How Token Concentration and Voting Costs Impact Compound Governance | HackerNoon

Table of Links

Abstract/Zusammenfassung

Publications

Acknowledgements

CHAPTER 1: INTRODUCTION

1. Introduction

   1.1 Overview of thesis contributions

   1.2 Thesis outline

CHAPTER 2: BACKGROUND

2.1 Blockchains & smart contracts

2.2 Transaction prioritization norms

2.3 Transaction prioritization and contention transparency

2.4 Decentralized governance

2.5 Blockchain Scalability with Layer 2.0 Solutions

CHAPTER 3. TRANSACTION PRIORITIZATION NORMS

3. Transaction Prioritization Norms

   3.1 Methodology

   3.2 Analyzing norm adherence

   3.3 Investigating norm violations

   3.4 Dark-fee transactions

   3.5 Concluding remarks

CHAPTER 4. TRANSACTION PRIORITIZATION AND CONTENTION TRANSPARENCY

4. Transaction Prioritization and Contention Transparency

   4.1 Methodology

   4.2 On contention transparency

   4.3 On prioritization transparency

   4.4 Concluding remarks

CHAPTER 5. DECENTRALIZED GOVERNANCE

5. Decentralized Governance

   5.1 Methodology

   5.2 Attacks on governance

   5.3 Compound’s governance

   5.4 Concluding remarks

CHAPTER 6. RELATED WORK

6.1 Transaction prioritization norms

6.2 Transaction prioritization and contention transparency

6.3 Decentralized governance

CHAPTER 7. DISCUSSION, LIMITATIONS & FUTURE WORK

7.1 Transaction ordering

7.2 Transaction transparency

7.3 Voting power distribution to amend smart contracts

Conclusion

Appendices

APPENDIX A: Additional Analysis of Transactions Prioritization Norms

APPENDIX B: Additional analysis of transactions prioritization and contention transparency

APPENDIX C: Additional Analysis of Distribution of Voting Power

Bibliography

5 Decentralized Governance

In this chapter, we present our research questions, methodology, and discuss the implications of our findings regarding the level of decentralization in governance protocols. To investigate this, we focus on the Compound governance protocol as a case study. Our analyses reveal that the distribution of voting power in Compound is highly concentrated among a small number of participants, which can significantly hinder the achievement of a fair and decentralized governance system.

The concentration of voting power poses a challenge to achieving truly decentralization in governance protocols. For example, when a small group of participants holds a majority of the tokens, they can make decisions that benefit themselves at the expense of others. Therefore, ensuring a fair distribution of tokens becomes crucial to foster decentralization in these protocols. In this chapter, we aim to analyze transaction data associated with Compound in order to assess the level of decentralization in Compound’s voting power. To guide our analysis, we propose the following research questions.

▶ RQ 1: How frequently are amendments proposed and voted on in the Compound protocol? This research question aims to investigate the activity level of the Compound protocol and its community engagement. For instance, by examining the frequency with which proposals are amended or voted, we can assess the level of participation and the extent to which the community is actively contributing to improving the protocol.

▶ RQ 2: What is the distribution of Compound tokens among its participants? How small or large is the set of voters who determine the outcomes for the amendments? This research question aims to investigate the distribution of Compound tokens among its participants. Hence, we can assess to which extent the Compound tokens is truly decentralized. Understanding this distribution is crucial for proposing fairness to the protocol token’s distribution.

▶ RQ 3: What is the cost associated with casting a vote in the Compound protocol? Voting in on-chain governance protocols, where the entire voting process happens on the blockchain, requires the payment of transaction fees that vary depending on the network congestion. These voting costs can disproportionately affect small token holders, potentially limiting their participation in the decision-making process. This research question aims to investigate the impact of voting costs on voter participation in the Compound protocol. It provides insights into the fairness of the decision-making process and shed light on potential barriers that may discourage certain participants from exercising their voting rights.

▶ RQ 4: What are the voting patterns of delegates, and do voters form coalitions? This research question aims to analyze the voting patterns of delegates in the Compound protocol and investigate whether voters form coalitions, where they align their votes as a collective group. The formation of coalitions among voters can lead to the marginalization of certain voters, as they consistently find themselves in a minority group. This undermines the core principle of decentralization and has the potential to compromise the security and effectiveness of the governance protocol. Specifically, instead of expressing their individual opinions on a proposal, voters may choose to mimic the voting behavior of their peers. Therefore, exploring the presence and impact of coalition formation can provide valuable insights into the decision-making dynamics among voters and help mitigate the concentration of tokens (or voting power) within the system.

