연구정보
Home 연구정보
국내외 연구기관에서 발표된 중국 연구 자료를 수집하여 제공합니다.
연구보고서
Swap Volatility and Systemic Risk in Hong Kong Banking: A Machine Learning Approach
Paul D. McNelis 2022-06-08
This paper examines sources of contagion emanating from within the Hong Kong banking sector as well as external sources of contagion. For robustness, the paper uses two measures of systemic risk for major Hong Kong banks. One is based on Forecast Error Variance Decomposition (FEDV) from Vector Autoregressive (VAR) estimation of daily realized
volatility. The other comes from Delta Conditional Variance at Risk (ΔCoVar) analysis with quantile regression of weekly share-price changes. Our sample period covers the past twelve years, encompassing the Global Financial Crisis, the downgrading of US Debt, Brexit, increased trade tensions between the US and China, and the onset of the COVID-19
pandemic.
We make use of recent advances in Machine Learning methods, in particular Elastic Net with Cross Validation, for estimation, as well as Neural Networks for dimensionality reduction. These methods are particularly useful for analysis of data sets with a large number of regressors. The goal is to isolate which variables serve well for out-of-sample prediction, not in-sample statistical significance, while economising on the number of parameters.
Two questions take centre stage. Do any banks stand out as net transmitters of risk to the banking system as a whole? Do any external factors emerge as additional sources of systemic risk for the banking system in Hong Kong? Controlling for financial market indicators and indices of Economic Policy Uncertainty in the US and China, we find that measures of
implied volatility on Interest Rate Swap Options contracts from both the United States and Hong Kong have strong effects on banking share price volatility. This result should not be surprising since banks are among the largest participants in swap-options markets.
이전글 | 미중갈등하에서의 중국 반도체산업 경쟁력 | 2022-06-08 |
---|---|---|
다음글 | 중국, 근로자 대상 개인연금제도 시범 도입 | 2022-06-08 |