Minimal Viable Futarchy DAO

This is a proposal for the minimal viable Futarchy DAO.
Futarchy is a governance form in which all decisions are made to optimize an agreed-upon success metric. In the case of the MVFD (minimal viable Futarchy DAO) this would be it’s own token price.

The concept would be:

1: Initialization period
Anyone can contribute ETH (or a token) and get pro-rata DAO tokens

This is only possible for a limited time. After that time the DAO switches into governance mode. Now new tokens can only be minted through governance

After the Initialization period, x% (10%) of the funds are out into a uniswap like market maker together with newly minted DAO tokens. The LP tokens are held by the DOA. This is done to have a constant price feed that is relevant for decision-making.

2: Governance period.
a) Proposal
Anyone can make any time a proposal (tx payload)
This will cause the DAO to redeem all the LP tokens for WETH and DAO tokens and split the WETH and DAO tokens into conditional token WETH - YES; WETH -NO and DAO - Yes; DAO - NO) It will not create the 2 markets: WETH - YES / DAO - YES and (WETH - NO / DAO -NO)

For 24h trading is allowed on those 2 markets. If the DAO token price is higher on the “yes” market (meaning more people are buying into the DAO if the proposal gets executed) the proposal is tentatively accepted. For security reasons another period will follow where everyone can sell DAO tokens at the “yes price”. This period will continue until no further seller is found (e.g. 24h extension if another 1% of all DAO tokens are sold). If all people that wanted to sell could sell at at least the “yes price” the proposal will be executed. If sellers could not sell at the higher price the proposal is not accepted.

Spam protection:
The mechanism above allows always to leave the DAO in case a malicious proposal comes in. Given that a proposal is only accepted if it will make the price increase an attacker would need to pay out all people that are leaving this higher price.
However - a malicious will cost peoples time (and gas costs) to signal that they would sell if that proposal should pass. For this reason, the DAO should have a spam protection mechanism. This could come in the form of a mandatory market order to buy “yes-DAO” tokens. This would mean that someone is willing to invest in the DAO if the proposal passes and they would actually lose that investment if the proposal would not pass. It would be investing in the DAO and betting on the proposal at the same time.

For simplicity, only one proposal should be pending at a time. If multiple proposals are seeking to get into the decision stage the system will prioritize the one that does the largest investment into the proposal.

Attack scenarios:
An attacker can always pass a malicious proposal that gives them all the assets the DAO holds.
Let’s assume the DAO holds $1m in assets and there are 1m DAO tokens. Now each DAO token should be at least worth $1. An attacker could start buying “yes” DAO tokens for $1.02
If all DAO token holders sell the attacker has bought $1,000,000 for 1,020,000 uns thus has gained nothing. However - if only 2% of all token holders do not sell the attacker will make a profit.

Factors that mitigate the attack:
a) DAO tokens might trade at a premium above the nominal value. If e.g. the token trades at $1.20 but the nominal value is only $1.00 the % of people ignoring the attack can be much larger and the attack would still not be financially successful for the attacker (token holders that would not sell still would lose money though)
b) It is unknown to the attacker up front how many people will sell at the “yes” price. Since the period will be continuously extended if people continuously sell the attacker needs to always assume that all people might sell eventually. If they don’t buy to the end the attack will fail.
c) If the attack (the proposal) does not succeed the attacker will lose the invested amount that is spent on the proposal succeeding.

A variation:

The investment from the proposers could at the same time be used as a price oracle for the “DAO-YES” token. A way to do it could be to place a “buy market order” on Gnosis Protocol. To get an average price over time the market order could be split up into 10 chunks that invest at 10 different price points a tenth of the money.

I think there is a griefing attack that is enabled by this mechanic, where an attacker could sell a fraction of their shares every 24 hours and push the decision out indefinitely.

A Moloch-style grace period might be more suitable (and more MVP).

So to continue this period 1% of the total amount amount (total DAO) needs to sell - so at max this could continue for 100 days. The idea is to give people the opportunity to exit if a malicious proposal should pass. If people keep selling it might make sense to let it running for a long time to let as many as possible people out.

Also, only DAO token holders can do this “attack” and as token holders they should not have an incentive to do this.

Ah I see. Yeah I missed the 1% piece.
That certainly limits the potential for griefing.