Date of Award
Spring 2024
Degree Name
Bachelor of Science (BS)
Advisor(s)
Yusif Simaan
Abstract
The purpose of this research is to explore the Silicon Valley Bank (SVB) collapse with a Value-at-Risk (VaR) model using quantile regression of options’ implied volatility. The aim is to enhance the understanding of financial contagion with forward-looking data and to examine the advantages of a quantile regression VaR model of options implied volatility over a standard VaR model. The methodology centered on a quantile regression VaR Python model of options implied volatility compared against traditional VaR models relying on historical returns. Test cases involved multiple moving-days windows along with various confidence/quantile levels. Data collection involved historical prices, actual returns, and daily implied volatility of the most in-the-money options for approximately two years prior to SVB’s collapse for SVB, KRE (regional banking ETF) and XLF (financial sector ETF). Findings demonstrate the quantile regression model provided more accurate risk predictions compared to a standard VaR model of historical returns. Overall, there were found to be advantages of incorporating options implied volatility as a predictive risk tool for policy makers, investors, and risk managers as financial contagion continues to increase. This research contributes to the risk management literature through demonstrating the effectiveness of forward-looking data over historical data in estimating potential future losses, and therefore, improving risk analysis.
Recommended Citation
Koutsonikolis, John, "Silicon Valley Bank (SVB) Collapse Value-At-Risk (VaR) Analysis With Quantile Regression Of Options Implied Volatility" (2024). Gabelli School of Business Honors Thesis Collection. 150.
https://research.library.fordham.edu/gabelli_thesis/150
