The original text comes from Photon Finance. The author is Dave Kajpust, CEO of DeFi development company Ekonomia and developer of Photon Finance. It was compiled by Baize Research Institute and slightly deleted.
Blockchain-based governance systems suck. Blockchain technology holds the promise of changing the world, but so far the governance systems we’ve created are pseudo-decentralized token voting where insiders and whales can also vote.
However, the design of governance systems has only just begun. A bear market is the perfect time to experiment, when markets are calm and thinkers and innovators are busy creating new trends.
In terms of governance systems, a good starting point is to develop a clear set of principles that will guide us in creating a durable decentralized governance system. As a development company, we have established the following key principles:
Strive to be as permissionless, deterministic and transparent as possible.
Voting power must be decentralized throughout the life of the protocol.
Users of the protocol should be given voting rights (such as owning in-game skins).
The incentive mechanism should focus on rewarding users who participate for a long time.
These four principles serve as a starting point to demonstrate to regulators that DAOs are different from corporations and should be regulated differently.
At Ekonomia, we have created eight new governance concepts that adhere to these principles. These concepts are somewhat unfinished - the goal of this article is to share them, get feedback, and work with others to improve them.
DAO Legal Pioneer
It’s important to start designing better governance systems at this moment, as regulators have set their sights on crypto protocols, as we’ve seen recently with Ooki DAO and LBRY. On Twitter, the consensus among crypto enthusiasts is that the CFTC and SEC overextended their powers in both incidents. But the reality is that there are many on-chain companies masquerading as DAOs today, and we need to do better.
Delphi Labs legal counsel Gabriel Shapiro recently explained in a tweet how to create a DAO that meets the appropriate legal thresholds.
We need to make DAO a true “decentralized autonomous organization.” Gabriel has another tweet that dives into some very interesting ideas. I summarized it:
Split the DAO into subDAOs that vote on protocol specific parameters.
All subDAOs are combined to form an uberDAO that votes on protocol parameters that affect everyone.
You must actively participate in the management of the agreement.
Fund the protocol, not the founding team - Instead of pre-selling tokens to investors, the protocol pre-sells tokens to early users of the protocol. These early users will have to deposit liquidity or pay fees to access their earnings in protocol tokens.
Continuing these thoughts, Gabriel posted a blog post about how DAOs work, from which I extracted some key points:
DAO tokens should only control code, code is law.
The DAO should not control people, and voting should not legally bind anyone to complete tasks. This means that the token confers legal power over the person, which looks a lot like a typical corporate contract. It increases the risk that the token is a security.
The main purpose of the DAO protocol is to provide users with autonomous digital infrastructure - a smart contract system that can be changed through on-chain voting. In reality, DAOs are just “MMORPGs” (massively multiplayer online role-playing games) used to adjust the system parameters of decentralized applications.
The second purpose of a DAO is to achieve off-chain, rough consensus on loose issues. These consensuses can occur on Twitter, in forums, or at meetings, and are non-binding and have no strict numbers.
Gabriel's perspective gave me ideas for developing new governance concepts.
For years, Crypto has been an underappreciated “renegade industry,” with good engineers and lawyers hard to find. But it has grown rapidly, and many governments are now working to regulate cryptocurrencies.
Luckily, a lot of incredible talent has entered the crypto industry over the past few years. We now have the legal and engineering expertise to design novel governance systems so that everyone can use it.
New concept of governance
I’m not a lawyer, I’m a technologist, so I think about governance in terms of DeFi protocols.
The common pattern among these concepts is that they all satisfy the four principles at the beginning of the article. Furthermore, these concepts have a lot of crossover with each other and can even be used together. All of these concepts are isolated ideas, but we are thinking of a way to bring them together to create a fully interoperable governance system.
1. Seed mobility
Gabriel once pointed out: "Pre-sale of tokens will forever taint the security of the token."
That's just the way it is. Over the past two years, we’ve seen the same VC names on many new crypto projects, and these projects raise funds in the same way as traditional startups. It’s also clear that there are a few crypto VCs who seem to be involved in every funding round.
