Insights in the Underlying Complex Mechanisms of Dutch Companies using Agent Based Modelling

This paper investigates the underlying complex mechanisms of the Dutch business sector using agent based modelling. Based on three stylized facts that have been observed in a data set of all firms in the Netherlands, the Complex Adaptive Trivial System (CATS) model by Gatti, Gallegati, Giulioni, and Palestrini (2003) is employed for the research. Several extensions are implemented in order to investigate whether the model can be improved in terms of replicating the empirical stylized facts. A calibration analysis using the Generalized Method of Simulated Moments has been performed to estimate the parameters of the original CATS model and its extended variants. Finally, an ex-post validation analysis compares the statistical characteristics of the simulated output with the empirical stylized facts. The results indicate that the CATS model is capable to reproduce all three stylized facts to a certain extent. Moreover, it is shown that some of the implemented extensions are an improvement of the original CATS model. w