Working Papers

The Economics of DeFi Lending: A Model of Smart Contract Parameter Choice

Abstract

We propose a simple model to analyze the optimal design of a decentralized finance (DeFi) lending protocol. We focus on two key smart contract parameters: the loan-to-value ratio and liquidation threshold. Our model reveals fundamental trade-offs between borrower leverage and lender security, with important implications for protocol governance and design. We show that the optimal liquidation threshold should be set to its maximum value, benefiting both borrowers and liquidity providers. However, this theoretical prediction stands in contrast to observed protocol implementations, where thresholds are typically set lower. The model also demonstrates how governance structure affects parameter choices: borrower-dominated governance leads to higher loan-to-value ratios, while liquidity provider-dominated governance leads to lower ratios. These findings have significant implications for protocol design, particularly in light of the efficiency of modern liquidation mechanisms enabled by flash loans and automated bots. Our analysis provides a framework for understanding how protocol parameters are determined through the interaction of governance structure, market competition, and stakeholder interests.

Security Audits in DeFi

Abstract

Decentralized Finance (DeFi) is an innovation with the potential to address fundamental issues in modern finance. However, in practice, its technology has also demonstrated susceptibility to various problems, including security vulnerabilities and coding errors, which have led to an increasing number of exploits and hacks. In response, protocols have adopted security audits as a means to mitigate these issues.

In this study, we collected original data on security audits and explored their potential signaling effects. Our findings indicate that protocols are audited more frequently than previously documented, including audits when contracts are already managing funds. Generally, the release of security audits results in modest increases in the returns of the protocols’ own currencies. Contrary to findings in other studies, we did not observe significant increases in the amount of liquidity attracted by protocols or in their operating volume as the number of audits increased. Therefore, our preliminary results do not support the notion that audit releases generally serve as a signaling mechanism.