Market Risk Management for China's Banking Industry: Evaluation of Value-at-Risk Approaches
Che, Xiaoying (2006) Market Risk Management for China's Banking Industry: Evaluation of Value-at-Risk Approaches. [Dissertation (University of Nottingham only)] (Unpublished)
Due to the rapid growth of the increasingly complex trading activities at large commercial banks and the numerous instances of financial market instability, there is a growing concern of risk management throughout the banking industry. One widely adopted technique to manage risk involves the use of Value-at-Risk, the benchmark for quantifying market risks. As VaR has become more widespread, the range of methods for computing VaR forecasts has broadened. However, in China's banking industry the use of VaR models is just at the early stages. Therefore this dissertation addresses the issue of evaluating some of the VaR methods for China's banking industry. These methods include the Historical Simulation, the RiskMetrics, the GARCH(1,1)-N, NGARCH(1,1)-t(d) and the Filtered Historical Simulation. And the approach to evaluate the performances of these methods is backtesting. The backtesting results indicate that the Filtered Historical Simulation might be the most accurate and reliable approach to forecast VaR in the context of China, while the RiskMetrics seems to be the least appropriate one. Furthermore, although the GARCH family models may capture the important features of returns, such as the autocorrelation, they may not be appropriate for VaR forecasts in China.
Actions (Archive Staff Only)