An innovative approach for optimising CCP default management through agent-based modelling

Abstract

This paper develops a rigorous model to analyse the market impact of liquidation. Invoking innovative techniques from agent-based simulation, we construct the order book's reaction function to liquidation. A methodology is then developed for establishing the optimal close-out strategy which balances the velocity of positional liquidation with that consequential market reaction function. This model can be used in parallel to more traditional calculations of market volatility to ensure that the overall risk capitalisation of a central clearing counterparty remains robust. In the case of financial markets, agent-based models often seek to account for the so-called stylised facts of financial markets. Stylised facts are simply empirical regularities that appear to be stable across markets and over time. Such facts range from statistical concepts, such as the distribution of returns, to more abstract notions such as stock market bubbles and crashes.