Introduction
Sirio employs a unique approach to risk management by conducting a detailed, asset-specific analysis of each token listed on the platform. This methodology focuses on key risk factors such as Smart Contract Risks, Counterparty Risks, and Market Risks. By performing comprehensive assessments of each asset, we can accurately classify the risk level of each token and establish precise risk parameters. Once we aggregate the data from individual assets, we use it to define global risk parameters.
Our streamlined approach enables dynamic and flexible risk management, which is crucial for adapting to changing market conditions and supporting lending and borrowing activities in emerging ecosystems. This methodology is inspired by the risk management practices of leading DeFi entities such as Aave, Gauntlet, and RiskDAO. Leveraging their open-source documentation has allowed us to gain insights into their frameworks. After defining the individual risk assessment for each token, we can calculate the overall risk score and establish global risk parameters, such as the Liquidation Threshold.
Our risk management process can be summarized in the following steps:
Risk per Asset: For each token we aim to list, we assign a Risk Factor that helps determine the token’s risk parameters: Caps, LTV and Reserves.
Supply & Borrow Caps: They represent the maximum amounts that can be deposited or borrowed for a specific token. These caps are crucial for managing various risks, such as infinite token minting or market manipulation. For example, if the Supply Cap for HBAR is 1,000,000, no more than 1,000,000 HBAR tokens can be supplied to the protocol. Similarly, if the Borrow Cap is 500,000, no more than 500,000 HBAR tokens can be borrowed.
Loan-To-Value (LTV): This determines the maximum amount that can be borrowed against a given deposited asset. For instance, if HBAR has an LTV of 80%, and you deposit $1,000 of HBAR, you can borrow up to $800 worth of any token listed on Sirio.
Reserves. Indicates the amount of tokens allocated to the protocol's reserves, supporting the business and providing a safeguard against rare but adverse events, such as a lack of liquidators for loans eligible for liquidation.
Once we establish the individual risk profiles, we aggregate them to define global risk parameters:
Liquidation Warning: Once the Liquidation Risk Level reaches this certain amount, users are alerted of potential liquidation risks on their dashboard.
Liquidation Threshold: This value indicates the level at which a loan becomes eligible for liquidation. The lower the threshold, the more conservative the approach.
Liquidation Prevention AI Model: Our AI model is trained using the parameters we set, such as the tokens' LTVs and Liquidation Thresholds. More conservative input parameters will result in a more conservative model.
Liquidation Fee: This represents the reward given to the platform after a liquidation. A lower penalty means higher reward for liquidators, so it increases the likelihood of liquidations, incentivizing developers and users to participate.
Our approach is divided into two phases: first, an explanation of the methodology, followed by its application to individual tokens. While our approach is inspired by top-tier risk management companies and lending protocols, it is specifically tailored to Sirio's unique environment—a protocol that has yet to launch on the mainnet within a developing ecosystem. This environment presents distinct risk factors, such as lower liquidity, higher volatility, and lower trading volumes—factors that more established DeFi platforms may not face to the same extent.
Our medium-term goal is to develop a publicly accessible Risk Dashboard, where we will update these parameters in real time, on a daily or weekly basis. Eventually, we aim to decentralize much of the risk management process through a governance system, while maintaining an admin risk panel for urgent tasks.
Last updated