Comparing network-based taxation schemes to reduce systemic risk in banking systems

Loading...
Thumbnail Image

URL

Journal Title

Journal ISSN

Volume Title

Perustieteiden korkeakoulu | Master's thesis

Date

2019-06-17

Department

Major/Subject

Complex Systems

Mcode

SCI3060

Degree programme

Master’s Programme in Life Science Technologies

Language

en

Pages

57

Series

Abstract

Difficulty in predicting financial crises may lead to attempts of getting ready to them instead. Authorities construct rules for financial institutions in order to secure the economies from catastrophic consequences of systemically important financial institutions defaults. In order to create proper stimuli for the economic agents to avoid deals exposing the entire financial system, the tax on the systemic risk was proposed. While systemic importance of financial institutions might be measured in different ways, different metrics (namely: degree centrality, SinkRank, acyclic DebtRank, cyclic DebtRank and 2-step DebtRank) are compared as candidates for the systemic risk tax basis. Economics, being in focus of complex systems disciplines, is very difficult to investigate due to the impossibility to set an experiment. In order to test a certain policy, one may use computer simulations. In this thesis, an agent-based model is used to investigate the simulated economy evolution under different credit taxation policies. For this special purpose, one more systemic risk metric was proposed (it is called 2-step DebtRank). In addition, an equity layer for the interbank obligations network was introduced in order to investigate interlayer interaction. As a result of the taxation schemes comparison, acyclic and cyclic DebtRanks demonstrated the best performance as systemic risk taxes; 2-step DebtRank appeared to perform closely to other two DebtRanks; SinkRank appeared to be inappropriate for this purpose. The introduction of the equity layer does not create many essentially new results, however, the interlayer antiphase oscillation of the systemic risk was observed.

Description

Supervisor

Saramäki, Jari

Thesis advisor

Saramäki, Jari

Keywords

agent-basent modelling, crises, interbank networks, banking risks, systemic risk, DebtRank

Other note

Citation