A shower-thought on governance tokens
(Disclaimer: Might be that someone else came up with this before)
A governance-token is a token representing a vote to adjust some parameters in some decentralised system (blockchain, in practice). I.e. you could say you hold one governance token for each country wherein you are eligible to vote.
In the blockchain-context those tokens are (in the ideal case) related or identical to a kind of stake — for example a security-token. Makes sense; the more invested you are, the more you have to say. Also determining personhood (like in a proper democracy) is difficult in itself.
And of course those tokens are tradable. This however poses the following problem: What if someone decides to buy a huge bag of governance tokens right before a vote, then vote, then sell them again?
The naive solution would be to say “well then we require a minimum holding duration or weight the vote by that timeframe!” which is not bad (in fact I just thought of it while typing) but does not align incentives properly. Also one might argue that it is not fair, but this debate will take a while and is luckily irrelevant for my point here. Also no one stops you from combining both approaches.
Anyways.
What does align incentives properly would be the following scheme.
- you can lock your valuable thing (i.e. security-tokens) which determines your voting power for up to a certain amount of time. This can be renewed at any time.
- Your votes at any time are then not only determined by your stake (how much of that valuable thing you have) but how long you are unable to sell it off from now on.
So if you for example lock in your stake for a year, and vote for something that will provide you short-term profits but harm the project, you just sucker-punched yourself. Incentives: aligned.
But wait, there’s more. What if people figured out a way to trade wallets? Then I could just sell a wallet with some locked stake, thus just adding the inconvenience. Recall that determining personhood is difficult, so I can split my funds over many wallets, each of which say “redeemable for 1 epsilon stake in 1 / epsilon amount of time”. Which is the same as 1 epsilon stake.
Of course trading wallets is also difficult but not impossible. So what if we don’t use wallets but some kind of access key with the property that you can’t check if the key works until the locking time elapsed? In other words — what if the error message for “funds are still locked” and “this is the wrong key, mate” are the same? The advantage here would be that it would be easy to scam-sell access keys, thusly crippling their trade as intended.
But. Here again we could contrive some kind of collateral-mechanic which basically says “I, the vendor, lock some collateral in this smart contract which you will be able to access if A) the locking-time has elapsed B) you tried to withdraw the locked stake and C) you failed. Otherwise, I get it back”. This does not sound practical at all but at least it’s something. Surely we can do better, though.
Another idea I had was to incentivise against splitting of funds, to keep them all in one wallet. Besides the issue of creating a central point of failure for the user the only mechanism I could think of without wasting even more water was to scale votes not only proportionally but superlinearly with stake. Which would be not at all fair.
But! And here comes the kicker. What if the locking-time also scales superlinearly? That way you get rewarded superlinearly for the risk you take also, which is worse in absolute terms for people who invested higher sums, thus offsetting their unfair advantage? I admit that last part feels very hand-wavery and should be treated mathematically.
Also I smell like watermelon now. Thanks, chemistry.
