The Financial Risk Measurement EVaR Based on DTARCH Models

Area of Science:

• Quantitative Finance
• Econometrics
• Financial Risk Management

Background:

• Financial risk measurement is crucial, especially for extreme events.
• Expectile value at risk (EVaR) offers a robust approach to quantifying financial risk.
• Double-threshold autoregressive conditional heteroscedastic (DTARCH) models capture asset return volatility with piecewise linear functions.

Purpose of the Study:

• To propose a novel weighted composite expectile regression (WCER) estimation for DTARCH models.
• To enhance the prediction of extreme financial risk using EVaR when asset return characteristics are nonlinear.
• To address limitations of existing DTARCH models by not assuming known threshold and delay parameters.

Main Methods:

• Development of the weighted composite expectile regression (WCER) framework.
• Application of expectile regression theory to DTARCH models.
• Simulation studies to evaluate the performance of the WCER estimation in finite samples.

Main Results:

• The proposed WCER estimation demonstrates adequate and promising performance in finite samples.
• The method effectively estimates DTARCH models without prior knowledge of threshold and delay parameters.
• Successful application of the approach to analyze the Hang Seng Index (HSI) and Standard & Poor's 500 Index (SPI) daily returns.

Conclusions:

• The WCER estimation provides a valuable tool for assessing extreme financial risk via EVaR.
• The method offers an improvement over existing DTARCH models by relaxing parameter assumptions.
• The empirical analysis on HSI and SPI validates the practical applicability of the proposed approach.