VAR AND EXPECTED SHORTFALL IN FOREIGN EXCHANGE RISK MANAGEMENT IN KAZAKHSTAN VARIOUS MODELS’ FORECASTING PERFORMANCE EVALUATION

Loading...
Thumbnail Image

Date

2023

Journal Title

Journal ISSN

Volume Title

Publisher

Nazarbayev University Graduate School of Business

Abstract

The thesis paper assesses methodologies, including Historical Simulation, Parametric Approach, ARMA-GARCH, and ARMA-GJR-GARCH, for quantifying Value at Risk and Expected Shortfall in currency pairs involving the Kazakhstani tenge. Both Parametric and GARCH-type models assume Normal Distribution, Student's t distribution, and Skewed Student's t distribution. Comprehensive analysis revealed no universally superior model; each exhibited distinct strengths and weaknesses, complicating the identification of an optimal risk measurement approach. Additionally, it is observed that the GJR-GARCH model does not consistently outperform the GARCH model, despite its extension to incorporate asymmetric volatility. GARCH-type models that assume skewed Student’s t distributions and Student’s t distributions for the residuals, demonstrated favorable outcomes for currency pairs originating from developed markets. Additionally, models exhibited improved predictions for currency pairs from developed countries during periods of floating exchange rate adoption. In contrast, more accurate results were observed for currencies of emerging countries in the period spanning 2009 to 2015

Description

Keywords

Type of access: Embargo

Citation

Kurmanbek, D., Oglanbek, A. (2023). VaR and Expected Shortfall in Foreign Exchange Risk Management in Kazakhstan Various models’ forecasting performance evaluation. Nazarbayev University Graduate School of Business