Status : Verified
| Personal Name | Fernando, Leilani G. |
|---|---|
| Resource Title | An Application of Selected Estimation Models for Peso-US Dollar Exchange Rate value-At-Risk |
| Date Issued | March 2025 |
| Abstract | With the globalization and uncertainty in the financial market, continuously enhancing and developing effective risk management practices such as adopting effective risk measures and processes that can capture true underlying risk of the portfolio is important. In this study, four Value-at-Risk (VaR) models are evaluated and compared for the Peso-US Dollar exchange rate VaR: (1) Autoregressive Moving Average – Generalized Autoregressive Conditional Heteroscedastic (ARMA-GARCH) type models with normal or Student-t distributions; (2) Static Extreme Value Theory – Peak-Over-Threshold (EVT-POT); (3) Generalized Autoregressive Conditional Heteroscedastic – Generalized Pareto Distribution (GARCH-GPD) approach of Mc Neil and Frey; and (4) Bayesian GARCH with Student-t innovation. The selected models are assessed using Backtesting for the accuracy, and Mean Relative Bias (MRB) and Root Mean Squared Relative Bias (RMSRB) for the conservatism of the models. Findings show that Dynamic EVT, GARCH type and Bayesian GARCH models may work well in terms of accuracy and conservatism. Among the three models, Bayesian GARCH approach may be considered as the best model for both 1-day returns series and 10-day returns series using 95% and 99% confidence levels as it shows consistency in terms of conservatism and accuracy. |
| Degree Course | Master in Statistics |
| Language | English |
| Keyword | Value-at-Risk (VaR); Market Risk Management; GARCH Type Models; Extreme Value Theory (EVT); Bayesian GARCH |
| Material Type | Thesis/Dissertation |
Preliminary Pages
11.01 Mb
Category : F - Regular work, i.e., it has no patentable invention or creation, the author does not wish for personal publication, there is no confidential information.
Access Permission : Open Access