve Wars under scrutiny, the extra step that would see the use of such LP token as collateral for a native stable gets interesting. Curve would benefit from having already a deeply liquid market for potential liquidations. The downside? Most assets held in Curve are stablecoin assets, and providing funding backed by stablecoin collateral isn’t very profitable for the protocol as we have seen for Aave before and for Maker’s PSM—that makes zero revenues. True, the $stETH+$ETH pool is #2 with $1.2b TVL and that could be leveraged. There is another protocol, however, that controls even more Lido $ETH than Curve, and that protocol is…Lido.

Lido → There are c. $6.7b worth of $ETH staked at Lido in exchange for $stETH. The protocol hasn’t shown interest in developing their own native stablecoin yet, probably busy with other issues and the uncertainty around the looming Ethereum merge. That doesn’t mean, however, that it won’t happen. A successful transition to PoS, together with perfect fungibility of $ETH and $stETH, could provide the perfect terroir for Lido to develop a single-collateral stablecoin backed by $ETH staked natively on the platform. The UX experience could be smooth, with staking happening behind the scenes for borrowers. Monetisation could happen in the form of interest rates, or as a share of $ETH staking yield, making it almost seamless for the borrower. Liquidations could be run through various avenues, including Curve $stETH pool (that might continue to be incentivised by Lido via $LDO distribution) and other $ETH liquidity venues. My guess is that Lido might continue to incentivise liquidations through the existence of a large and deep market for $stETH—as it would help them keep value within the platform. If someone from Lido wants to bump ideas on the matter my Twitter DMs are open.

The great convergence has started, and it will most likely continue as long as the bear market stays with us. The competitive landscape of tomorrow might look way messier and more aggressive than what it is today, with a contracting effect on protocol margins and revenues. It is curious that no-one has embarked on a mission to attack the widest margin of the whole industry, that thick 3% US treasury spread captured by Circle when swapping dollars for crypto dollars. Somebody should wake up. I am making it a note to myself.

Subscribe to Dirt Roads.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

TLDR: Nouns has allocated 14,835 ETH since inception across 292 proposals via on-chain governance, Prop House, and small grants. 61% of these proposals have been completed, 37% of them are in progress, and 2% were incomplete*. While the ROI of specific allocations is difficult to measure, Nouns’ bid price, number of unique bidders, and treasury growth has held up over the past year.

Number of proposals and ETH allocated by category:

Completion rates of proposals by category:

*Incomplete includes abandoned/canceled/uncertain projects

Introduction

The Nouns community continues to fascinate us at Llama. We have published a couple of pieces focused on their unique model including Extending Nouns Through Invisible Traits by Austin Green and Designing a Nounish DAO: Analysis of Auction Frequency by gumfather. We will now dive into how effectively funds are allocated.

To date, ETH 26,695 has been amassed in their treasury (as of August 4, 2022) that are being allocated to community projects and special initiatives through three separate allocation mechanisms: small grants, Prop House, and on-chain governance proposals. We will review the efficacy of this segmented grant structure by answering the following questions across their allocation programs:

• How do Nouns’ funding mechanisms work?

• How effective has Nouns' allocation structure been in facilitating project completion?

• How can Nouns improve this structure as it continues to grow?

How is Nouns funding structured?

The primary source of funds for Nouns DAO are the auctions that take place every 24 hours. The Nouns treasury receives 100% of the proceeds from these auctions (denominated in ETH). Governance of the DAO is based on the Compound Governance model and is the main governing body of the greater Nouns ecosystem.  Each Noun is an irrevocable member of the Nouns DAO and entitled to one vote in all matters of DAO governance. We will dive more into the topic of Nouns governance in future reports so, for now, let’s stay focused on the topic at hand: how grants funds are distributed.

To allocate these funds, Nouns DAO has elected to manage allocation governance through three different programs: Small Grants, Prop House and Proposals. Each grant program is aimed at allocating funds for projects of a certain size (denominated in ETH) and tackle a specific set of DAO requirements.

As someone who has spent a fair amount of time researching grants and incentives programs across Web3, the different funding bands were an interesting approach and gave some segmentation in terms of types of projects and overall level of governance needed for review and approval.

How effective are each of these programs in facilitating project completion?

Small Grants

The Small Grants Committee was created through Nouns Proposal 13. The Committee controls a flexible pool of capital for smaller projects, provides retroactive funding for completed projects, and funds projects where the ask is too small for an official proposal. These grants are anywhere from 0.1-10 ETH. The capital used to fund these proposals is controlled by a funding committee consisting of four Nouns and the Nouns DAO itself (who is mostly there as a backup and symbolic owner). Disbursement of funds requires two of the members to sign a multisig transaction.

The program was originally provided 20 ETH as a pilot and immediately used some of those funds to retroactively reward a few of the community members to the tune of 2.75 ETH. Over time Nouns has passed more proposals (85, 43 and 31) to provide additional funds to the Small Grants Committee. These proposals added 50 ETH in January, 100 ETH in March, and 200 ETH in June (all in 2022).

The process to request funding for small grants is straightforward. All one needs to do is open a post on the Nouns Discourse forum with the prefix “Small Grants” and include some basic information including a project overview, how it will grow Nouns (or how it has already) and how much funding is needed (or retroactively owed).

Nouns Small Grants (like all Nouns grants) are relatively transparent with the projects and the amounts disbursed available both on this Notion page along with the the small grants funding page on Nouns Center.

The total number of grants that have been completed at time of writing is 105 proposals or ~95% with ~226 ETH disbursed from the treasury. There are currently 5 active projects with ~22 ETH in funding, all of which are on schedule. The Small Grants team does a good job of keeping projects on task and completing them based on the timelines agreed upon in each of the proposals.

The projects are tracked across a few different categories including Art, Charity, Community, Marketing, Operational, Physical, Staking and Tech. Some of the categories do overlap (meaning some might be called both Operational and Tech, for instance) so it is a bit difficult to give exact totals related to each category. Overall, Art and Community are the most active with a combined total of 74 (again, some of these overlap) while others like Physical, Marketing and Tech accounting for lower single and double digit totals.

Both Charity and Staking had no projects listed; that may be an area for Nouns to consider funding or focusing on.

Prop House

Prop House is an interesting model that looks to have creators submit their ideas to win auctions ranging from 2-10 ETH. In some ways, this functions a bit like Gitcoin Grants, where funding rounds are launched and anyone can create a proposal for community members to vote on and express their support. Where this differs a bit from Gitcoin is in that the amount for each round is predetermined versus an open bid where the amounts can change based on the number of contributions coming in.

The original auction house, known as Nouns Launchpad, was started in late March 2022 and has had seven separate rounds with anywhere from 2-5 ETH in funding allocated for each round. After the success of this model, separate community prop (short for proposal) “houses” have been created as a means to decentralize the applicant pool and boost the number of proposals that are receiving funding. This approach was formalized in proposal 62 where the concept of these independent houses was introduced. At a high level, these houses would all include different and distinct NFT communities that are focused on specific types of auctions that all serve the greater Nouns Ecosystem.

An interesting quote from the information available that sums it up quite nicely:

“If the ERC-721 standard is thought of as the social graph, Prop House may be looked at as the permissionless power source that communities gather around. Akin to reddit, I believe we will see NFT communities for every niche imaginable.”

At the time of writing, there were a total of 84 projects with 278 ETH total granted. The projects had the same categories that were present in Small Grants (Art, Charity, Community, Marketing, Operational, Physical, Staking and Tech) with a similar funding makeup; Art and Community carry the lion’s share of funded projects. There were a couple of entries for Charity totaling around 4 ETH with Staking still being at zero.

There are 84 total projects to date on Prop House, out of which 22 are completed (26%), 35 are on schedule (41%), 3 are behind schedule (4%), 3 abandoned and uncertain (4%) and 21 that currently do not list a status (25% - these are mostly Lil-Nouns projects with a few from NounPunks). Overall it seems that they have still done a good job of ensuring progression of these projects considering they are larger and more intensive than what is being proposed with the smaller grants.

On-Chain Proposals

The last area we’ll examine in terms of grants is their on-chain Proposals. This category consists of larger funding requests ranging from 10-1,000 ETH. For this category, only Noun Owners (“Nouners”) are able to request funding. If someone would like to make a request for funding without owning a Noun, they can seek sponsorship from a Nouner by posting their idea in the Discord or Discourse or by contacting groups like Nouncil or any of the numerous Nouns’ sub-DAOs to see if they may be willing to help.

In looking at the requirements for a proposal it does ramp up a bit in terms of details needed including a detailed breakdown of how the funds will be allocated and what are the success metrics in play for the project. Nouns does ask how these proposals are structured to provide both a high level overview of the task and then be very specific in follow-on sections to detail out.

Once the proposal is submitted it will undergo what is known as a “Voting Week” which is a minimum of seven days. There is a visual overview of this process is shown below:

The categories for these proposals mirror those previously mentioned with a couple of additions: Investment and a more broad category termed “Other.” A total of 123 projects at time of writing were brought to proposal, with 98 of those passed for a total of ~14,309 ETH. The projects that didn’t pass were either defeated (voted down), expired, vetoed or canceled.

As we review further, the mix of projects funded shows more diversity in terms of category spread versus previous programs. As many of the projects are spread across multiple categories it is hard to break down specifically the funding and count for each. Marketing and Community seemed to lead the way with mid double digit counts and 4,000 to upwards of 5,300 ETH respectively. This seems consistent with what the community believe is the goal of Nouns governance (as per the inaugural Nouns survey) - to ‘proliferate the meme’. Other categories such as Art saw lesser numbers in terms of overall funding versus previous categories while Charity and Staking saw increased amounts relative to previously reviewed programs.

