Indirect Estimation of the Parameters of Agent Based Models of Financial Markets

Agent based models take into account limited rational behaviour of individuals acting on financial markets. Explicit simulation of this behaviour and the resulting interaction of individuals provide a description of aggregate financial market time series. At least for some parameter settings, the outcome of such simulations exhibit marked similarities with actual financial market time series. The goal of this paper is twofold. First, we compare simulation results of agent based models with observed time series based on characteristic moments like ARCH--effects or excess kurtosis. Second, we try to estimate the parameters of the agent based model from the observed data using a simulated indirect estimation method based on the characteristic moments. The paper presents details of this estimation approach and first results for the US/DM exchange rate.