AGENT_BASED COMPUTATIONAL ECONOMICS FINANCIAL MODELLING

Welcome to ACEFINMOD.com


Welcome to ACEFINMOD. This web site, started in 2009, is dedicated to show-case the research carried out since about 2002 by a team of researchers and scientists associated with the Centre for Computational Finance and Economic Agents. The work reported here reflects a complex adaptive systems perspective and the methodology of Agent-based Computational Economics to build large scale data base driven financial and economic models for purposes of market and policy design. Much of the work was pioneered while Sheri Markose was Founder Director of CCFEA, while on going work represents collaborations on projects funded by grants or those done as PhD research under Sheri's supervision.

 

Digital Modeling of Financial Contagion

This project builds a multi-sector agent based simulator of the US financial system which is driven by the FDIC data set. The model is first built sector by sector starting with the CDS market.

Financial network analysis and systemic risk from contagion are the key objectives of this project.

Quick link to CDS Network Simulator beta 03

Quick link to CDS Network Simulator beta 02

Older version beta 01

"Too Interconnected To Fail" paper

Qiuck link to varius sections

Financial Network of CDS

CDO-ABX Model

Wholesale and Interbank Money Market

 

Financial Databases

The Database Manager is a powerful tool for accessing data from public sources, visualizing it and chanelling it to the different large scale simulators. Issues such as homogenizing, synchronizing etc of the diverse data sets needed for a single simulator platform is explained (see Tutorials)

To date we have dealt with the FDIC, BIS, DTCC and Datastream.


Systemic Risk from Financial Derivatives: Network Analysis

At least since the NBER paper by Michael Darby (1994), systemic risk from financial derivatives has been considered as a threat due to the sheer size of the OTC markets (over $603 tn) and the concentration of the market. To this may be added the correlated movements in the underlying of derivatives, interest rates, exchanges rates, sovereign debt etc that trigger multiple obligations. Here we provide the first comprehensive network analysis of the derivatives markets that can highlight structural fragility.

 

Next Generation Electronic Trading Simulators for London SETS (NGETS)

In this project, the London SETS electronic order book market is fully rebuilt and is the basis of a simulator to design and build algo strategies in real time. Detailed analysis of the structure and statistical properties of the orderbook are possible.

Run NGETS app

 

Sovereign Risk Contagion

In this project, the BIS bilateral data on immediate obligations between countries relative to their assets is visualized in a financial network. We proceed to analyse the financial contagion affecting the PIIGS (Portugal, Italy, Ireland, Greece, Spain) & UK.

Run SOVEREIGN CONTAGION app

Markets as Complex Adaptive Systems

Lectures on CAS pioneered by Sheri Markose at the Centre for Computational Finance and Economic Agents since 2002-2009 follows the perspective of Stephen Wolfram, John Casti and Chris Langton. The sine qua non of CAS is its capacity to produce novelty and surprises, new objects that were previously not there. This framework that reflects ubiquitous phenomena such as the Liar, surprises and the arms race of innovation to escape from hostile agents- is outside the ambit of mathematical economics and game theory.


CLIMACE, Green Transport and Smart Infra-structure

CLIMACE investigates a multi-agent model for digitally mapping the transition in economies

from high to low carbon.

Green Transport and Smart Infrastructure Models

covers research in this field


 

Extreme Financial Events Modelling

Extreme non-Guassian events mark financial markets. The research reported here is based on the work done by Sheri Markose and Amadeo Alentorn on developing a closed form solution for equity options assuming a Generalized Extreme Value (GEV) Risk Neutral Density Function.

Work on regime switching models for portfolio management is also reported.