50 of the 98 granted projects have been completed (51%) with another 38 on schedule (39%), 6 behind schedule (6%) and 4 either uncertain, canceled or abandoned (4%). Like the other programs it seems there is a relatively high success rate for the on-chain Proposals even with the added overhead of governance that is introduced.

Concluding Thoughts: Path to the Future

The allocation mechanisms at Nouns are set up to fund interesting experiments quickly, reward talented contributors, and proliferate the Nouns meme.

Proliferating the Nouns meme is a difficult goal to attribute ROI for specific projects. However, Nouns’ bid price, number of unique bidders, and treasury growth has held up over the past year. Allocations have been relatively successful in facilitating project completion. 61% of the 292 projects have been completed and 37% are in progress.

The Small Grants program has allowed 100+ grants to be completed in the past year. The program has created as little friction as possible for builders and creators to charge ahead with their ideas. When completion rates on smaller grants start to tail off (e.g., with proposals on Prop House), Nouns team appear engaged on key forums and aware of the progress on these, with support being offered to ensure progress.

With Nouns continuing to grow, it is inevitable that the size of funds requested, and complexity of on-chain proposals for which the funds are being requested, will continue to grow. Nouns community would benefit from requiring proposals to define success metrics in advance, so the efficacy of the funds distributed can be monitored with greater rigor, and the treasury can move forward with active insights. Where proposals are complex and involve a lengthy development phase, Nouns should consider introducing milestone based payment structures so that the community can be rewarded for ongoing efforts.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

Introduction

Nouns is an NFT project on the Ethereum blockchain. Launched in August 2021, it has grown into one of the most exciting experiments in community formation and governance in DeFi today. While there are many unique aspects of the Nouns project, perhaps the most iconic is its distribution mechanism: exactly one Noun is auctioned per day in perpetuity, with the proceeds flowing directly to the community treasury to be managed by Nouns holders. This mechanism has generated significant capital inflows with remarkable durability - Nouns routinely sell for 100+ ETH, helping to bootstrap a treasury of more than 25,000 ETH by July 2022 [1].

Given the initial success of the Nouns project, it is no surprise that it has already inspired rich experimentation as other communities iterate on its essential characteristics. One such example is Lil Nouns, a Nouns fork that reduces auction duration from 24 hours to just 15 minutes. As more projects consider a Nounish distribution mechanism, the question of optimal auction frequency is likely to become increasingly debated.

In this post, I will evaluate the impact of auction frequency on a variety of key metrics: revenue, community participation and community growth. Critically, this evaluation attempts to assess the impact of auction frequency on these metrics holding all else equal (e.g., art, provenance, quality of initial team).

For each key metric, I provide an analysis from first principles on the potential impact of auction frequency. To bolster my arguments (or in some cases, identify limitations), I reference data from Nouns and Lil Nouns. Ultimately, I provide a framework for how various types of DAOs and / or projects who are considering a Nounish auction structure might determine their optimal auction frequency.

Revenue

Revenue is a critical consideration for any DAO or project. In a Nounish structure, auctions are the direct source of recurring revenue. In this section, I evaluate the impact of auction frequency on treasury growth and provide a perspective on how durable that growth might be over time.

How are treasury inflows impacted by auction frequency?

Daily treasury inflows can be calculated with a simple equation:

Daily treasury inflows = Average winning bid price * the number of auctions per day

As we adjust auction frequency, we would expect these two factors to move in opposite directions: a lower-frequency mechanism seems likely to result in a higher-average bid price compared to a higher-frequency mechanism. While this may seem obvious, it leaves us with a crucial question: what is the optimal frequency to maximize daily inflows?

The surprising answer: Assuming rational bidders, it is not clear that auction frequency should have any impact on overall treasury growth. Assuming that bidders ascribe the same total value to the project (based on total treasury size, etc.), they should be willing to pay the same daily amount to purchase one day’s worth of votes regardless of auction frequency.

For a concrete example, let’s assume that the current treasury for a Nounish project is 100 ETH after 100 days of auctions. For a project with one auction per day, this implies a ‘book value’ of 1 ETH / vote. We would expect the next auction to be settled with a winning bid of 1 ETH (and therefore a total daily inflow of 1 ETH to the treasury). For a project with ten auctions per day, this implies a ‘book value’ of 0.1 ETH / vote. We would expect the next auction to be settled with a winning bid of 0.1 ETH – since this project has ten auctions per day, this results in a total daily inflow of 1 ETH to the treasury. In both cases, we have a total daily inflow of 1 ETH.

However, there are several reasons why this may not be the case in practice:

• Rarity: This has been a key driver of NFT interest and value – less frequent auctions may increase perceived rarity (even absent explicit rarity built into the actual tokens) and therefore increase both interest and average bid price relative to higher frequency auctions

• Forks: Nouns DAO has already been forked several times (e.g., Lil Nouns). There may be some premium ascribed to Nouns over its forks, as each forked community increases the value of Nouns itself (either directly or indirectly), that would result in a higher average bid price and faster treasury growth for the original community. Though not a perfect analogy, we can see that derivative NFT collections generally command lower prices than the original collections (e.g., Mutant Ape Yacht Club has a significantly lower floor price, at ~20 ETH, than Bored Ape Yacht Club at ~100 ETH [2])

• Auction settlement mechanism: After auctions are completed in Nounish structures, a settlement transaction is required to mint a Noun and begin a new auction. This does not happen automatically – someone must trigger the settlement to launch the next auction. Therefore, auctions do not always start immediately after a prior auction has completed. While this may not have a large impact on lower-frequency auctions, it can dramatically decrease the number of auctions compared to the theoretical daily maximum for higher-frequency auctions. Lil Nouns is a good example: it has averaged ~41 auctions per day over its lifespan, whereas the theoretical maximum would be 96 auctions per day. This may lead to decreased inflows relative to expectations, as bidders cannot predict precise auction frequency with certainty and may therefore deviate from perfectly rational bidding behavior. If Lil Nouns bidders initially bid with the expectation of ~100 auctions per day, this may have established an average winning bid price below where it would have been if they had expected ~50 auctions per day. Price expectations are sticky and difficult to displace, potentially making this self-fulfilling and reducing treasury growth

• Extreme auction frequencies: In practice, there is likely to be a ‘sweet spot’ of acceptable auction frequencies for this type of model. Projects that choose a frequency that is extremely high (e.g., one auction every fifteen seconds) may find that there are limits to rational price discovery – even if the average bid price should, in theory, find a level where total treasury growth finds equilibrium with lower-frequency models, it may not occur in practice due to the level of consistent attention and participation demanded by such rapid auctions. This could result in more volatile winning bids, and therefore less efficient auctions. Gas fees are also likely to become a more significant factor at higher frequencies (as they make up a higher percentage of total bids and purchases), which may depress demand

• Post-launch change in auction frequency: While we might expect that the initial choice of auction frequency would not impact treasury growth, the impact of a shift after launch would likely be different. As Kiran suggests here, if an existing DAO like Nouns were to increase its rate of emissions it would likely result in an uptick in treasury growth. This is because the expected value of an incremental holder has already been established (by the lower frequency). To make this clearer - a Nounish project might start at one auction per day and establish an average winning bid of 100 ETH. At this point, each additional Noun auctioned would have an expected value of 100 ETH (based on the number of Nouns and treasury size). If the number of auctions per day increased to four, each incremental Noun would still have an expected value of 100 ETH. As long as there were enough bidders with a reserve price above 100 ETH, we would expect the price to remain steady even with the faster emissions rate. On the other hand, there is nothing to necessarily suggest that starting the project with a faster emissions rate would lead to this outcome. In fact, we might expect the project to establish an average winning bid of 25 ETH, finding equilibrium with the initial 100 ETH / day value

We can look to early data from Nouns and Lil Nouns to begin to hypothesize about the impact of auction frequency on revenue growth. Since inception in August 2021, Nouns DAO has grown its treasury significantly at an average rate of ~100 ETH per day. After 339 auctions (as of 7/17/22), this has resulted in the inflow of 33,917 ETH since inception.

The advent of Lil Nouns, in May 2022, provides an opportunity to compare the relative impact of more frequent auctions with the original Nouns DAO mechanism.

While Lil Nouns has conducted 2500+ auctions, it has a more limited number of total days for treasury growth analysis. We have compared its first 68 days in operation to both the first 68 days of Nouns (to hold constant the impact of starting a community) and the same 68 days of Nouns (to hold constant the impact of market conditions).

We can clearly see that Nouns has been able to generate more treasury growth historically (at ~100+ ETH per day compared to ~24 ETH per day for Lil Nouns). This contradicts my hypothesis, but could potentially be explained by a combination of the ‘rarity’ phenomenon and the nature of Lil Nouns as a fork (and explicit ‘lesser’ derivative) of Nouns, which may depress its overall value compared to Nouns. As both projects mature, it will be interesting to see if there is any convergence in total daily revenue.

How durable will treasury growth be over time?

Sustaining treasury growth over time is another key consideration for Nounish DAOs. An analysis of the number of unique bidders and winning bid variability over time can provide insight into the impact auction frequency might have on project sustainability.

While not strictly an input for treasury growth, the number of unique bidders over time on a Nounish project can serve as a proxy for general interest in the project and provide insight into its long-term treasury outlook. We would expect, all else equal, that auctions with a higher number of unique bidders tend to result in higher average winning bids and therefore that projects with a higher number of unique bidders over time will tend to be more successful and last longer.

