Dr Monira Al-Oud PhD Student (January 2009 - March 2013) Computational Finance and Economics Research Laboratory Computational Finance and Economics Laboratory Department of Computing and Electronic Systems University of Essex Email @essex.ac.uk: memalo

Agent-based Market Analysis

Monira works on the High-frequency Finance Project. The proposal is to use an agent-based approach to model trading behaviour in high-frequency markets. The research will focus on the foreign exchange market. We shall follow the approach by Richard Olsen (Visiting Professor at CCFEA). The initial part of this approach is factual description: the task is to observe the micro-behaviour of traders in the market and define stylized facts. This will be performed on anonymised data provided by OANDA. The next step could take two forms:

to model the traders;

to understand how the stylized facts emerged.

The approach is to study the interaction between various trading strategies, such as strategies based on directional changes. Results of interactions between different traders will be observed. Such observations will be used as feedback to change the model of interactions. The aim is to find models under which stylized facts are exhibited. Apart from finding conditions under which stylised facts occur in this market, we intend to look for conditions under which collapses occur.

The approach is to build on Dr Serafin Martinez-Jaramillo's PhD work at University of Essex, which used artificial markets to study behaviour of financial markets. In Serafin's work, a framework called CHASM, which stands for Co-evolutionary Heterogeneous Artificial Stock Market, was developed. CHASM supports modelling of fundamental characteristics of real markets by instantiating different types and numbers of agents (including fundamental, technical, noise or hybrid agents). Technical traders are based on EDDIE (developed by Dr Jin Li at University of Essex), which was able to learn and improve its performance in the market. Martinez-Jaramillo used CHASM to examine the statistical properties of end-of-day data series in the stock market. Following that, he identified conditions under which artificial markets exhibit stylised facts, i.e. the prices generated collectively by the agents resemble prices in real markets.

Monira was supervised by Edward Tsang (2009-2012) and Maria Fasli (2012-2013). She was examined by Peter McBurney (External Examiner) and Sam Steel (Internal Examiner). She passed her viva on 21st March 2013.

This work is related to but independent of Shaimaa Masry's work.