A couple weeks ago, a16z crypto announced their investment in MakerDAO, the creator of the Dai stablecoin1. For $15M, a16z purchased 6% of MKR tokens at a discount to the spot price (around 45%). Prior MKR holders were not involved in the decision to make this deal.
This deal is notable for many reasons, and Yannick covers them well in his analysis of the deal.
One particular thing stuck out to me: the negative reactions some people had to the nature of the deal. They felt that MakerDAO:
should not “sell out” to Silicon Valley
should not give “unfair” discounts to prestigious investors
should have consulted the community before making the deal
In making this deal, the MakerDAO team did not violate any explicitly stated or implied rules. Except, perhaps, in calling their organization a DAO (decentralized autonomous organization).
MakerDAO is not obligated to consult with MKR holders and can sell their team tokens at whatever price they wish to whomever they’d like. Their explicitly defined governance mechanisms only cover risk management. Yet there remained an expectation that all stakeholders would be consulted.
Set aside the problems with the name “MakerDAO” (inaccurate to describe the system today, but useful to describe the project’s ideals) and we can see the root of the problem: tokens are not equity2, but people expect tokens to behave like equity.
Tokens vs equity
What you get with equity
Ownership: A pro-rata claim on the residual value of the company
Control: Governance rights
These things are legally enforced. The average holder of equity need not worry about unexpected changes to their ownership or control.
Most cryptoassets behave like psuedo-equity
Ownership: Pro-rata ownership of the current supply3 of a cryptoasset
Control: An informal promise that token owners will be informed and consulted in governance decisions
These equity-like expectations are not legally enforced. Sometimes, they are enforced “on-chain” either in the base protocol design or through smart contracts. For example, monetary supply that nobody can change or an on-chain governance system.
But by and large, ownership and control are provided in “good faith” by those with power in these networks (if at all). The powerful meet the expectations of the network’s stakeholders in self-interest. If they were to fail expectations, the value of the network would decrease.
Psuedo-equity can create conflicts of interest
Problems with conflating tokens and equity emerge in situations like this a16z MakerDAO deal. There are many reasons why token holders in the MakerDAO ecosystem were not consulted. For instance, making the deal in public may have jeopardized the success of the deal. The decision may be in the best interest of all token holders, yet there remains a feeling that they were stripped of their control.
Bigger problems arise when there are conflicts of interest.
Here is a toy example to illustrate the point:
Investors fund a team by buying tokens, expecting value to be captured by the token
The tokens never capture value, but the team is able to create a separate, valuable business
Because the success of the separate business does not flow into the token, the investors do not benefit
Or how about this one: investors buy tokens from a team and the team decides to not commit any meaningful work to the project and instead just use the funds for whatever they feel like.
While this might seem like an extreme example, the concept of potential conflicts of interest are unavoidable when teams have both equity and token components. What’s best for the value of equity won’t always be what’s best for the token.
It doesn’t take an active imagination to come up with many other possible conflicts of interest, which is a bit ironic because a popular promise of public blockchain-based systems is to align the incentives in a network.
Aligning the incentives of owners is part of why equity (and the reams and reams of laws surrounding equity) “works.” It works so well that we complain that internet giants align with shareholders too much and users too little–sometimes trading user privacy or wellness for higher advertising returns.
For internet oligopolists, the conflict of interest is between the owners and its users. Crypto networks are supposed to turn users into owners, erasing that conflict. What we’re seeing with the conflation of equity and tokens is that the conflict simply moves positions:
In equity, a conflict of interest can develop between owners and users
In tokens, a conflict of interest can develop between owners and less powerful owners
Disambiguating tokens and equity
You might protest that Bitcoin does not have this conflict of interest. That is mostly true, and leads me to my final point.
Like so many things in crypto, a half-measure doesn’t work well:
Kind of private can mean not private
Kind of censorship resistant can mean not censorship resistant
Kind of governed by it’s stakeholders can mean not governed by its stakeholders
When you kind of have ownership and control, you don’t actually have it. Conflicts of interest are avoided when ownership and control are clearly defined:
Security token: when you issue tokenized equity (ownership and control just like equity)
Leaderless network: when there aren’t distinct classes of owners, so you are unable to create conflicts of interest amongst owners (ownership and control not like equity unless explicitly designed like equity)
Owners don’t conflate tokens with equity when tokens behave literally like equity or nothing like equity. The awkward middle4 where power pools to a “ruling class” creates systems highly vulnerable to conflicts of interest. This describes the majority of token projects.
To be clear, the problem is not that tokens are not equity, it’s that owners conflate tokens with equity. The risks of conflation are greatest in networks with multiple classes of owners because they are more vulnerable to conflicts of interest. These networks may either evolve into security tokens or leaderless networks, reducing risk of conflicts of interest. But until then, it’s important to disambiguate tokens from equity.
Treat the properties of tokens literally. If expected behaviors are not in a contract (smart or dumb), owners must rely on trust in others. And that defeats the purpose.
Special thanks to Alex Hardy for his feedback after I published this post. I’ve since gone in and made some corrections.
Important because the supply is more easily changed in tokens than in equity. ↩
It’s possible that for some future leaderless network, the path must pass through the awkward middle. As Rune (a founder of MakerDAO) commented following the announcement of the deal, “It is simply not possible for a decentralized community to bootstrap a fintech product from scratch with the flexibility required to compete in the marketplace.” If the networks of the future must pass through periods of awkward ownership and control, we cannot simply avoid the problems of pseudo-equity by designing only security tokens and leaderless networks. ↩