The topic of unique bidders can be considered over two time frames: long-term (over weeks and months) and short-term (within a single day or week):

• Over the long term, projects that are able to sustain a high number of unique bidders are more likely to maintain their treasury growth, whereas projects with a dwindling number of unique bidders are likely in trouble

• In the short term, high volatility in the number of unique bidders may indicate a suboptimal auction frequency. Auctions with relatively fewer unique bidders may be more likely to be resolved with lower winning bids, which reduces the efficiency of the mechanism and slows treasury growth (see previous section for more details on this point)

While the impact of auction frequency on the number of unique bidders (and ultimately project longevity) is not yet entirely clear, I’ve laid out several hypotheses on the potential impact:

• Short-term: Lower-frequency auctions are likely to have somewhat lower volatility in the number of unique bidders, as participants have more time for price discovery and can monitor the auction relatively infrequently over a longer time period. Higher winning bid prices also make the relative cost of gas for bids cheaper, encouraging more bids in each auction. Conversely, higher-frequency auctions demand a more consistent level of interest and attention throughout a given time period to sustain a consistent number of unique bidders. While there is likely minimal impact within some reasonable band of auction frequencies, extreme frequencies may result in high volatility

• Long-term: The impact of auction frequency on long-term volatility of unique bidders is not entirely clear. Lower frequency auctions likely limit the universe of potential buyers given the elevated average winning bid, which may result in an earlier tapering of interest. On the other hand, higher frequency auctions require a sustained level of interest over many more auctions

As we compare Nouns and Lil Nouns, we can see that Nouns, with a longer auction duration, attracts more unique bidders per auction than Lil Nouns (~8 for Nouns compared to ~4 for Lil Nouns). It seems that, despite the lower barrier of entry for Lil Nouns given the price point, the longer duration auctions for Nouns is actually a more important determinant for unique bidders (at least so far). Interestingly, Nouns has largely sustained the number of unique bidders over time at ~6-8 per auction, suggesting no significant tapering in interest to date.

Despite the relative stability in the number of unique bidders, there has been relatively high variability in winning auction prices for Nouns and Lil Nouns. This brings up a natural question:

“Why the focus on unique bidders? After all, a steady supply of unique bidders over time did little to insulate Nouns from wild price swings over time (dropping from 150+ ETH in the first several months to as low as ~50 ETH, before converging back to the overall average winning bid price of 100 ETH). There must be more important factors at play, right?”

This is true –  there are undoubtedly other factors influencing the average bid price for Nouns over time. However, the number of unique bidders provides a useful proxy for interest in the project over time, while allowing for the abstraction of confounding variables (e.g., broader market conditions, performance of ETH). It also seems reasonable that Nounish projects may experience more dramatic price swings early in their evolution, as the market attempts to determine a fair price. This trend is clearly demonstrated with Nouns, which has experienced far less variability in recent months as the project has matured.

https://app.hex.tech/6f57775d-7717-4ae7-8924-21da06597348/app/3ac71844-beb4-4416-869c-242f761542b6/latest?embedded=true&embeddedStaticCellId=3c16853c-dafd-4b8c-8b19-6e9f42f39b85&display=iframe&height=550&width=750

Community participation

Community participation is another critical consideration for new projects and DAOs. For the purposes of this analysis, we will consider the impact of auction frequency on two measures of community participation: number of proposals and voter participation.

What is the impact of auction frequency on the number and diversity of proposals generated within a community?

A larger community, holding all other factors equal, will tend to generate a greater number of ideas and proposals. On an absolute basis, therefore, a higher auction frequency seems likely to result in a larger number of proposals.

Additionally, this larger community is likely to be more diverse, and therefore generate a more diverse set of proposals, due to the structural elements of a higher frequency auction. A lower bid price, and more opportunities for entry, expands the set of potential community members relative to a lower frequency auction that may select for a more homogeneous population (e.g., significant economic means, deep interest in crypto projects).

We can already start to observe this trend when comparing Nouns and Lil Nouns: Nouns has averaged one proposal every ~3 days since inception (114 proposals over 343 days) whereas Lil Nouns has averaged one proposal every ~2.4 days (28 proposals over 68 days) [3].

The impact on the number of proposals per holder is less clear. It may be that in a lower-frequency community the average individual holder is more engaged due to the higher entry price and greater accountability fostered by the smaller community. On the other hand, lower-frequency communities may suffer from a lack of enthusiasm or decline in interest over time that leads to a decline in proposals per holder. These communities may find themselves with a relatively disinterested legacy holder base that is minimally bolstered by a slow flow of new entrants. Higher-frequency communities, meanwhile, have a more consistent flow of new holders that may serve to inject new ideas into the community and maintain more consistent interest.

So far, Nouns is outpacing Lil Nouns on this measure. Given that there are ~ 1,400 unique holders of Lil Nouns, compared to ~230 for Nouns [4], we can calculate an average number of each in existence over the lifespan of the project (~700 and ~115, respectively [5]). Though Lil Nouns has produced proposals at an overall faster rate (with ~25% more proposals / day than Nouns ), the difference in proposals generated clearly falls short of the difference in unique holders.

What is the impact of auction frequency on voter participation?

Voter participation rate, defined as the percentage of total possible votes that are cast for each governance proposal, can be a useful metric when evaluating the overall level of community engagement and interest in a project.

A lower auction frequency, while restricting the total number of community members and ultimate votes, is likely to result in a community of more engaged voters who tend to vote more often. Since holders in these communities tend to pay more on average for each vote, and each vote has a larger impact on the ultimate result, there is a larger incentive for individual holders to cast votes for each governance proposal.

On the other hand, a higher auction frequency seems likely to result in lower voter participation. As previously noted, we’d expect the average holder in these communities to have less significant economic means. Voting occurs on-chain and requires participants to pay gas – this may make the average holder in a community with more frequent auctions less likely to vote as they face the double whammy of 1) gas fees and 2) a lower relative value of their vote compared to a holder in a lower-frequency community. Ultimately, this might result in a community where only the largest holders regularly participate in governance.

This hypothesis has been supported so far by the average voter participation rate for Nouns and Lil Nouns, with Nouns averaging a ~30% voter participation rate (roughly double the rate for Lil Nouns). These rates have been durable to date, with voter participation rate generally fluctuating by ~10% around these averages and no noticeable drop-off for Nouns since inception.

*Excludes ‘Canceled’, ‘Active’, ‘Passing’ and ‘'Pending’ proposals

^Voter participation rate defined as number of votes divided by total eligible supply at time of each proposal

Community growth

Defined simply as the total number of holders in the community, community growth is controlled directly by the auction frequency: more frequent auctions will result in a larger, more rapidly growing community.

On an absolute basis, it seems obvious that more frequent auctions will result in both 1) more total votes and 2) more unique holders. Indeed, Nouns has conducted just shy of 350 individual auctions since its formation in August 2021. Lil Nouns, since its formation in May 2022, has already conducted nearly 3000 auctions.

Larger communities, all else equal, will likely make it easier to market and drive new interest to the project – a community of thousands (often) has more reach than a community of hundreds. Additionally, providing more frequent opportunities to join the project may make it easier to sustain and grow interest from the broader public.

However, there are ways that lower-frequency projects can compensate for these potential drawbacks (and that we’ve already seen from Nouns). Fractionalization provides an opportunity for more members to join the community (and at a lower price point). In the future, we may also increasingly see entire DAOs purchase individual NFTs to join lower-frequency communities – this provides a similar benefit to fractionalization in growing the community.

Implications for auction frequency

Based on this analysis, we can begin to lay out a framework for how Nounish DAOs might select an auction frequency:

DAOs favoring a more active community, with a greater number of proposals and diversity of membership, should likely select a more frequent auction structure. These DAOs would prioritize overall community growth, and the network effects associated with a larger holder base, over maximizing community engagement.

On the other hand, DAOs prioritizing a more engaged community may favor a less frequent auction structure. This structure may help to foster a more tightly-knit community, but would likely result in fewer total proposals and (potentially) a less diverse holder base. The cap on absolute community growth should be considered carefully, as the DAO will have fewer holders to propagate its mission and ensure continued interest. At the same time, this slower growth can lead to increased ‘prestige’ from membership and a stronger brand. The Nouns one-auction-per-day mechanism, as an example, has proven to be an exceptionally strong meme that has driven significant interest in the project, helping to propagate the Nouns brand and mission widely.

There is not an obvious right answer today on the question of optimal auction frequency for Nounish DAOs. As more experiments are conducted, and existing projects continue to provide real-world data, we will hopefully get a clearer picture. Nouns, and subsequent similar communities, should encourage these experiments to determine the best way to build the community and grow the treasury.

[1]  Treasury balance as of July 19, 2022 was 25,530 ETH

[2] Mutant Ape Yacht Club is a 20,000 NFT collection, which is double Bored Ape Yacht Club. Nonetheless, the difference in price is ~5x and therefore cannot be purely explained by the increased supply

[3],[4] As of 7/17/2022

[5] Simplified calculation assuming linear growth in number of unique holders over time

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

Nouns is an Ethereum-based NFT project with an innovative structure. Rather than having a fixed supply of ERC-721 tokens that are distributed upfront, 1 Noun is generated every day and auctioned off to the highest bidder. The proceeds from the auction go to a treasury that is governed by the holders.

This interplay between the Nouns auction, treasury, and governance is one of the most exciting DAO experiments. It's a great example of how properly designed mechanisms can cause capital and labor to form around taking real-world action, like commissioning custom artwork to be displayed around NYC or bootstrapping a luxury eyewear brand.

There is a vast design space of treasury distribution strategies, artwork generation algorithms, auction mechanics, and more. It is expected that nounish forks will rapidly iterate and some of these adaptations will make their way into Nouns.

