Stability of Financial Markets: An Evolutionary Approach

This project is concerned with the stability and dynamics of financial markets. How to reduce or eliminate the latent risk of a period of market instability (such as a crash) is a topic of an ongoing debate among politicians, regulators and practitioners, as witnessed, for instance, by the current discussion on the potential risk of the largely unregulated hedge fund industry at the 2007 G8 summit. Though these events are rare, their occurrence often has strong adverse effects on the macroeconomy. Our o bjective in this project is to improve the understanding of this risk and, in particular, its causes. The approach is based on the development and analysis of evolutionary models of financial markets. The main strengths of this approach are the explicit m odelling of the dynamic interaction of investors, the incorporation of endogenous change and adaptation of investment behaviour and the numerical tractability of large-scale systems with thousands of active investors. This model has excellent potential t o offer new insights as it aims at capturing two main aspects of financial markets which defy standard finance/economics approaches. These are the two major R&D challenges in this project. First, investors' behaviours are shaped by market interaction rath er than being a priori specified. This is achieved by a genetic programming approach. Second, the model captures different market microstructures in a unified framework. This facilitates the comparison of limit order markets (e.g. stock markets), dealersh ip markets (e.g. foreign exchange) and novel market-clearing mechanisms derived from theory. The applications comprise the statistical determination of financial market stability and the distribution of risk through the characterization of individual beh aviour and its interdependencies. The effects of extreme events can be quantified in simulation studies to shed light on empirical findings on market crashes.