EURO 2024 Copenhagen
Abstract Submission

EURO-Online login

3068. A network-based model for dynamic analysis of the banking sector

Invited abstract in session TD-6: Advancements of OR-analytics in statistics, machine learning and data science 15, stream Advancements of OR-analytics in statistics, machine learning and data science.

Tuesday, 14:30-16:00
Room: 1013 (building: 202)

Authors (first author is the speaker)

1. Riccardo Bastianutti
Economics, Unisalento

Abstract

This paper introduces an innovative model to interpret the dynamics of banking sector, moving away from the traditional reliance on individual firm variables. Instead, it proposes a network approach to capture information from inter-firm interactions.
The model utilizes Principal Component Analysis (PCA) to extract the first principal component, interpreted as a proxy for each firm's weight within the banking sector. A graph is constructed, with the first principal component acting as the weight for links between firms, and the firm with the highest weight serving as the network's center.
Applied to a dataset of banking sector firms, the model effectively captures sector dynamics. Active nodes, those with an indegree greater than zero, are identified and analyzed over time. Results are compared with the S&P 500 index, revealing a strong correlation (above 0.91).
Additionally, a predictive model is implemented on the probability density function (PDF) of the first principal component to analyze sector evolution. Predictive models demonstrate an error of less than 5%, highlighting the model's robustness.
The proposed model boasts several advantages over existing approaches. Firstly, it provides a more comprehensive view of the banking sector by considering inter-firm interactions. Secondly, it identifies key players in the banking sector and their impact on dynamics. Lastly, it offers predictive capabilities for the banking sector's future evolution.

Keywords

Status: accepted

Back to the list of papers