The Nouns Artwork

Each noun has a seed consisting of five traits: a background, body, accessory, head, and glasses. There are currently 2 backgrounds, 30 bodies, 137 accessories, 234 heads, and 21 glasses that can be combined together to create over 40M unique pieces of artwork. The contracts use Ethereum block hashes as a source of pseudorandomness to pick each trait number. For example, Noun #1 has background: 1, body: 20, accessory: 95, head: 88, glasses: 14. You can visit the playground to see all the different artwork. The Nouns DAO can add new trait parts through governance.

Rarity

The most common NFT distribution mechanism has been the 10K PFP collection. In this model, a fixed supply of tokens are available to be minted by the public. The artwork of these NFTs are usually variations of an animal or anthropomorphized object. The minter does not know which token they’ll receive, making the process similar to a lottery. Certain traits appear less often which means they have increased value due to rarity. For the most durable collections such as CryptoPunks and BAYC, we’ve seen rare NFTs assigned a large premium over the floor.

Although Nouns don't have a concept of rarity like typical 10K PFP collections, certain tokens are deemed more culturally valuable than others and have fetched premiums above the average auction price. Noun 1 went for ~6x above the average settlement price in ETH terms, Noun 69 was ~3x, and Noun 316 went for ~2x because of its clean aesthetic.

Rather than solely relying on the token ID number or aesthetics, another way to increase the auction frequency of premium nouns is by introducing invisible trait parts that can be updated once at the owner’s discretion. Invisible trait parts would preserve how artwork is generated, but add a new body, accessory, and head that the market would likely value.

Mechanism to Enable Invisible Traits

Nouns or a fork could accomplish this by adding new artwork for the body, accessory, and head traits. Rather than containing a visible trait, this “artwork” would just be empty bytes. This means that if a Noun seed includes the number that corresponds to this invisible artwork, the NFT will appear to be missing a trait. Potential examples shown below:

This mechanism is compatible with the existing smart contracts and could be added via a governance proposal.

Invisible and Mutable

Although purely invisible traits might have some novelty, this feature could also include the ability for owners to update invisible traits once. Introducing the concept of editable traits to the protocol could have some interesting behavior.

Nouns with invisible traits could be sold on secondary marketplaces while preserving the option to update later. They can be customized by the owner by using a playground-like UI. The owner can also preserve the invisibility of the trait and lock the ability to update forever.

Looking Forward

We are all collectively writing the playbook for CC0 and exploring how nounish communities should try new mechanisms. The Nouns community should encourage forks to experiment and see which features are most useful for growing the treasury and spreading the meme.

Llama is looking for ideas to grow the Nouns ecosystem. We are ramping up our working group and are looking to add contributors. Fill out our form and we’ll get back to you.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

As Llama builds and implements economic infrastructure for DAOs, it’s vital to have a robust source of data from which communities can create reports and assess the results of Llama’s treasury actions. Llama seeks to make its reporting and analytics as accessible as possible by abstracting away the difficult parts of the analysis. After all, what good is open, on-chain data if it’s not accessible?

For good reason, the weapon of choice for this reporting is Dune Analytics. Dune provides a great environment for doing on-chain analytics work. There is an easy-to-learn graphical frontend for making dashboards, with Dune taking care of hosting and all the other backend plumbing. The queries are written in SQL, which has been around forever and many people already know. Dune decodes smart contract data so that it's findable via human readable table names instead of the binary mess that are the internal blockchain data structures.

Building a Dune Dashboard

If you want to build a Dune dashboard, you need a few things first:

You need to know SQL

This is the easy part. If you know selects, joins & window functions then you are already 90% of the way there. If you don't know this, a couple of online courses will get you up to speed quickly.

You need to know how the protocol works

A crypto protocol typically has a collection of smart contracts which execute on the blockchain.

When these contracts have their functions called, or they do something and emit an event, a record is written to the chain which gets picked up by Dune. What these calls and events do, and what the parameters mean, is the puzzle. Some protocols have great documentation, others less so. Often documentation is not up to date with the latest deployed contracts. Seemingly simple things, like calculating APR/APY, or how fees are accounted for, are unique to each protocol. Be prepared to comb through Github and read the deployed contract code on Etherscan to figure out how things actually work.

You need the contracts in question decoded on Dune

Smart contracts are compiled into bytecode which is executed on the blockchain. This bytecode isn't human readable, nor is the data which they write to the chain. Thankfully Dune takes care of this for us.

Using the ABI of a smart contract as the key, this bytecode can be translated back into human readable function calls and event data. Dune handles this via a contract decoding process. This requires a protocol to have published their contract ABIs publicly, either by verifying the contract on Etherscan or publishing a copy elsewhere. It's surprising that this isn't always the case.

If you're lucky, the protocol you want to make a dashboard for already has the contracts decoded. If you're really lucky, they'll even be decoded correctly. Decoding requests are submitted by users, so the input data is not always reliable and hence there can be problems with some contracts. Thankfully, Dune has some friendly and helpful people to help out with decoding issues.

Building for TribeDAO - Putting it into Practice

As part of Llama's work with TribeDAO, I built a dashboard to track the lending rates for the FEI stablecoin across Aave, Compound and Rari Fuse.

This dashboard needed to calculate the amount of deposited FEI in each of these platform pools, as well as the APY received by depositors for both FEI and competing stablecoins. Fortunately, Rari Fuse is a fork of Compound (with each pool a separate instance of Compound) so the code and the accounting was very similar for both protocols.

In these Compound-like protocols, a user deposits FEI into a Pool and receives an amount of cTokens in return. These cTokens act as a fungible deposit receipts, representing a redeemable claim on the underlying assets. As interest is earned on the deposited tokens, the exchange rate of the token/cToken increases (calculated by the smart contract) and a user can redeem their cTokens for an increased number of underlying assets at a later date.

An Easy Problem

Calculating total deposits for FEI (or any token) on a Compound-like protocol is simple in Dune.

The deposit and withdrawal transactions emit a Mint event (deposit FEI, mint cFEI) or a Redeem event (withdraw FEI, burn cFEI) which contain the quantities of the cToken and underlying FEI in the transaction. Here is an event from Etherscan illustrating this for the Rari Fuse Pool 8 FEI deposit contract:

Once we understand how Mints and Redeems relate to the total deposits of FEI, we can query it on Dune. This is simply a matter of grabbing all these events in a query, then summing by day and calculating a rolling total.

A Harder Problem

The two key things which differentiate Rari Fuse from Compound are that each Fuse pool is a full separate instance of the Compound protocol, and it is permissionless, so anyone can create a pool and add assets to it. These features have led to an explosion in Pools and listed assets - there were 200 pools and almost 800 listed assets at the time of this article.

Dune only had 200 assets showing in the tables, and there was no table available which mapped Fuse Pools, cTokens and underlying assets. Further investigation revealed that cTokens are created in at least three different ways, with some being via multicall operations (where call data is obscured) and some via unverified contracts. It was impossible to reconstruct the full picture of Fuse assets via call & event data only.

There was, however, another solution... querying the Fuse smart contracts directly! Dune allows users to submit code which creates custom tables or views, known as abstractions. These can be used to cache data for faster access or add external data to Dune for access in queries. In order to create an abstraction, there were some key pieces of information I needed to understand first:

• The FusePoolDirectory contract has a method called getAllPools() which returns a list of pool names and contracts.

• Each pool contract has a method called getAllMarkets(), which returns the list of cTokens in a pool.

• Each cToken contract has a method called underlying() which returns the contract address of the underlying asset.

To solve this problem for our dashboard (and for everyone else who wants to work with Fuse contracts in Dune), I generated an abstraction table using the contracts above. To do this, I wrote code to iterate through the Fuse smart contracts and generate a SQL table creation query for an abstraction. I used the python library eth-brownie for this, although it can be done in web3.py or equivalent libraries in other languages. The code can be found here and the resulting abstraction is available in Dune as rari_capital.view_rari_fuse_ftokens.

Smart contract developers typically don’t have Dune reporting as a high priority when they develop their protocols. Enabling rich reporting through emitted events has a gas cost, so we can expect to see only the bare minimum of event data written on-chain. Maintaining a flexible approach to available data sources is the key to navigating these problems as an analyst.

Llamas Delivering

As Llama prepared the Quarterly Financial Report for TribeDAO, it became apparent that some of the data was going to be very difficult – or even impossible – to get via our Dune dashboards. With the FEI-Rari merger, Rari Capital income needed to be included in the financial reports. This presented a number of non-trivial challenges to work through.

Income from Rari’s Fuse Pools is generated for TribeDAO via the Platform Fee, a percentage of the interest charged on borrowing. While calculating this fee may sound simple enough, there were several complicating factors:

• This cut is usually 10%, but not always.

• This fee isn't reported out via any events from the Fuse contracts and the fees accrue inside the Fuse contracts for retrieval by the Rari team when required.

Thus, there are no token transfer events which can reliably account for the Fuse fees earned by the protocol. This is unlike many other treasuries where there is a stream of token transfers to designated treasury accounts.

At Llama, we strive to produce the highest quality work possible, even if that means testing the boundaries of the impossible. Instead of excluding Rari income from the report, we found a way to solve our Fuse pool income problem by directly querying the contracts. Here’s how we did it:

Each Fuse cToken contract has a totalFuseFees() method which returns the fees accrued in the contract. By writing custom code to query an Ethereum archive node, we were able to build a picture of the platform fees accrued in each Fuse Pool over time.

