OpusARM
Artificial Intelligence for risk management
Operating off-chain, OpusARM (Automated Risk Manager) is responsible for implied volatility curve calibration, on-demand option pricing, and risk management.
An efficient Risk Management System (RMS) needs to perform various tasks, such as:
Continuously track market conditions and relevant risks
Identify exogenous and endogenous trading and investing patterns.
Calibrate OpusAMM’s pricing and risk management decisions.
Support its LP’s and LT’s to ensure positive outcomes.
Incentivize non predatory trading flows.
Take advantage of systematic arbitrage option trading opportunities.
Eliminate threats (ex. predatory trading practices and market manipulation).
OpusDigital adopts Reinforcement Learning and Swarm Learning techniques to optimize these risk management decisions. The implementation is both local and global.
Risk management at portfolio level
Each portfolio (for example, FX Options for a given currency pair form a portfolio) adopts its own algorithm to manage its risk with total or partial degree of automation, depending on the maturity of that particular portfolio and on the asset class. The methods to implemet this are, for the example of FX Options:
AI-driven algorithms: to predict price, volatility and rates change and fine-tune the volatility surface to keep the portfolio on or above the efficient risk/reward ratio frontier.
Quantitative algorithms: to asses arbitrage opportunities such as relative value volatility arbitrage, and to calculate hedging requirements to stay within the acceptable risk boundaries.
Human intervention: to study the behaviour of the market and of the system, collect data, and drive research and development to the most critical areas.
Risk management at the global level
OpusDigital is formed by various portfolios (FX Options for various currencies, Crypto Options, NFTs...) and although every portfolio is risk-managed locally, the OpusARM ensures that the risk of the entire echosystem is controlled. This is of key importance, because while LT's are responsible for their portfolios, LP's and shareholders are passively earning the market making income from the inflow coming from all the portfolios: a central control system protects their assets.
This is implemented using various key technologies:
Blockchain: as one of Opus' key differentiating factors, every transaction coming from the various portfolios is mined in Opus' proprietary blockchain. This ensures safety, compliance, and enforces that all transactions follow unique accounting and risk principles: if a transaction of some portfolio will not align with the echosystem risk boundaries, it will simply be rejected by the blockchain and reverted.
Artificial Intelligence: particularly Swarm Learning, in which the data coming from the various portfolios are used to continuously train the central model, aimed to optimize the validation parameters of the central blockchain. Every feedback from each node is therefore, by reflection, propagated to all other portfolios, in a continuous learning which makes the echosystem always more optimized as time goes.
Last updated