Abstract

This paper builds on the study by Engle and Manganelli (2004), titled “CAViaR: Conditional Autoregressive Value at Risk by Regression Quantiles,” published in the Journal of Business and Economic Statistics. It investigates the application of Conditional Autoregressive Value at Risk (CAViaR) within the stock markets of Iraq, Jordan, Oman, Saudi Arabia, Kuwait, and Egypt. The study analyzes daily time-series data of stock market indices from early 2011 to the end of 2023. CAViaR estimates for two risk levels (1% and 5%) are presented across four different models: Adaptive, Indirect GARCH, Asymmetric Slope, and Symmetric Absolute Value. Overall, the results indicate that the Adaptive model performs effectively at both risk levels, highlighting the positive impact of the variable. Additionally, the findings reveal performance differences among the various VaR models, underscoring the importance of selecting the appropriate model for risk analysis.