Addressing these research questions is key for improving the protocol’s fairness and achieving more decentralization in the distribution of voting power. In the following section, we discuss our methodology for gathering the necessary data related to Compound protocol from our Ethereum archive node.

Relevant publication

The results presented in this chapter have been submitted, and we are currently awaiting a decision (Messias et al., 2023b).

5.1 Methodology

To analyze the voting power concentration among Compound token holders (or voters), we adopt a data-driven approach. Our methodology involves collecting events triggered by transaction executions when voters cast votes, create proposals, cancel proposals, transfer tokens, or delegate their voting rights to another address. We cover events from the inception of the Compound token and Compound governance protocol. To address the possibility of a single entity owning multiple addresses, we have developed a methodology to infer address ownership and group them accordingly. This approach allows us to identify and consolidate addresses that are likely controlled by the same

entity. This process utilizes data from well-known blockchain explorers and publicly disclosed information regarding address ownership. For those addresses that we were not able to infer their ownership we renamed them to their specific wallet addresses.

To gather the data, we deployed an Ethereum archive node on a server with 64 cores (with a base clock frequency of 2.25 GHz that can be boosted to 3.4 GHz), 256 MB L3 cache, 252 GB of RAM, and 21 TB of NVMe-based storage. The archive node took about 4 weeks—a relatively long time, though not unexpected—to fully synchronize with the Ethereum blockchain. We used Web3.py (web3.py team, 2022), a Python library for interacting with Ethereum nodes, to query and retrieve the information that we need from the archive node. Figure 5.1 summarizes our methodology for the Ethereum data gathering.

5.1.1 Smart contract events.

Smart contracts in Ethereum can generate and dispatch events for signaling various types of activities (e.g., ERC-20 token transfers or state changes) within the contract. We can subscribe to these events, or analyze them later since Ethereum persists the events in the blockchain via the “logs” field of the transaction receipt attribute. In this thesis, we leveraged these logs to filter transactions that triggered specific events, e.g., sending, receiving, or swapping tokens. We also filtered and analyzed transactions that triggered events related to governance protocols to track the evolution of each proposal, including when it was created, when users started voting, and when it was executed or canceled.

5.1.2 Data set collection

We gathered various details on Compound tokens and Compound governance contracts between March 3, 2020 (block #9,600,000) and November 7, 2022 (block #15,917,000) from our Ethereum archive node. This 32-month study period includes Compound’s entire lifetime (from its inception). We illustrate our methodology and data-analysis pipeline

in Figure 5.1. We obtained 213,220 Approval events, 12,095 DelegateChanged events, 75,820 DelegateVotesChanged events, and 1,886,618 Transfer events for Compound tokens (refer to Table 5.1). We also collected various events (refer to Table 5.2) related to the Compound Governance contract for analyzing various aspects of the proposal creation and voting processes.

5.1.3 Inferring wallet address ownership.

Since an entity can control multiple wallet addresses in the blockchain, identifying the ownership of these wallets helps in grouping together the accounts that are owned by the same entity. However, this task of wallet-address ownership determination is challenging due to the inherent anonymity of blockchains (Antonopoulos, 2014; Antonopoulos and Wood, 2018). This task is further complicated because owners are only identifiable if they choose to voluntarily make their identities public. To address this challenge, we combine wallet ownership information from two widely used data sources: Etherscan (Etherscan, 2023b) and Sybil-List (Sybil, 2023b). The former is a blockchain explorer that helps in identifying the top holders of various cryptocurrencies, and the latter, a Uniswap governance tool for discovering delegates addresses (Sybil, 2023a). It uses cryptographic proofs for verifying wallet addresses voluntarily disclosed by the wallet owner. From these two data sources, we gathered the owners of 3191 public wallet addresses. We used these addresses to infer the owners of 17 (51.52%) of the 33 unique addresses associated with proposal creation, 114 (3.42%) out of 3335 proposal voters, and 265 (0.13%) out of 210,598 token holders. By analyzing the top 10 most influential voters for each proposal, determined by the number of delegated tokens they possessed when casting their vote, we were able to infer the ownership of 67 (50.37%) of these 133 unique addresses. Finally, as an entity can control more than one address, we grouped the addresses we identified belonging to the same entity together to conduct our analysis.