Understanding investors profile using a network approach

The objective of this project thesis is to analyze and understand the investors present in the client base of BNP Paribas Fortis, with which this collaborative project was carried out. In this thesis, we decided to opt for a network approach, considering the investors as part of a graph and try to detect communities. This master thesis will cover the path that led us from a list of links between investors and products to a final partition of investors presenting homogenous investment profiles. Each of the detected group will be subject to in-depth study of its characteristics in order to make sure that the clusters are coherent. Finally, this thesis proposes a method to extend the detected segmentation to new users that were not part of the sample subject to the clustering. Attention is paid on the one hand to the recognition of each cluster and on the other hand to the facility of customization of the model to make it easy to tune and adapt to the company’s need.