Abstract:
The paper identifies the business models followed by banks in the euro area, utilising a proprietary dataset collected in the context of the supervisory reporting of the Single Supervisory Mechanism. The concept of a 'business model' has been neglected by economic theory and is defined here with respect to the set of activities performed by banks. We adopt a clustering methodology to provide evidence for the existence of distinct business models. Clustering is combined with dimensionality reduction optimally, given the nature of our dataset which features a large number of dimensions for each bank ('fat' data). The method produces a level and a contrast factor which are intuitive in the economic sense. Four business models are identified alongside a set of 'outlier' banks that follow unique business models. The risk and performance indicators of each cluster are examined and evidence is provided that they follow distinct statistical distributions.