From Corporations to Nations: How the Meta-DAO is Going to Change Everything (Part 3)
Introduction
Existing human organizations have human rulers. Whether the organizations be corporations or national governments and whether the rulers be CEOs, board members, prime ministers, congressmen, or ayatollahs, the structure the same: humans appointed to serve the interests of the group.
These organizations suffer from the principal-agent problem. Simply put, these rulers often take self-serving actions that hurt the group because it is in their interest to to so. Examples of this include board members being too risk-averse in order to avoid lawsuits, politicians appointing loyal but unqualified cronies, and regulators defending rent-seeking incumbents to maintain good job prospects.
We propose an alternative organization where control is handed to executable computer code, namely eBPF instructions stored on the Solana blockchain. Decisions would be made automatically by an algorithm that tries to maximize the wealth of the humans involved in the organization. We will describe this algorithm in this post.
Members
We define members as single-product profit-seeking entities. Members are analogous to Helium’s sub-DAOs. Each member has its own treasury, from which it can pay grants, payroll, and suppliers. Each member also has its own token, which functions like common equity in a company.
We refer to a collection of members as a Meta-DAO. For instance, there could be a DeFi-focused Meta-DAO with two members: one that manages an AMM product and one that manages a yield vault product.
Improvement proposals
Members take actions via improvement proposals. Each improvement proposal contains a list of commands, where a command contains a member and a Solana instruction that the member can execute. If a proposal passes, instructions are executed by their corresponding members.
Making decisions
To decide whether or not to execute an improvement proposal, a Meta-DAO takes the following steps:
- Estimate, in some base currency, the impact of the proposal on each member’s valuation.
- Sum all the estimated impacts.
- Execute the proposal if and only if the resulting sum is positive.
For example, suppose that a proposal for member A to raise its prices is active. This proposal could benefit A by increasing A’s profitability. On the other hand, if member B shares a brand with A and member B depends on a low-cost brand image to market its product, the proposal could hurt B.
Suppose that the Meta-DAO calculates that the proposal will add $10M to A’s valuation and diminish B’s valuation by $15M. Since 10M + -15M = -5M, a negative number, the Meta-DAO would not execute this proposal.
This algorithm is ideal, by our previous definition of ideal, because it only takes an action if it grows the total wealth of the system.
Estimating impact
We use a system based on Robin Hanson’s futarchy to estimate the impact of a proposal on a member’s valuation.
While a proposal is active, investors can lock arbitrary tokens into a vault in exchange for two convertible instruments: one that converts back to the token if the proposal passes and one that converts if it fails. We call the former a convertible-on-pass token and the latter a convertible-on-fail token.
Investors may trade these convertible instruments freely to speculate on how a proposal will affect a member. For example, if an investor believes that member token ABC would be worth 10 SOL if a proposal were to pass but the market price of convertible-on-pass ABC is 9 convertible-on-pass SOL, he can realize a 1 SOL gain by doing the following:
- Lock up 9 SOL
- Swap the 9 resultant convertible-on-pass SOL for 1 convertible-on-pass ABC
- If the proposal passes, claim the 1 ABC and sell it for 10 SOL once the market realizes that the investor is right
- If the proposal fails, use the 9 convertible-on-fail SOL to get back the original 9 SOL
In other words, the proposal passing locks in the investor’s buy, but a proposal failing would simply revert his trade and he would get back his original inputs.
We can use the market prices of these convertible instruments to estimate the impact of a proposal on a member’s valuation. A pass price, or the market’s estimate of what a member’s token would be worth if the proposal were to pass, is simply the price of convertible-on-pass member tokens, quoted in convertible-on-pass base tokens. Likewise, the fail price, the market’s estimate of what the token would be worth if the proposal failed, is just the price of convertible-on-fail member tokens, quoted in convertible-on-fail base tokens.
Once we have the pass price and the fail price, calculating the market’s estimate of the proposal’s impact on this member’s valuation is a simple matter of subtracting the fail price from the pass price and then multiplying by the token’s total supply.
Concerns
Token prices aren’t a good reflection of underlying valuation
In a world where cryptocurrencies created as a joke can fetch $70B valuations and a subreddit can run up a failing retailer to $34B, it’s reasonable to question the efficacy of financial markets. Do market prices really reflect intrinsic value?
Our answer to this question is no, they do not. However, to quote Robin Hanson:
The key issue is not absolute accuracy, but accuracy relative to other institutions, on the same topics, given similar resources. We have some data on this. In addition to lab studies, a few studies directly compare real speculative markets with other real info institutions. For example, racetrack market odds improve on the prediction of racetrack experts; orange juice commodity futures improve on government weather forecasts; stocks fingered the guilty firm in the Challenger crash long before the official NASA panel; Oscar markets beat columnist forecasts; gas demand markets beat gas demand experts; betting markets beat Hewlett Packard official printer sale forecasts; and betting markets beat Eli Lily official drug trial forecasts.
Markets are not perfect, but the evidence shows that they are superior to the alternatives and hard to manipulate. The reason why those examples come to mind easily is because they are sensational - out of the ordinary. If you look at most publicly-traded companies, they tend to trade based on their fundamental metrics. See, for example, the correlation between SaaS companies’ revenue multiples and their growth rate. There are outliers, of course, but they tend to be outliers for a reason (higher profitability, bigger moat, a potentially-lucrative project in the works, et cetera).
In the early stages, a Meta-DAO may have limited interest and low liquidity in the pass-fail token markets. What’s the bootstrap mechanism/market structure for these pass-on-success/fail markets?
This is still being worked upon, but some ideas are:
- Traditional ‘liquidity mining’ / ‘network mining’ incentives, where members allocate a large portion of their token supply to compensate early market participants
- Other social benefits, such as NFTs, to those early adopters
- Some form of human oversight in the early stages, such as a multisig that can veto improvement proposals.
If you have ideas here, please don’t hesitate to reach out.
Conclusion
We have demonstrated a design for a human organization where the managerial decisions are made by eBPF instructions on the Solana blockchain. Because this system is not human-led, it doesn’t suffer many of the problems of prior approaches to human organization.