This was only part of the solution – the accrued fees then needed to be adjusted for withdrawals already made to the Rari treasury.  Finally, converting everything back to USD required querying on-chain oracle contracts to get exchange rates for the fees accrued over time. In the end, we were left with an accurate picture of Rari’s Fuse earnings over time, which we integrated into our Financial Reports.

Reporting on a protocol’s performance and financial metrics is complex but crucial work. At Llama, we endeavor to produce the most accurate and advanced reports possible. Tools like Dune make sense in a lot of applications, but it's important to understand their limitations and how they can be extended. Llama will use a range of technologies to solve these problems, and build the tools required if they aren't already available.

For more information on Llama’s reporting, please visit our Dune profile at dune.com/llama.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

A lot has been written about the importance of diversifying DAO treasuries in the name of liquidity/capitalization. Today, let’s talk about some of the methods DAOs with publicly traded governance tokens might utilize to raise capital. Specifically, we’re going to focus on the pros and cons of private versus public capital and three specific structures for public capital auctions [1].

Something to note: these are all ways to diversify at scale (i.e., in ways other than simply selling small chunks of tokens on an exchange). Any DAO with a floating token can obviously sell on the open market, but they usually can’t make a dent in their capitalization needs without putting material selling pressure on their governance token given the relative lack of liquidity. These are all methodologies that allow DAOs to alleviate that concern while still raising the funds it needs to operate.

Private Capital

While many compare diversification/capital from a VC fund to a traditional, off-chain private capital raise, the proper TradFi analog is probably a PIPE (Private Investment in Public Equity). Regardless of your choice of analog, tapping institutional investment funds is a simple and common way for protocols to raise money in DeFi.

The common practice thus far has been fairly simple: the DAO’s governance token is exchanged for stables or Ethereum at some discount to its current market price. Discounts have varied but in general the common practice currently appears to be in the ballpark of 30-50%. If that sounds hefty, that’s because it certainly is, but it does usually come with a vesting period in the range of 2-4 years and often includes a one-year cliff, so it isn’t necessarily a steal for the investors. On the one-hand, all the VCs who took part in Lido’s ~$70M raise at a steep discount last May probably feel pretty good about that decision. The investors who funded POOL’s stablecoin diversification around the same time? Probably not so much. Whatever the “right” discount vs. lock-up tradeoff is will likely take some time to sort itself out, but this is more or less what would be considered standard today.

The Positives of Private Capital

This heavily depends on the specific investors involved and what each DAO values as a community. There is certainly merit in the “stamp of approval” marketing you get from being able to boast brand name investors. The price of LUNA ran over ~100% in the week after Galaxy Digital’s $25M investment was announced. Correlation is not necessarily causation, but it’s probably fair to say that this particular private investment discount more than paid for itself immediately.

In addition to the marketing value, you also (theoretically) get a long-term partner that can bring relationships, know-how, and additional governance wisdom to the table. Included in that is a large shareholder who isn’t likely to just dump your tokens as quickly as individual token holders might [2].

As with anything, both of these positive side-effects will vary from investor to investor. Not all VC firms hold the same cachet and marketing value and not all will be incredibly helpful, hands-on partners. This probably goes without saying but in venture not all capital is created equally.

The Downsides of Private Capital

The big downside here is obvious: token holder dilution at a significant discount to market price. This can be especially painful in DeFi where you may feel like you have a bunch of other options to raise capital. To be perfectly blunt about it, there simply is not an analog in the traditional investing world (private or public) where this size of discount to the “market value” of equity is swallowed when fundraising [3]. That doesn’t necessarily make it an unfair trade, but it does make it understandable that protocols and communities would have reticence to diversify this way.

Going down this route is, quite simply, a trade-off between capital at a discount to market and the partnership, stability, and marketing value that specific capital brings. That’s the trade.

Public Auctions

Another way to raise capital without hitting exchanges or negotiating with VCs is simply auctioning off some of your treasury to the public in what effectively amounts to a retail crypto version of a diluted follow-on public offering [4]. Auctions can be limited to existing token holders, an invite-only group of potential new token holders, or not limited at all; part of the beauty of public auctions is that both the audience and mechanism for price discovery are, at least to a certain extent, up to the DAO.

The Positives of Public Auctions

There are a few different things to like here. DAOs get a chunk of capital without being required to slowly diversify in the open market or rely on an institution to play ball at a price they are comfortable with. Auctions will always be at a discount to market price (this is the case by necessity) but the size of that discount will likely be nowhere near the terms you would have to give a venture capitalist. The auction itself can also serve as a marketing event to drive protocol awareness and usage, this value can be compounded if DAOs choose to align the timing of the auction with a product release or something similar. On their face, most free market auctions also stick very cleanly to the ethos of DeFi and crypto more broadly, to the extent you and/or your community are particularly passionate about that.

The Downsides of Public Auctions

Whereas a private capital raise has one massive negative to consider (dilution at a highly discounted price), public auctions have a bunch of smaller issues to think about.

First is the need to market your event. Auctions only work if people show up; pricing terms are heavily dependent on demand. The larger the amount of funds you wish to raise, the more true this fact is. Second is the negative signaling that comes along with selling treasury tokens at scale – this is true of basically any diversification strategy (i.e., “if you’re so confident in your token then why are you selling it?”) but is compounded in this instance by the fact that the sale itself hinges on the aforementioned marketing. That said, negative connotations can be mitigated by being clear about your intentions and rationale for the capital raise (i.e., “we’re raising money specifically to fund initiatives X, Y, and Z which we’re super stoked about and you should be too because…”).

Lastly, off-exchange auctions require a discount to be effective (otherwise participants would just buy on exchanges) and thus require at least some degree of vesting terms to avoid price arbitrage. Given this fact, it’s important to consider the negative effects of a mass vesting event and the downward sale pressure that could be caused by your auction structure and terms. This can be mitigated by including vesting terms as an auction variable or by using a granular linear vest for all participants following an initial cliff. The reality is that individual “retail” investors are never going to be as stable of token holders as institutions will be; your auction terms will need to account for that.

There are obviously a number of different ways to structure a public token auction, today we’re going to consider three. To illustrate the differences, let’s look at a made up set of bidders in a fictional auction. Additionally, let’s assume that a protocol is looking to raise ~$250k in stables with a current market price of $50.

To create a hypothetical “demand” curve, I went ahead and created 100 made-up auction participants with randomly generated buy prices, using $35 as my minimum and $45 – a 10% discount to market – as my maximum. I also randomly generated hypothetical desired token quantities using a minimum of 50 tokens and maximum of 250. Lastly, I took the desired fundraise size of $250k and divided each potential bid price by it to create a theoretical “supply” curve for our auction as well. The results are below:

Note that this isn’t a normal supply curve (hence the quotations above) given we are focusing on a single event and a fixed fundraise figure rather than supply and demand on an exchange. Normal supply would obviously increase with price rather than decrease.

Now that our bounds are set, let’s talk about three different auction types and apply each of them to our set of bidders. One very important point to stress here: as with any economic model this exercise is both oversimplified and assumes bidders are rational actors and behave efficiently with regard to price discovery. There are likely a bunch of behavioral (i.e., FOMO) and technical (i.e., gas prices) realities that lead to imperfections in some of our conclusions here but I think the exercise is both interesting and directionally instructive when thinking about what roles various auction structures play in determining price. This is particularly true moving forward as the crypto community scales and crypto markets ultimately become more efficient.

Methodology #1: Batch Auction

In a Batch Auction [5], a dollar value of capital (or a fixed number of tokens) and a minimum price is set by the DAO beforehand, and a bidding period for the auction is scheduled. Bidders place their bids (both quantity and their own personal maximum limit price) throughout the bidding period. Taking all the bids into account, a clearing price at which the tokens will ultimately be sold is calculated as the lowest price at which the dollar value of capital can be filled while including the quantity of all the higher bids (pro-rating the lowest price bid with the lowest price to fill the remaining tokens/dollar value).

Using the supply and demand curves we created earlier, you can see how an auction structured this way would have theoretically ended up shaking out:

As mentioned before, the particularly nice thing about batch auctions is that DAOs are able to control some of the variables that might matter to them - you can flexibly focus on fixing the dollar value raised (or a fixed number of tokens distributed if you’d rather think about it that way) and you can easily set a minimum price. Assuming enough demand to simultaneously satisfy those two variables, a Batch Auction is able to accomplish price discovery at an individual level, group pricing preferences together, and calculate an efficient, “fair” price and distribute a DAO’s tokens accordingly based on where the supply and demand curves meet.

Methodology #2: Streaming Auction

In a streaming auction [6], a fixed number of tokens is set and a bidding period for the auction is scheduled. Throughout the bidding period participants are able to freely deposit and withdraw funds into a staking pool that represent their “bids.” The catch is that these deposits don’t individually come with a desired price, instead the price is implied based on the number of deposits in the pool, with the size of the pool and current implied price freely available for participants to see, allowing for bidders to hop in and out of the pool if the price is below or above the price they are individually willing to pay. The auctioned tokens are distributed linearly and continuously across the deposited funds in the pool. If you are confused it’s because this isn’t a straightforward concept on its face – Locke (a facilitator of such auctions) produced a simple, helpful video to explain more or less how it works.