In comparison, an Initial Coin Offering ( 1 C 0 ) seems fairer because anyone can join. The rules of participation are the same for everyone and can be verified on-chain. But regulators launched an “attack” on 1C0 in 2018 and forced new projects to raise funds from venture capital firms. This has led to the scenario we are in today - most tokens are VC-backed and risk being considered securities.
We need to improve our fundraising process to make it more equitable. Blockchain technology allows us to be creative and innovate new ways to raise capital. This is where the concept of “seed liquidity” comes into play.
How does it work? Users invest in liquidity for a period of time before the protocol launches and receive governance tokens. This is all done without permission via smart contracts, which allow anyone to participate.
With this concept, you can raise funds and build a real-time protocol, and everything is done on-chain! There are no meetings with VCs or angel investors, no term sheets, and no information asymmetry between early-stage insiders and the public.
The development teams would do better if most crypto projects raised funds that were deposited into the protocol instead of the development team's account. Some protocols do very similar fundraising mechanisms. OlympusDAO raised funds in DAI to launch their protocol. Early participants can receive OHM in return at extremely low prices. The main difference here is making the protocol’s liquidity more sticky and structural, as OlympusDAO “bonds” typically don’t last longer than 1 week.
Providing long-term liquidity locks, rather than traditional fundraising, can also be used to fund protocols to ensure their success. The protocol can plan for seed liquidity locks, Series A liquidity locks, Series B liquidity locks, and so on until it becomes a self-sustaining protocol.
The beauty of these liquidity locks is that they are priced in an open, competitive market, creating fair pricing for everyone and no insider advantage. The governance token cannot be purchased, as it is distributed to users who provide liquidity. It has no price until the protocol is launched and trading begins. This makes it not function like a security, where profits are only expected.
2. Liquidity mining partners
There are also scenarios where DeFi protocols can work with well-known DAOs/protocols that can provide users or liquidity. We call them “liquidity mining partners”.
Let’s take an example using Photon Finance.
Photon Finance is a stablecoin protocol built with a modular design – each module takes different collateral and has different PHO stablecoin minting capabilities. These modules contain assets from other DeFi protocols such as AAVE, MKR, FRAX, COMP, BAL, DAI, LUSD, and more.
DeFi is open and permissionless, so there are synergies in connecting protocols and creating cross-protocol incentives. In the past, we've often seen "partnerships" between protocols announced in press releases. But what's the point? Often, they are just "vague agreements" for cooperation between two companies. This makes no sense for the protocol. Because any cooperation agreement should be completed through on-chain code.
Here’s how a liquidity mining partnership works:
Provide partners with rewards of up to 20% of the governance token TON.
Some DAOs with large amounts of liquidity and a large number of users want to receive this 20% reward.
The DAO can now begin directing liquidity or its users to Photon.
Photon builds a revenue aggregator like Yearn. TON rewards can be sent directly to the DAO or DAO users based on the success of the project.
Lending protocols like Aave or Euler are incentivized to use PHO as an asset and send rewards directly to their protocol’s DAO.
Offer decentralized exchanges (DEX) a fixed percentage of TON rewards and let them compete to build deeper liquidity.
The design will create a "game" for Photon, and other DeFi protocols or DAOs that help Photon the most will receive the most TON tokens. You can also require the DAO to hold TON tokens for a period of time and design rules for unlocking TON rewards.
Yield mining has historically been predatory in crypto history, with most miners giving up rewards. The “Liquidity Mining Partner” concept provides a sustainable way for protocols to work together.
3. Multi-protocol token voting
This concept is related to the idea of governance mining. Governance mining aims to bring different DeFi DAOs together to work together to provide a better user experience for all DeFi users. DeFi protocols are interconnected and often have a lot of crossover. These intersections manifest in many areas, such as governance recommendations and smart contract integration.
So why isn’t there a governance system that can accept voting from two different tokens?
We haven’t seen it happen yet because DAOs are currently designed more like a company, with the interests of a single group determining the future of a single protocol.
However, this multi-token governance will be easy to implement if we create subDAOs in the protocol. For example - suppose a module of Maple Finance is submitted to Photon Finance that lends Maple LP tokens representing USDC to third parties and allows PHO to be minted against it.
