An Analysis of Voting For Stable Cryptocurrencies

Abstract

We begin by introducing a simplified model for the vote in a stablecoin protocol. Under these simplifying assumptions, we mathematically determine the best votes for a single voter in the absence of rewards. We assume that demand for the coin is tied to the exchange rate and observe that when voters believe that demand will decrease when they vote too dishonestly they have incentive to vote closer to the truth. This creates a scenario where voters must optimize the amount that they can deviate from the true value, taking into account effects on future demand. We then add rewards and penalties to this model under more simplifying assumptions and use a simulation based approached to analyze how one individual will vote under different reward and penalty scenarios. We find that even small rewards and penalties provide a powerful incentive to vote closer to the truth. We also analyze voting over longer periods of time via a simulation and determine that in the presence of rewards, wallet maximizing voters do not destabilize the exchange rate of a stablecoin any more than voters who draw their votes from a normal distribution.