Let’s assume for our hypothetical auction that we set aside 6,500 tokens to be continuously auctioned off over the course of our bidding/staking period. Again, we are assuming all bidders behave efficiently and rationally (a big assumption in general and a massive assumption in crypto, particularly today). Regardless, we ran the auction amongst our fake set of bidders and this is how it played out:

From a price discovery perspective the outcome is very similar to that of a Batch Auction. At least in theory, when auction price discovery occurs at a group level, irrespective of the precise auction mechanism, you are going to get a crowdsourced price outcome where supply meets demand. One positive to this particular structure is it is simply more transparent than the Batch Auction (i.e., participants are able to see current pricing terms and get in and out based on their own preferences). Bidders obviously enjoy that level of transparency/empowerment and it also could theoretically serve to generate perceived demand (read: FOMO) if it looks like more and more bidders are entering the auction pool. That said, transparency can just as easily cut the other way if you aren’t ready and willing to act as a stabilizing agent and do a bunch of staking in your own auction to keep the price afloat should things go poorly.

The primary negative with regard to the Streaming Auction is control. Unlike other auction choices, DAOs must by necessity set auction supply based on a fixed number of tokens (hence the straight red line in the graph above) rather than having the ability to set a fixed dollar value. This is compounded by the fact that if a DAO wishes to set a minimum price the primary way to do this is to simply stake in your own auction which may lead to your DAO acquiring large chunks of its own governance token back into treasury. DAOs can obviously account for this by planning to auction off more tokens but the variance between what your own staking position might net you is fully dependent on the demand your auction generates and thus more or less out of your control.

Methodology #3: Dutch Auction

In a Dutch Auction, a fixed number of tokens or dollar value of capital is set by the community beforehand, and a bidding period for the auction is scheduled. Over the course of the auction, the terms slowly change and become more favorable for bidders the longer the auction goes on – but with the catch that once the allotment is filled, the auction is over. DAOs can set-up a “fixed-vest” or “fixed-price” auction or some combination, depending on preference. In a fixed-vest scenario the bidding would begin at the current market price and scale down linearly over time to some set discount (5%, 10%, whatever you like). Bidders can buy at whatever price they please, subject to a set vesting period to avoid immediate price arbitrage. Conversely, in a “fixed-price” scenario the discount to market is set beforehand and the vesting period floats (starting larger and gradually getting smaller) over the course of the auction.

Assuming a “fixed-vest” Duction Auction with a floating price applied to our supply and demand curves illustrates the material difference relative to the other auctions structures:

Note that this outcome obviously assumes a few different things (the auction is truly done “blindly,” there is enough participation to avoid collusion, etc.) but because Dutch Auction price discovery is done and transactions are cemented individually rather than on the group level, it theoretically allows protocols to capture the most possible value throughout the auction.

That all being said, there is certainly a negative connotation with this type of structure given you are obviously selling the same token to different people at different prices (the specific economic term for this type of pricing is quite literally “price discrimination” and it’s most closely associated with airlines, so that should tell you something). An emotionless economist would tell you that this sort of negative connotation is nonsense - everyone is transacting at prices that they are clearly willing to pay. And at a discount! That said, I don’t know a ton of emotionless economists in crypto - my guess is that while more efficient and better for your current token holders, this wouldn’t necessarily be a super popular route if you are particularly concerned about PR.

In Conclusion

When DAOs make the decision to diversify their treasury and raise funds there are a number of different variables to consider with regard to source of capital and pricing structure. This is nowhere near an exhaustive list but hopefully can be utilized by protocols and communities in the future when approaching this decision and thinking about the pros and cons of various strategies.

[1] This is not an exhaustive list. There are many types of auctions.

[2]  DeFi is probably way too nascent for publicly traded protocols to be thinking about shareholder optimization yet, but this stuff does matter.

[3] The only real comparison I can think of is underpriced IPOs but at least in that case you can half-heartedly argue there was no set market price to begin with.

[4] Note that an at-the-market offering is a version of a follow-on that would be more analogous to algorithmically diversifying over time via an AMM. We’re basically just focusing on a single event, at scale version of this.

[5]  Would refer to Gnosis Batch Auctions as an example of this.

[6]  Would refer to Locke Protocol’s Streaming Auction as an example.

Stay Up to Date on Llama 🦙

Follow us on Twitter, check out our website, and subscribe to our Substack.

Artwork credit: 0xEFRA.

Introduction

Metagovernance has been a recurring topic in the discussion of governance in different crypto protocols. In this article, we will define metagovernance as holding one DAO’s token in order to influence decisions in another DAO(s). This definition is borrowed from this post.

Metagovernance is possible because of on-chain composability between protocols. One protocol could acquire voting power in another protocol and have direct influences through on-chain voting. Metagovernance has been used to bootstrap liquidity, deepen liquidity, and increase yield. Some of the prominent players utilizing metagovernance in these regards are Convex Finance, Abracadabra Money, Yearn Finance, Redacted Cartel, and Frax Protocol.

Although these account for most value transfers in the metagovernance space, a few more nascent and less explored areas of metagovernance are strategic collaborations such as the metagovernance event among Fei Protocol, Index Coop, and Aave. Fei Protocol established a strategic position in Index Coop’s governance token and listed itself onto Aave, in which Index Coop holds voting power.

In this post, I will brainstorm on the future of metagovernance and the potential future where metagovernance becomes a more common theme in the protocol governance landscape.

DAOliticians for metagovernance

If the metagovernance trend continues, I hypothesize that different interest groups will form political-party-like DAOs to advance their agendas. I will call these political-party-like DAOs “DAOlitical parties,” and key individuals representing these parties “DAOliticians.”  They are pronounced like political and politicians but replace the po- sound with dao.

Metagovernance introduces a venue for voters to unionize, decreasing friction to vote and increasing their collective voice (and/or yield). The most primitive example of this idea is Convex Finance. Before Convex Finance, $CRV (Curve Finance’s governance token) holders could lock their $CRV within Curve Finance to boost their LP position and vote on the gauges. This posed a great challenge for $CRV holders without an LP position. Convex Finance created a marketplace to match LPs with $CRV token holders to solve the lack of LP issue. To crowdsource voting power among $CRV holders, Convex Finance introduced an interesting dynamic in which $CRV holders could exchange 1 $CRV for 1 $cvxCRV. After the exchange, $CRV holders would have additional yield compared to locking their tokens on Curve Finance directly. In exchange for this higher yield, $CRV holders delegate their $CRV voting power to Convex Finance. This voting power is then controlled by the holders of $CVX, Convex Finance’s governance tokens. Through this design, Convex Finance was able to acquire more than 50% of the voting power in Curve Finance. So, one may even say that Curve Finance is controlled by Convex Finance. There are problems with this design as it encourages vote bribing but it is a primitive version of unionization nonetheless.

Taking a step back, the interaction between Curve Finance and Convex Finance can be abstracted into the following:

A group of people within Curve Finance wanted to maximize their yield on their $CRV (their agenda). However, individual voice is not enough to shift the pendulum. So, they pooled their voting power together (unionizing) in Convex Finance (the union) to achieve it.

Although these agendas currently revolve around additional monetary value, they could evolve into more specific agendas without a direct monetary value. For example, if an interest group were to form around decentralized stablecoin’s adoption, it could adopt a similar format as Convex Finance. Since this interest group would no longer be purely driven by financial returns, it could cast its voting power just for decentralized stablecoin adoption even though there might be higher ROI voting strategies.

In the union model of metagovernance, an important role would emerge. I call it the DAOlitician. A DAOlitician’s main role would be convincing others token holders to join their cause and delegate tokens to their union. Surprisingly for Convex Finance, a dedicated DAOlitician did not exist. This is possible because, in terms of monetary gains, numbers speak loudly for themselves. However, for more nuanced topics, such as stablecoin adoption or staking derivative adoption (eg. stETH), DAOliticians would be extremely important in generating buy-ins and winning mindshares from token holders.

DAOliticians are similar to politicians but also different in important ways. We will use the United States’ representative model in comparison since many other democratic systems share similar structures. First, a DAOlitician’s voting power is different from another DAOlitician’s, but all politicians share the same voting power. For example, each congressman/woman’s vote is counted as one, regardless of the number of constituents in his/her district; but DAOlitician’s voting power is determined by the votes delegated to its union, which is different. Second, politicians are elected with term limits but DAOliticians are on a continuous election cycle (if we assume vote delegation can change easily), reflected in their voting power’s change. Third, in the union model, the DAOlitician’s responsibility is narrower than the politician’s. Politicians normally govern on a wide range of topics from public health to military spending. However, for a DAOlitician in the union model, his/her duty mainly revolves around one set of issues. This narrow scope for unions would be an issue for it to scale. Since each token holder can only delegate to one union, what if he/she wants to advance more than one union (decentralized stablecoin adoption and staking derivative adoption)? Since naturally, token holders would care about different agendas all at once, I propose a higher level DAOlitical party would form on top of different unions to fulfill this need.

DAOlitical Parties for Multiple Protocols

In the previous section, we explained how Convex Finance can be thought of as a union within Curve Finance, how this union model can be applied for other interest groups, and why a DAOlitical party is necessary. In this section, we will continue this train of thought and discuss the implication of metagovernance when interacting with more than one base layer protocol.

Cross-protocol DAOlitical parties will form because most token holders hold more than one token and to achieve any sets of agendas would require more than one base layer protocol. In the decentralized stablecoin adoption example, while Curve Finance is important for trading experience, other factors such as decentralized stablecoin’s utility are equally important. These utilities include but are not limited to: borrowing+lending, education+onboarding, treasury diversification, and many more. Therefore, having voting power in one base layer protocol, such as Curve Finance, is not enough. The DAOlitical party should hold a wide range of tokens (or voting power in those tokens) to successfully advance its agenda.

