This paper provides a new methodology for the analysis of multiple long run relations in panel data models where the cross section dimension, n, is large relative to the time series dimension, T. For panel data models with large n researchers have focussed on panels with a single long run...

The study compares model approaches in predictive modeling for claim frequency and severity within the cross-border cargo insurance domain. The aim is to identify the optimal model approach between generalized linear models (GLMs) and advanced machine learning techniques. Evaluations focus on...