Decentralized Finance and Governance: Designing Automated Market Makers and Voting Systems
OPEN ACCESS
Loading...
Author / Producer
Date
2024
Publication Type
Doctoral Thesis
ETH Bibliography
yes
Citations
Altmetric
OPEN ACCESS
Data
Rights / License
Abstract
This dissertation studies several aspects of decentralized applications deployed on blockchains with the aim of improving their efficiency and economic viability. It focuses on automated market makers (AMMs), voting in decentralized autonomous organizations (DAOs), as well as the more general problem of aggregating issue-wise opinions of individuals.
We begin by examining the economic viability of AMMs from the protocol's perspective: Is it possible for AMMs to sustainably retain a share of their trading fees, otherwise earned by liquidity providers, for the protocol? We approach the problem by modeling how to choose a pool's take rate, i.e., the fraction of fee revenue that remains with the protocol, optimally to maximize the protocol's revenue. Our model suggests that AMMs with a portion of loyal trade volume can sustainably set a non-zero take rate, even without losing liquidity to competitors with a zero take rate. Furthermore, we determine the optimal take rate depending on a number of model parameters, including the pools' portion of loyal trade volume and the competitors' take rates.
Next, we focus on improving AMMs from the perspective of liquidity providers (LPs) by reducing their losses to arbitrageurs. We study a novel AMM design in which all trades are batched and executed at a price equal to the marginal price after the batch trades. We derive the trading function of such an AMM, and show that it is "function maximizing" (hence naming it FM-AMM): For given prices, it trades to reach the highest possible value of a given function. Competition between arbitrageurs guarantees that an FM-AMM always trades at a fair, equilibrium price, and arbitrage profits (also known as loss-versus-rebalancing) are eliminated. This also eliminates sandwich attacks because all trades occur at the uniform batch price. Finally, we show that our results are robust to the case where the batch trades on both the FM-AMM and on a traditional constant function AMM.
Subsequently, we empirically examine LPs' losses to arbitrageurs. In addition to quantifying the potential of AMMs preventing these losses (such as the FM-AMM), the results also evaluate the historical profitability of AMM liquidity. We find that losses to arbitrageurs exceed the fees earned by LPs across many of the largest AMM liquidity pools. We also investigate how arbitrage losses change with block times. As expected, arbitrage losses decrease when block production is faster. However, the rate of the decline varies significantly across different trading pairs, with the reduction for 100ms block times ranging between 20% and 70%.
In the second part of the thesis, we study voting and aggregating opinions. First, the voting process of decentralized autonomous organizations (DAOs) is empirically studied. We analyze the state of three prominent DAO governance systems on the Ethereum blockchain: Compound, Uniswap and ENS. Using a comprehensive dataset of all governance token holders, delegates, proposals and votes, we analyze who holds the voting rights and how they are used to influence governance decisions. While we find voting power in the DAOs to be concentrated in the hands of a few addresses, our analysis shows that these influential stakeholders generally vote in alignment with the broader community.
Finally, a challenge occurring in DAO governance as well as in democratic systems more broadly is discussed: Finding a compromise between the opinions of a group of individuals on a number of mutually independent, binary topics. We quantify the loss in representativeness that results from requiring the outcome to have majority support, in other words, the "price of majority support". Here, individuals are assumed to support an outcome if they agree with the outcome on at least as many topics as they disagree on. Our results also quantify Anscombe's paradox, which states that a topic-wise majority outcome may not be supported by a majority. To measure the representativeness of an outcome, we utilize two metrics. First, the outcome that agrees with a majority on as many topics as possible is considered. Second, we maximize how often a voter's opinion on a topic matches the outcome on that topic. For both metrics, we consider the maximum value for which an outcome of this representativeness is guaranteed to exist, and establish lower and upper bounds on this value. For the number of majority topics, we prove matching lower and upper bounds of (t+1)/2 where t is the total number of topics. For the number of matching voter opinions, we show that an outcome with 16.6% fewer matches, compared to the overall best outcome, can always be found when requiring majority support for three topics. In general, we prove that the "price of majority support" is higher than 10%, and numerically compute a better lower as well as a non-matching upper bound.
Permanent link
Publication status
published
External links
Editor
Contributors
Examiner : Wattenhofer, Roger
Examiner : Knottenbelt, William J.
Examiner : Danos, Vincent
Book title
Journal / series
Volume
Pages / Article No.
Publisher
ETH Zurich
Event
Edition / version
Methods
Software
Geographic location
Date collected
Date created
Subject
Blockchain; Decentralized finance; Automated market maker; market making; Batch trading; Arbitrage; Arbitrage profits; Arbitrage opportunities; MEV; Blockchain governance; DAO Governance; Decentralized autonomous organizations (DAOs); Liquid democracy; Voting rules; Approval voting; Majority support; Anscombes paradox
Organisational unit
03604 - Wattenhofer, Roger / Wattenhofer, Roger