A Tree-Based Approach to Modelling Stock Exchange Index Returns in EU Countries

In this paper, we examine the stock exchange index returns for the panel of 27 EU countries in the last ten years. Our method of choice is a special kind of tree models, namely model baþed recursive partitioning (MOB). The MOB algorithm uses the generalized M-fluctuation test to examine the parameter stability for a given node. In this investigation, we attempt to explain the returns of the EU stock exchange indices with the help of gross domestic product, interest rates, and other financial and macroeconomic variables. The model-based recursive partitioning algorithm yields four terminal nodes pointing to GDP growth and rate of inflation as the splitting variables. During different inflation and GDP growth regimes, the observed explanatory variables impact the stock exchange index returns with varying intensity. The results are discussed and ýnterpreted in light of the current economic sitüation.