Then the parameters of this Maple module can be voted on by TON and MPL holders. MPL holders know more about the Maple pool than TON holders, so the voting weight can be slightly higher. The two communities can come together and vote for the best outcome for both parties.
This concept also creates a very interesting legal issue. I believe it enables tokens to function less like securities. If you could build a multi-token voting system across multiple DeFi protocols, tokens like TON and MPL could potentially be used for the governance of over 10 DeFi protocols. Suddenly, holding governance tokens has nothing to do with a single protocol built by one company. Governance tokens now have voting rights to participate in the entire Ethereum DeFi ecosystem.
This is the future of DeFi that I most hope to see, and it may even lead to the emergence of governance standards for multi-protocol voting.
4. Dynamic multi-protocol token voting
Another problem with traditional token voting is that protocols are often controlled by a small number of people, usually large investors (whales) or founders. You can see this problem in almost every DeFi project. This is especially problematic in the early stages of a protocol when governance tokens are in low circulation. It may have 10% of the circulating supply, with all the tokens owned by insiders. In the eyes of regulators – this looks like 100% centralized, even if the protocol’s long-term plan is to transition to decentralization.
In the previous section, multi-token voting was described as involving modules of 2 protocols that can come together and vote on parameters. But what stops us from allowing protocols to be governed by multiple tokens? Let’s look at the diagram below to see how the Photon example works:
Explain this diagram:
TON starts with 6% of circulation but only 35% of voting power.
Other voting rights have been split into:
CRV = 30% — Since Curve is the largest stablecoin DEX, it makes sense to reward the community by voting for Photon.
UNI = 20% – Uniswap has a TON/PHO mining pool, and Uniswap has a larger token distribution, which makes UNI a good candidate.
AAVE = 10% – AAVE tokens are well distributed and using PHO as collateral in Aave would be a good strategy.
MPL = 5% — If you create the Maple module described earlier, it's worth including the Maple community. However, it is smaller than the previous three protocols and is prone to accepting PHO as collateral in the short term.
As more TON is issued on a linear release schedule, TON voting power increases linearly, while the voting power of the other four decreases.
Likewise, this concept may make these tokens function less like securities, as they now have utility across all DeFi.
However, this concept requires careful consideration. In one instance, when Photon first launched, CRV and UNI token holders combined had more voting power than TON holders. Clearly there has to be some kind of plan here to ensure these other communities don't collude to do evil through Photon.
5. Governance based on usage categories
Changing the traditional structure of DAO, you must be a liquidity provider to manage the LP pool, you must be a trader (during the active period) to receive trader rebates, and you must be a developer to manage the code to receive developer rewards (during the active period)
Osmosis already does this by allowing LP tokens to have governance rights. Osmosis LPs can vote on parameters of a DEX-specific LP pool, such as transaction fees or AMM curves.
Photon Finance could consider giving PHO stablecoin holders voting rights, similar to how DAI holders have voting rights in the Maker protocol.
6. Voting based on NFT
NFT-based voting isn’t new, but it hasn’t really caught on. There are many opportunities for NFT voting, especially when combined with other types of voting.
Let’s look at an example – a DEX can allocate different tiers of NFTs based on the total transaction fees each user accrues over time. What will this ultimately look like?
It will allow real traders of the protocol to participate in governance.
If combined with users’ governance token balances, it can provide more votes to users who use the protocol and hold governance tokens. This will be more advanced than veCRV as veCRV is independent of the volume of transactions made by the user.
Sybil attacks are difficult because accumulated transaction fees cost users.
If using Soulbound NFTs, voting rights cannot be sold on the secondary market.
With this concept, users who regularly use the protocol will be rewarded, as well as those who have skins in the game.
But it's a tough balance. If the rewards are too hard to earn, people won’t care. If rewards are too easy to earn, they become meaningless. One solution is to reward NFTs, similar to gaming seasons. The protocol can host 6-month trading tournaments, with the most active users climbing the ladder to have the most voting power. After the event, the rankings will be reset and a new tournament will begin.
This will allow the protocol to adapt to the fast-paced, ever-changing crypto world. It should provide some flexibility to the protocol’s voting mechanism without compromising the core voting mechanism.
