Adaptive Information Systems and Modelling
in Economics and Management Science
Artificial Financial Markets and Investment
(coordinators: E.J.Dockner and J.Zechner)
Traditional models in finance rest on the assumptions of informationally efficient markets and rational agents that have symmetric information, identical (homogenous) preferences and that build their expectation on all available information. Under these assumptions many stylised facts observed from real financial markets as well as the existence of many financial contracts and even the existence of important financial institutions cannot be properly explained. In a world with rational expectations and efficient markets everything is driven by unanticipated uncorrelated shocks (technological shocks, changing preferences, etc.)
The main objective of Initiative #6 is to explore how information is acquired and aggregated in financial systems, how agents form their expectations and whether or not asset prices and returns are predictable. For that purpose models and theories are introduced that explicitly account for one or several of the following features: boundedly rational agents, heterogeneous preferences, market imperfections and information asymmetries between parties within the financial system. In particular three main areas of research will be explored over the next period. These are continuations of the work that has been done during the first period of the SFB.
The first one deals with alternative models of expectation formation and their role in the pricing process of financial securities. Different models with adaptive agents are introduced and the following problems analysed: (i) under which alternative expectations hypotheses do equilibrium prices in a financial market still converge to the rational expectations equilibrium? (ii) are alternative theories of expectation formation and the resulting market dynamics capable of explaining stylised facts in financial markets? The second area deals with empirical issues that are related to market efficiency and information processing. In particular the issues of market based volatility models and their interactions with option prices will be explored in detail for various markets and sample periods. These studies help us to shed light on the dynamics and, hence, on stylised facts of financial markets and on the efficiency of these markets. The third area of research stresses the role of learning and information acquisition within corporations or regulatory agencies.
complete text of the proposal
Participants from our group:
Thomas Dangl (till August 2001)
Participants from Josef Zechner's group:
Finance and Physics