What we learnt from the rise and fall of the DigixDAO autonomous organisation
DigixDAO voting itself out of existence is a good example of a DAO working as it should.
Governance remains one of the tougher problems in blockchain. It's essentially the question of who's in charge of a decentralised public network and how to imbue a network with a practical sense of direction and keep development going, without sacrificing decentralisation and handing any one entity the keys to the castle.
Decentralised autonomous organisations (DAOs) are one of the answers to the problem. Far from being monolithic organisations, as the name would suggest, DAOs are essentially self-governing communities. But boiled down to their most basic parts, decentralised autonomous organisations (and many other governments) are essentially a set of rules and a series of votes.
Now one of the world's few live DAOs, DigixDAO, has decided to hit the self-destruct button by passing a vote in favour of Project Ragnarok. Somewhat paradoxically, this is a great example of a DAO working well.
For any of this to make sense, you need to have an understanding of the DigixDAO tree of life and the different components involved in it. The following are the main ones:
- Digix - a company
- DigixDAO - a decentralised autonomous organisation
- DGX - a gold-pegged stablecoin
- DGD - a utility cryptocurrency
Digix is the name of the company that created the DigixDAO and the cryptocurrencies within it. Digix is a for-profit business and a separate entity to the DAO. As an entity, Digix has different motivations than the DigixDAO.
The DigixDAO ecosystem has two tokens. One is the DGX stablecoin, whose value is pegged to gold. Each DGX token is worth one gram of gold. Then there's the DGD token, which is essentially a symbol of "citizenship" in the DigixDAO ecosystem. DGD ownership gives people the right to vote on proposals in the DigixDAO ecosystem and earn rewards for active participation.
The goal of DigixDAO is (was) to foster a community around the DGX stablecoin and to promote its growth and use. To this end, DigixDAO was put in charge of managing a treasury and allocating funding for projects it thinks will help achieve these goals.
Digix created the DigixDAO in 2016, by simultaneously bringing it into existence, giving it members and giving it a treasury to manage. It did this through an initial coin offering (ICO) where people exchanged Ether (ETH) for DGD tokens.
The Ether was locked away into the DigixDAO treasury to fund projects as approved by the community of DGD token-holders. Whether those ICO participants were investors or donors is something of a grey area.
The project raised 450,000 ETH, worth some US$7 million at the time. A bunch was spent over time, but there's still about 380,000 ETH, worth $65 million today, left in the DigixDAO treasury. If you're curious, the two treasury addresses are here and here.
So, when designing its DAO, Digix essentially needed to program an entire government. It had to create a system for the community to moderate itself, prevent abuse and decide on funding for good projects while sieving out the scammers and junk, all in a way that doesn't contain any obvious exploits or downsides. As you can imagine, this is extraordinarily complex and difficult.
There are three main mechanisms controlling people's behaviour in this system and keeping it all from spiralling out of control:
- Moderators. Individuals who hold a certain amount of DGD tokens automatically become moderators. The idea is that people with a larger stake in the success of the DigixDAO ecosystem can be trusted with additional responsibility.
- Rewards. Moderators and other active participants are entitled to DGX rewards, taken from DGX transaction fees and pulled from the plumbing of other network money sinks.
- Underlying incentives. Whether you're requesting funding, voting or moderating, DGD holdings are required for all participants. The assumption is that this aligns everyone's incentives.
As complex as this kind of social engineering is, the end result is rather simple looking. People who want funding put their project through a series of community votes. With enough yes votes through all stages, the proposers are entitled to claim the agreed amount of ETH from the treasury.
And in a way, the DGD tokens are collateralised by the ETH in the DigixDAO treasury, not unlike the way DGX tokens are backed by gold.
The self-destruct button, Project Ragnarok, was introduced by the community for voting just like any other proposal. The main point of it was to create a way for DGD holders to (re)claim their share of the treasury.
The Ragnarok contract itself is relatively straightforward. It's basically a contract that lets people destroy DGD in exchange for a share of ETH in the treasury at a proportional rate. The formula for this exchange rate is basically the total amount of available ETH in the treasury divided by the total number of DGD tokens. It comes out to 0.19 ETH per DGD token.