7. Migration and upgrade
Upgrading via migration is not a new concept, but it's worth mentioning. Upgrading contracts on the chain through proxy contracts and governance voting is cumbersome and risky. Compound lost $80 million in $COMP when one of its upgrades had unintended consequences.
Where possible, it is safer to migrate to a new version of a durable, non-upgradeable smart contract and should be used as often as possible. The best part is that no vote is required. The new version can be deployed on the chain, and users can upgrade through "liquidity voting". This is how Uniswap upgrades from V 1 → V2 → V3. It allows funds to be migrated slowly, which also prevents the risk of errors as they can be caught early before everyone migrates.
8. Permission-free triggers
DeFi protocols need to constantly make efficient, secure, and up-to-date decisions. Some examples:
What oracle should it pull data from?
Which DEXs should liquidity be focused on?
What cross-chain bridge should be used to bridge other chains?
When should the protocol open new liquidity mining rewards?
These proposals must be initiated by someone. Typically, it is initiated by the development team via multi-signature. This is a centralizing factor and is not good for regulators. This is why protocols should consider using hard-coded triggers to automate protocol changes without human intervention.
Let's look at some examples:
1. Centralized DEX liquidity dividend
Protocols should strive to centralize DEX liquidity into a single, reliable DEX protocol. Centralized exchanges (CEX) like Binance almost always have the deepest liquidity, which is a huge disservice to DeFi. This allows Binance to grow in a dominant position and earn a lot of revenue from it. As we have seen time and time again, centralized exchanges cannot be trusted.
The protocol attempts to centralize DEX liquidity through liquidity mining incentives, but this is not enough. Protocols should strive to have more than 70% liquidity on a single DEX.
Two protocols that have achieved significant success here are Osmosis and Olympus.
OSMO holds over 80% of the trading volume on Osmosis. Osmosis provides high incentives for staking and providing liquidity for OSMO, so there is no point in storing your tokens on a CEX.
100% of OHM's trading volume is conducted on DEX due to the nature of its rebase mechanism and strong vision for decentralization.
While both won this outcome with high inflation, there is a more sustainable route - lower transaction fees when DEX liquidity is greater than 70%. This can be hardcoded into the protocol:
When DEX liquidity is lower than 70% → 0.30% handling fee.
When DEX liquidity is greater than 70% → 0.05% handling fee.
When DEX liquidity is greater than 90% → 0.01% handling fee.
This is a positive incentive that should drive the community to centralize liquidity on one DEX to improve decentralized trading. Liquidity fragmentation is a big problem, especially when it is split between CEX and DEX. Attracting more liquidity will also increase trading volume, which should keep liquidity providers happy.
In the case of OlympusDAO, the protocol holds over 95% of all DEX liquidity. This is a genius move that allows them to decide the parameters of the DEX on their own terms while still earning almost all of OHM’s trading fees.
2. Upgrade the oracle machine
Protocols often have difficulty deciding how to upgrade oracles. One problem with decentralized oracles is that low liquidity can quickly change prices.
If a protocol has a high enough liquidity threshold, it can automatically use DEX oracles. If it falls below that level, it will fall back to off-chain oracles, such as Chainlink, which are less susceptible to on-chain manipulation. Or it could leverage an external decentralized oracle, such as the UMA Optimistic Oracle, which can report prices and curtail bad behavior through a dispute resolution process.
3. Transfer funds from the treasury and trigger bonuses
Protocol incentives typically require multiple signatures from the team to be transferred from the treasury. This creates centralization risks that regulators can point to. Therefore, the protocol can hardcode the release date of the token to a specific date, eliminating the need for manual management and maintenance.
Taking it a step further, you can also trigger token unlocks based on performance criteria. For example - a stablecoin protocol could hardcode a condition that it must reach a total supply of $100,000,000 by a certain date for founders and investors to receive bonuses.
Summarize
These ideas only scratch the surface of what can be built with decentralized governance. We are testing these governance mechanisms by coding and iterating on them to see which ones work best. In the meantime, we will refine our ideas around the four principles of governance design we proposed in the introduction. We are working hard to create a governance system that realizes the vision of a “decentralized autonomous organization.”
The future is bright.