OpusAI

What is OpusAI

OpusAI is a Swarm Intelligence ecosystem aimed to optimize Opus' portfolios in terms of risk and return. Its architecture consists of two main parts:

  • A set of decentralized Nodes, which can be identified by specific portfolios, each one running its own Machine Learning model to automate its risk management and market making operations.

  • A centralized Blockchain, whose code is proprietary but whose data is publicly visible, which learns from the Nodes and validates their operations by setting risk and accounting standards every Node must adhere to.

It can thus be identified with the union of the ARM and AMM and on the AI algorithms to automatically improve the performance of these two components.

The goal of OpusAI

OpusAI is engineered in such a way that the entire set of portfolios is optimized in a fully automated way: each Node continuously learns from its own set of trades, and at the same time communicates (via the central Blockchain) to all the other Nodes the current status of its risk, liquidity, P&L, PV et cetera. The central Blockchain then assures that, the ec0osystem is globally healthy.

This is similar to what is done in trading desks in the main investment banks, hedge funds and wealth management venues, if we consider the Nodes as the various desks and the central Blockchain as the CRO and CFO departments. Indeed, each desk has its own set of clients, has the freedom to take risks, has its budget and liquidity requirements. All desks submit their portfolios to the central system who then decides whether to approve a new trade, based on its global policy and current risk situation.

The advantages

This system brings numerous advantages:

  • Cost-effectiveness: the automation at the Node level removes the need of a team of traders for every portfolio, and the central Blockchain validation process makes the settlement and approval process fast and secure without the need of a risk management team. This in turn allows the portfolio to apply more convenient spreads and attract more volume.

  • Compliance: by assuring the Blockchain's validation process is compliant with the regulator's rules (for example KYC and AML practices, regional restrictions, data retention laws etc...), such compliance is automatically propagated to all the nodes, thus assuring a 100% compliant echosystem.

  • Cutting-edge technology: fostering AI and Blockchain, the most advanced technologies and algorithms are applied to give a competitive edge to the benefit of the shareholders, users and liquidity providers.

Example of a training loop

We outline a simplified example of how this echosystem works and how it automatically learns from the market. We assume that the system is already mature during this example: initialization topics will be covered in a dedicated documentation.

  1. A liquidity taker arrives to OpusDigital platform: he or she wants to open a FX Option on the EURUSD pair, at a given strike and maturity.

  2. The OpusAMM calculates the price based on the particular algorithm (with or without AI) running on the EURUSD FX Options portfolio.

  3. The OpusARM checks that the new trade is in line with the risk boundaries of both the portfolio (local ARM) and of the entire echosystem (global ARM). If it is, the trade proposal is delivered to the LT, and if accepted, the deal is closed. If it is not, the trade is not possible and the user is notified.

  4. Assume the trade is accepted. The OpusARM will monitor the trade continuously and collecting data on P&L, risk dynamics, and lifecycle. Based on its performance, the global AI will register whether the trade is profitable or not, and will adjust its acceptance parameters for the next trades.

  5. If now another LT arrives, even at a different portfolio (say ETH crypto options), it will submit a transaction to the global ARM, which is now updated with the knowledge taken from the previous trade (and all the others in its history). In this way, every portfolio benefits of the training from the entire echosystem.

Last updated