Index funds are great DAOlitical party vehicles. They hold a significant stake in multiple tokens’ voting rights. Although they currently do not have any stated agenda on voting, other protocols could acquire index funds’ governance tokens to propose such changes. An early example of this is the interaction among Fei Protocol, Index Coop, and Aave. Fei Protocol acquired a long-term position in Index Coop and utilized Index Coop controlled Aave’s voting power to list $FEI onto Aave. You can see here for a detailed walkthrough of the event.

A possible future may be represented by the diagram above. For each base layer protocol, there will be different unions formed around it aiming for different purposes, such as yield maximization, tokenomics reforms, or any other protocol-specific issues. On top of these unions are DAOlitical parties (could be in the form of an index fund). These DAOlitical parties combine different agendas together. DAOliticians attract capital and also voting power for their DAOlitical parties. This reminisces the political structure in the real world. The parties are political parties and each union is a lobby or interest group. The base layers are fundamental issues such as health, military, education, etc.

Implications for Our Future

While many early versions of this structure already exist, such as Redacted Cartel + (Convex, Frax, OHM) and Index Coop’s Metagovernance Committee, no protocol has a clearly stated agenda. Even if an agenda exists, it is hard to differentiate it with others. We believe this is expected at the stage of crypto. Most stakeholders share similar beliefs about how protocols should operate. However, as crypto matures, we will see differences between stakeholders increase. One related example is the current trend of DeFi, crypto games, and NFTs slowly moving into their own camps on Crypto Twitter (CT). They might all benefit from the same base layer protocols, but each has a different agenda.

Another important trend to note is the financialization of votes in the crypto world and its impact on this structure. Since each vote has a price, which is the token price, the amount of capital a DAOlitical party controls will determine its voting power. While similar to real-world politics (where a political party’s budget is important), in crypto, a DAOlitical party will never win given a smaller capital size (while a political party with less funding could beat a better-funded opponent). This over-financialization might not be the most suitable voting model in the future if we were to consider some protocols as public goods.

The role of DAOliticians has been taken on by protocol founders: Joey for Fei Protocol, Do Kwon for Terra, and Sam for Frax Finance. These DAOliticians are active in different forums to spread their ideas. They can take up this responsibility because they are experts in their respective domains and have a personality that people like. However, as the protocols and different verticals mature (presumably with more regulations), the role of DAOliticians will be shifted to external individuals who can better facilitate cross-protocol cooperation.

Closing Summary

In this piece, we brainstormed about the future of metagovernance. We begin with the union model, inspired by Convex Finance, on how token holders could unionize to achieve a common agenda. Then, we introduce the concept of DAOlitical parties to allow users to contribute to more than one union. Lastly, we discussed a future where DAOlitical parties expand their control to more than one base layer protocol through DAOliticians.

Crypto is an amazing testing ground for new political and coordination systems because of its composability and interoperability. With great power comes great responsibilities. We know many inefficiencies exist in our current political systems and the DAOlitical space could only be different if we are conscious about the design choices we are making and the things we are optimizing for. Although there are differences between the real-world political system and the DAO world, these differences alone are not enough. We hope this article serves as a way to jumpstart this conversation and design a sustainable system together.

Written by Yuan Han Li (@yuan_han_li) and BraveNewDeFi (@bravedefi).

There has been a cascade of talent flowing into DeFi in the last year. The DAO structure provides community members and contributors alike with greater autonomy and the freedom to work without the hindrance many of us have experienced in traditional workplaces that depend on cubicle culture and a panopticon-like approach to supervision.

Many DAOs are trying to onboard more non-technical contributors and scale their communities. No DAO uses the same approach, but it’s undeniable that small working groups empowered by token holders can more effectively scale a DAO than putting every decision to a governance vote.

Community-driven working groups are the most effective way for DAOs to engage their communities, tap into the talent lurking in their social channels, and give community members skin in the game and the resources needed for success.

Llama contributors have helped spin up a number of working groups, such as Aave’s Grants DAO, GitcoinDAO’s Public Goods Funding Workstream, Nexus’ Mutant Marketing, and PoolTogether’s Treasury Working Group. Below, we provide an overview of certain approaches, best practices, and ways in which DAOs can use specialized groups to tap the talent in their communities and put those skilled users to work for the community.

What’s the point of working groups anyway?

Scaling Governance Velocity

Depending on whether they have been imbued with executive power, working groups scale governance in one of two primary ways.

The first way working groups help scale governance is by directly removing voting overhead. Specifically, a DAO can empower working groups with the ability to execute, allowing dynamic working groups to move fast and execute without putting every decision to a vote. For example, Yearn’s Governace 2.0 proposal outlines a thoughtful way to delegate executive power to numerous working groups. Similarly, grants working groups at Aave, Uniswap, and Compound are allocated a budget every quarter to fund grants and report back to governance. If the working group has discretionary power, then this means that governance will no longer be burdened with making certain decisions. This reduces the voting overhead of tokenholders, allowing the DAO to work efficiently while voting less.

Working groups can also operate on a purely advisory basis without the need for executive power (e.g., PoolTogether’s Treasury Working Group or Nexus Mutual’s Investment Hub). A working group formed via governance receives its mandate from the community. Specifically, token holders vote to grant a working group legitimacy within the DAO and provide the group with a de facto mandate to investigate, research, and advise the DAO on certain decisions the DAO ought to take. Advisory working groups provide some form of expertise to their community, which empowers the community to make more informed decisions. There are a variety of DAOs that make important financial decisions without fully understanding the potential outcomes or impacts a vote may have. Creating an advisory working group scales governance by reducing the time it takes to achieve consensus among members of the DAO. Increasing the pace and quality of decision making is the key contribution these groups provide.

Put simply, good working groups reduce coordination costs within DAOs; they catalyze effective action with less bureaucracy.

Scaling Operations

DAOs can increase productivity by adding specialized working groups. Instead of taking a bottom-up approach to every decision, a DAO can form two-pizza teams to tackle many issues at once. It's a simple division of labor. Over time, these working groups will develop expertise, attract better contributors, and become more effective. Every DAO can achieve greater forward momentum this way.

Whether it's factories, semiconductors, corporations, or even blockchains, we have seen specialization and modularity win out time and time again. DAOs are no different.

Now that we are clear on how working groups can improve your DAO, let’s dive into a framework to help decide what powers and responsibilities DAOs ought to delegate to its working groups.

Powers and responsibilities to delegate to working groups

When a DAO is evaluating new working groups, there’s one sound principle to keep in mind: form follows function. Each working group should be granted only enough power to achieve their established objective and key results.

Working groups should always have a founding charter that the community can refer back to.

• What is the purpose and objective of the working group?

• Is the working group set up to be temporary or permanent?

• What degree of power is required?

• What type of skillsets/roles are required?

• How is decision making performed in working groups with executive power?

Let’s review PoolTogether’s Treasury Working Group (TWG) as an example.

During the formation process, the founding members of the TWG outlined their mandate and core deliverables. The mandate is to:

• Ensure safety and prudence go into every decision

• Drive future protocol growth

• Identify sustainable solutions

• Present objective data to the community

PoolTogether’s TWG acts in an advisory capacity but has access to enough funding to pay TWG contributors. This allows the TWG to attract skilled contributors but does not lead to counterparty risk issues with treasury funds.

Checks and Balances

Working groups are granted legitimacy by the DAO and held accountable by its members. This is why the founding documents are crucial to the long-term success of any community-driven team. Should incentives between the DAO and the working group become unaligned, then checks need to be carefully constructed to restrict or remove power from a working group.

For instance, Yearn’s Governance 2.0 uses the “constrained delegation” model, which allows YFI holders to grant yTeams (working groups) certain powers. If yTeam contributors are effective and act in alignment with YFI holders, then token holders need not take action. Should a yTeam become unaligned or misdirect their contributions, then the community can redirect them, restrict powers, remove team members, and even dissolve the team.

Best practices:

• Limit runway to 3-month or 6-month timeframes

• Give token holders power to restrict or remove powers

• Allow the DAO to redirect working groups

• Allow the DAO to dissolve working groups

• Though not a Check or a Balance, it is also important that the working group be filled with members with the right skills/roles

Transparency and Reporting

There should be formal and informal reporting requirements in place when working groups are established. After all, working groups should provide proof of work if they are receiving compensation from the treasury.

Examples of formal reporting:

• Quarterly/biannual performance reviews (see here for a performance review of Aave’s Grants DAO)

• Weekly working group updates (see here for an example of our work with PoolTogether)

• Regular budget expense analysis (Aave Grants provides a list of funded and outstanding grants that are updated in real-time)

Examples of informal reporting:

• Open discussion in community social channels (e.g., Discord, Telegram, etc.)

• Presence in forum discussions

Transparency in discussion is the best approach to informal reporting. PoolTogether’s TWG has read-only channels where TWG members can discuss deliverables with a general channel in the Discord server for community members to add comments and suggestions or ask questions.

While radical transparency is the goal, there are many different approaches that can work for each DAO. The point of transparency and reporting, at the end of the day, is so that the broader DAO can easily verify the work that working groups do without holding them back. Whatever form of reporting a DAO chooses should enable the community to evaluate a working group’s performance and determine whether or not to renew their funding, or possibly dissolve the working group altogether.

Budgets

As stated above, a working group should have a well defined set of objectives, key results, and deliverables outlined when requesting a budget. Each DAO should grant as much funding as necessary to allow a working group to achieve their objective.

The usual instinct is to start with a small amount of funding and scale up a working group; however, funding is the primary way to attract talent to a community-driven team. Constraining funding too much will likely constrain the ability to attract quality contributors. A DAO should strive to create a culture where talent and performance is rewarded.

