Spillover Effects among Financial Institutions: A State-Dependent Sensitivity Value-at-Risk Approach

Item Type Journal paper
Abstract In this paper, we develop a state-dependent sensitivity value-at-risk (SDSVaR) approach that enables us to quantify the direction, size, and duration of risk spillovers among financial institutions as a function of the state of financial markets (tranquil, normal, and volatile). Within a system of quantile regressions for four sets of major financial institutions (commercial banks, investment banks, hedge funds, and insurance companies) we show that while small during normal times, equivalent shocks lead to considerable spillover effects in volatile market periods. Commercial banks and, especially, hedge funds appear to play a major role in the transmission of shocks to other financial institutions. Using daily data, we can trace out the spillover effects over time in a set of impulse response functions and find that they reach their peak after 10 to 15 days.
Authors Adams, Zeno; Füss, Roland & Gropp, Reint
Journal or Publication Title Journal of Financial and Quantitative Analysis
Language English
Keywords Risk spillovers; state-dependent sensitivity value-at-risk (SDSVaR); quantile regression; financial institutions; hedge funds
Subjects economics
HSG Classification contribution to scientific community
Refereed Yes
Date 30 May 2014
Publisher Cambridge University Press
Place of Publication Cambridge UK
Volume 49
Number 3
Page Range 575-598
Number of Pages 24
ISSN 0022-1090
ISSN-Digital 1756-6916
Publisher DOI https://doi.org/10.1017/S0022109014000325
Depositing User Dr. Zeno Adams
Date Deposited 29 Oct 2012 16:18
Last Modified 20 Jul 2022 17:14
URI: https://www.alexandria.unisg.ch/publications/217575

Download

Full text not available from this repository.

Citation

Adams, Zeno; Füss, Roland & Gropp, Reint (2014) Spillover Effects among Financial Institutions: A State-Dependent Sensitivity Value-at-Risk Approach. Journal of Financial and Quantitative Analysis, 49 (3). 575-598. ISSN 0022-1090

Statistics

https://www.alexandria.unisg.ch/id/eprint/217575
Edit item Edit item
Feedback?