Network-based computational techniques to determine the risk drivers of bank failures during a systemic banking crisis

Research output: Contribution to journalArticle

Abstract

This paper employs a computational model of solvency and liquidity contagion assessing the vulnerability of banks to systemic risk. We find that the main risk drivers relate to the financial connections a bank has and the market concentration, apart from the size of the bank triggering the contagion, while balance sheets play only a minor role. We also find that market concentration might facilitate banks to withstand liquidity shocks better while exposing them to larger solvency chocks. Our results are validated through an out-of-sample forecasting that shows that both type I and type II prediction errors are reduced if we include network characteristics in our prediction model.
LanguageEnglish
Pages174
Number of pages184
JournalIEEE Transactions on Emerging Topics in Computational Intelligence
Volume2
Issue number3
Early online date23 May 2018
DOIs
StatusPublished - Jun 2018

Fingerprint

Computational techniques
Banking crisis
Bank failure
Market concentration
Solvency
Contagion
Systemic risk
Prediction model
Prediction error
Liquidity shocks
Balance sheet
Out-of-sample forecasting
Liquidity
Computational model
Vulnerability

Keywords

  • Solvency
  • liquidity
  • interbank loans
  • network topology
  • banking crises
  • systemic risk
  • systemic crisis
  • bank failure

Cite this

@article{ce05fa66dcaf48b0b5d3520598946c52,
title = "Network-based computational techniques to determine the risk drivers of bank failures during a systemic banking crisis",
abstract = "This paper employs a computational model of solvency and liquidity contagion assessing the vulnerability of banks to systemic risk. We find that the main risk drivers relate to the financial connections a bank has and the market concentration, apart from the size of the bank triggering the contagion, while balance sheets play only a minor role. We also find that market concentration might facilitate banks to withstand liquidity shocks better while exposing them to larger solvency chocks. Our results are validated through an out-of-sample forecasting that shows that both type I and type II prediction errors are reduced if we include network characteristics in our prediction model.",
keywords = "Solvency, liquidity, interbank loans, network topology, banking crises, systemic risk, systemic crisis, bank failure",
author = "Andreas Krause and Simone Giansante",
year = "2018",
month = "6",
doi = "10.1109/TETCI.2018.2805319",
language = "English",
volume = "2",
pages = "174",
journal = "IEEE Transactions on Emerging Topics in Computational Intelligence",
issn = "2471-285X",
publisher = "IEEE",
number = "3",

}

TY - JOUR

T1 - Network-based computational techniques to determine the risk drivers of bank failures during a systemic banking crisis

AU - Krause, Andreas

AU - Giansante, Simone

PY - 2018/6

Y1 - 2018/6

N2 - This paper employs a computational model of solvency and liquidity contagion assessing the vulnerability of banks to systemic risk. We find that the main risk drivers relate to the financial connections a bank has and the market concentration, apart from the size of the bank triggering the contagion, while balance sheets play only a minor role. We also find that market concentration might facilitate banks to withstand liquidity shocks better while exposing them to larger solvency chocks. Our results are validated through an out-of-sample forecasting that shows that both type I and type II prediction errors are reduced if we include network characteristics in our prediction model.

AB - This paper employs a computational model of solvency and liquidity contagion assessing the vulnerability of banks to systemic risk. We find that the main risk drivers relate to the financial connections a bank has and the market concentration, apart from the size of the bank triggering the contagion, while balance sheets play only a minor role. We also find that market concentration might facilitate banks to withstand liquidity shocks better while exposing them to larger solvency chocks. Our results are validated through an out-of-sample forecasting that shows that both type I and type II prediction errors are reduced if we include network characteristics in our prediction model.

KW - Solvency

KW - liquidity

KW - interbank loans

KW - network topology

KW - banking crises

KW - systemic risk

KW - systemic crisis

KW - bank failure

U2 - 10.1109/TETCI.2018.2805319

DO - 10.1109/TETCI.2018.2805319

M3 - Article

VL - 2

SP - 174

JO - IEEE Transactions on Emerging Topics in Computational Intelligence

T2 - IEEE Transactions on Emerging Topics in Computational Intelligence

JF - IEEE Transactions on Emerging Topics in Computational Intelligence

SN - 2471-285X

IS - 3

ER -