It may also be a potential arbitrage opportunity for anyone who's fast enough. The value of DGD can be perfectly measured at 0.19 ETH, making it possible to see when it's objectively over or undervalued.
The idea was for Project Ragnarok to recur each quarter, providing periodic opportunities for community members to pull the plug. It didn't come to that though, as the apocalypse passed with overwhelming approval the first time it was put to vote.
The final count was about 670,000 DGD (over $20 million) for yes and 20,000 DGD (about $620,000) for no. Digix politely abstained from voting, but it was quite vocally opposed to the dissolution.
To be clear, Digix and DGX will continue operating as usual. It's just DigixDAO and DGD that will cease to exist.
All in all, it's an excellent example of a DAO running its course as nature intended without any major disasters, so feel free to mentally clap DigixDAO off the stage. It also provides an invaluable set of lessons for anyone trying to build a DAO in the future.
The most obvious lesson to be learned is that DAOs are really, really hard to get right.
Even in this case, where the DAO was arguably working as smoothly as intended, some frictions were obvious and it's easy to see why people were so keen to get their money out.
Consider, for example, the proposal from Silver Gold Bull in which it boasted of its "yearly revenues in the hundreds of millions" and then asked for $750,000 in funding for adding DGX to its digital asset exchange and a marketing campaign aimed at its existing clients.
While listing fees are quite normal, $750,000 is pretty outrageous for an exchange that doesn't even exist yet, and it's not entirely clear why the DigixDAO community should be footing the bill. It's the kind of deal best made on purely commercial grounds and with clear contractual obligations for each party.
A much easier and more typical decision for the community is around questions of whether to donate $6,000 of ETH to the person managing the DigixDAO Twitter accounts or whether to award $4,300 to the DigixDAO community manager.
It's easy for the community to vote on simple yes or no decisions like "should we keep doing this?" It would be considerably harder for the community to reach agreement on more complex questions like "how much should social media managers be paid?" or "how do we quantify the benefits of Gold Silver Bull advertising DGX to its existing clients?"
If there's a rule of thumb buried around here, it's probably that DAOs can handle small and simple decisions, but big, complex multi-faceted decisions are beyond them.
And in that case, it's quite clear that DigixDAO really doesn't need to maintain a $65 million treasury. Keeping it in the treasury is simply not a good use of money when it could be earning interest instead. Without actually crunching the numbers, it looks like the potential interest earned on $65 million would have been more than enough to fund every request the DigixDAO ever approved.
And with DigixDAO as an example, it's also clear that there are potential issues with small groups, consolidation of power and economic unsustainability in DAOs.
There are over 11,000 DGD addresses, but the apocalypse was passed with only 58 votes. Despite the rewards for doing so, the vast majority of people still elect not to participate. The same problems are found everywhere people have the vote. Susan B. Anthony would probably be very disappointed.
There's also something mildly disconcerting about this kind of system where voting power is literally based on money and where more money means more votes. It naturally consolidates power, which isn't ideal when you also want to encourage more participation and tap into the power of group decision-making to unlock better decisions. Plus, it puts a system at risk of certain attacks. For example, if a large holder uses their clout to drive funding to projects of their choice or starts selling their votes.
It's also brought interesting observations in the accrual of value in ecosystems and some of the lessons around that. The idea may have been that the value of DGD was initially based on its ETH backing, but as the ETH treasury is spent, the DigixDAO ecosystem grows, which naturally shifts that value towards DGD.
We can see that this never really happened. Like almost every cryptocurrency, DGX and DGD are only seeing a limited amount of real-world use. It's perhaps as a consequence of this that DGD prices have very closely tracked the value of ETH in the treasury.
That's not great because one could argue by the numbers that if DGD's total market cap mirrors the total value of ETH in the DigixDAO treasury, the money spent from the treasury must be getting wasted. It's obviously an oversimplification, but three years may have been long enough for DGD holders to decide the experiment had failed.
There's also the question of value accrual in other moving parts of this system, such as the equity of the Digix company. The sheer novelty of this kind of arrangement, where a business creates a decentralised autonomous organisation to help grow its own product, may pose a few financial headscratchers for investors.
At the end of the day, even though the sun's going down on DigixDAO, the experiment shone a lot of light on DAOs as a whole.
Disclosure: The author holds BNB and BTC at the time of writing.