Best practices:

• Budget funds should be held in a multisig

  • At least 4/7 signers to avoid any central point of failure

  • In special cases where many transactions require synchronous signing, a working group may opt for a lower threshold

• Regular budget analysis should be a requirement

  • Expenses should be tracked and justified

  • Gives token holders one way to evaluate performance

• Contributor payment should be at least 25% in native governance tokens

  • Stablecoin payments are necessary for cost of living, taxes

  • Skin in the game is necessary to align interests

  • Performance bonuses should be in additional governance token compensation

By the DAO, For the DAO

DAOs provide a powerful way to align interests around a common goal, but as communities grow, the goals become multifaceted. The heavier the burden on token holders, the greater the chance voter fatigue will plague a DAO. Allowing for specialization within a DAO and granting limited power to community-driven working groups will enable DAOs to scale, put talented contributors to work, and realize greater success. We see temporary or permanent working groups as the optimal way to sustainably scale a DAO without sacrificing the qualities that make DAOs such an effective way to coordinate around shared interests.

Written by: Daniel Schlabach (@dmschlabach) and 0xJosh (@0xJosh_).

The overwhelming majority of DAOs hold over 90% of their treasury in the native tokens.

While this demonstrates conviction in the future of the token, it also poses serious risks to the DAO, hampering its ability to make reliable budget projections and to fund innovation and ongoing operations.

There are many ways a DAO can get access to stablecoins. These include selling native tokens in the open market for stablecoins, exchanging native tokens for stablecoins with strategic partners, and borrowing stablecoins against treasury collateral, among others.

This post outlines a template for treasury diversifications via strategic partnerships. It also covers why DAOs need stablecoins in the first place and what community-run strategic partnerships look like.

The Need for Stables: Why DAOs Need to Diversify Their Treasury

Having >90% of the treasury in native tokens is both dangerous and inefficient from an operations perspective. Instead, DAOs should have 1-2 years' worth of operating expenses in stablecoins.

Diversifying a portion of the treasury into stables allows the DAO to:

• Lower risk and volatility in the treasury - many governance tokens are highly volatile and unpredictable, causing treasury value to fluctuate on the order of millions of dollars.

• Operate from a position of strength in a market downturn - ensuring the DAO can continue to grow and take advantage of opportunities, even in bear markets.

• Have a stable, predictable source for funding innovation and operations - it's hard to budget when your treasury swings 20% (or more) overnight.

• Lower operational costs - swapping native tokens for stables each quarter is inefficient and costly.

• Put tokens in the hands of long-term, like-minded partners - rather than paying contributor salaries in tokens or selling on the open market, DAOs can ensure their tokens end up in the hands of aligned partners. This also removes a source of selling pressure from the market.

If the diversification is done thoughtfully, the DAO can find a win-win: gaining stables to pay contributors while also distributing their token to those who will help further the DAO's mission.

Community-Centric, DAO-Native Strategic Partnerships

Here, we propose an approach to DAO strategic partnerships that places the community at the heart of the process.

DAOs are a remarkable sociological and financial innovation. They bring together contributors from all corners of the world to work on interesting problems. By placing a high emphasis on trust, fairness, and transparency, DAOs are community-first by default.

A strategic partnership should also be community-first. Like any other DAO action, community involvement and feedback should be a part of any treasury diversification. Done correctly, the process can be as efficient and agile as behind-doors processes.

A DAO's strategic partnership should also be done in a DAO-native manner with the DAO's mission in mind:

• Be lean, efficient, and agile - DAOs shouldn't have to compromise on efficiency. Empower a few representatives from the community to run the process.

• Regularly collect feedback from the community - Involve the community without turning every decision into a proxy fight by soliciting community feedback at key, predefined parts of the process.

• Incorporate the DAO's mission, culture, and values - Each DAO has a different mission, culture, and values. There is no "one-size-fits-all" list of values that DAOs should use when conducting community-led strategic partnerships. Instead, the community should help decide what values to prioritize when looking for strategic partners. For instance, some communities may look for VC partners that can help with immediate scaling and provide hiring advice, while others may look to DAOs for partnerships and growth strategies.

• Use the diversification to forge important partnerships - The diversification should be viewed as a way to engage with like-minded partners. As much as possible, DAOs should include other DAOs, individual contributors, and DAO-friendly VCs in the diversification. Those participating should be partners in the truest sense of the word.

Steps for Running a Community-Centric Strategic Partnership Process

Starting from these first principles, we developed an actionable framework for running such a process:

1. Form a 5-6 person treasury diversification working group from the community to run the process.

2. Draft a comprehensive governance forum post, detailing the logistics of the diversification and the types of strategic partners desired.

3. Vet strategic partners based on select criteria established by the DAO and working group, and ask these partners to speak to the community (on a platform such as Discord or Twitter Spaces). Potential partners should elaborate on their interest, their value-add, and vision.

4. Have the working group deliberate and form a shortlist of potential strategic partners. Share this shortlist (with rationale for each partner) with the community and allow for public debate on governance forums and/or community town halls.

5. Draw up the final details of the diversification following community feedback and conduct an on-chain governance vote for the proposed new strategic partners.

This framework is designed with the objective of being open, transparent, and community-led.

Of course, many of the details within these steps depend on the community and its governance structure. For larger DAOs, the community may require Snapshot votes at each step of the process. For smaller and more agile DAOs, consensus may be reached simply through conversations on Discord and governance forums.

Gitcoin's Treasury Diversification

Llama recently proposed a $15 million treasury diversification with Gitcoin using this template.

Gitcoin is building and funding the open web. By providing infrastructure and tools that help coordinate developer funding, crypto education, and project development, Gitcoin is one of web3's most recognized and integral platforms.

Over 90% of Gitcoin's treasury today is held in $GTC tokens, despite the fact that they have $3-4 million in operating costs each quarter. Using the framework for community-first treasury diversifications above, Llama proposed that Gitcoin diversify $15 million of $GTC into stablecoins (24% of the vested treasury and 2% of the total treasury at the time of the proposal).

After approaching the community and asking candidates to apply for the working group, GitcoinDAO formed a 6-member group composed of key Gitcoin community members.

These community members have worked together to come up with criteria for strategic partners that match Gitcoin's values. The ideal strategic partners:

• Deeply understand Gitcoin and its mission (and is DAO-native/DAO-friendly)

• Are thoughtful and open-minded

• Are long-term thinkers that view the relationship with Gitcoin as a partnership

• Act with the highest integrity

• Have a clear value-add to Gitcoin and can help drive Gitcoin's mission forward

The working group's goal is to assemble a contingent of partners that bring a diverse range of experiences, backgrounds, and skill sets to Gitcoin. The group should include a mix of visionaries, operations experts, technical analysis pros, and financial gurus who all understand and deeply believe in Gitcoin's mission.

Conclusion

Like DAOs, community-centric strategic partnerships are here to stay.

By involving the community in key processes such as treasury diversifications, DAOs can ensure that they build long-term sustainability in an inclusive and open way. The best organizations succeeded because they were able to bring people with them - DAOs should be no different.

Written by: Shreyas Hariharan (@HelloShreyas).

More complex behavior has moved to the digital world, but our tools to coordinate in a digitally-native way haven't evolved at the same pace. DAOs are the defining coordination mechanism of humanity. They champion values of decentralization, programmability, and community ownership. Llama aims to be at the forefront of this innovation by building infrastructure that helps DAOs flourish.

DAOs enable communities to encode decision-making and transparently govern over a shared pool of resources, i.e. a community treasury. Over the past year, Llama has worked with some of the most prominent DAOs including Aave, Uniswap, Gitcoin, PoolTogether, Fei Protocol, Index Coop, FWB, among others. We’ve helped them develop treasury strategies, design liquidity programs, construct on-chain indices, and build dashboards and financial reports.

Through our work, we have established best practices for DAO to DAO relationships. We do not view ourselves as external parties, but rather as active contributors within the DAOs we work with. While the boundaries of a B2B relationship are strict, the boundaries of a D2D relationship are fluid. We encourage our community members to own our partners’ native tokens and be active participants in their governance programs. A Llama <> Aave treasury working group should attract the best within both Llama and Aave to help allocate Aave’s treasury effectively.

Contributors are at the heart of what makes Llama special. We have 35 contributors who operate by a set of shared values:

• Pack animals: Llamas trust each other and generously give credit where it’s due.

• Experimentation: Llamas move fast and experiment.

• Optimism: Llamas believe that the future is malleable. Anything that isn’t impossible according to the laws of physics is possible with sufficient knowledge.

• Rigor: Work done by llamas is of the highest quality.

We continuously monitor community health by tracking various KPIs. DAOs enable fluidity of contribution and collaboration, which means it’s often much easier to attract top talent than it is to retain it. Our metrics include:

• % of contributors who are active

• % of contributors retained after 3 months

• Time between joining & first contribution

• Time between contribution & payment

The Future of Llama

To date, our focus has been on customized treasury solutions. This has allowed us to identify generalizable problems across communities and develop frameworks for important treasury actions such as stablecoin diversification, liquidity incentives, token swaps, and more. We are now building workflows that enable communities to allocate their treasuries effectively and assess results. We are building economic infrastructure for DAOs.

As Llama enters a new chapter, we want this new direction to be reflected in our visual identity. We worked with VectorDAO to refresh our brand, color palette, and website.

Treasuries require innovation, not just management. Rather than sitting idle, they should be used to catalyze growth, sustain communities, and advance the DAO economy. We are building infrastructure that unlocks this potential.

We are growing our full-time team, our community, and the DAOs we work with.

• If you are an engineer interested in DAOs, email jobs@llama.xyz.

• If you want to join the Llama community, apply here.

• If you are part of a DAO that wants to collaborate with us, fill out this form.