The Macroeconomic Determinants of Volatility in Precious Metals Prices in Ethiopia Using GARCH and Riskmetrics Models
No Thumbnail Available
Date
2014-06
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Abeba university
Abstract
Modelling and forecasting volatility for the price of precious metals has become a fertile field of
empirical research in financial markets. Since volatility is considered as an important concept in
many economic and financial applications. The objective of this study was to model and forecast
the volatility dynamics in precious metals prices in the Ethiopian market using GARCH family
and RiskMetrics models using data from January1998 to January 2014.
The price return series of gold and silver show the characteristics of financial time series such as
leptokurtic distributions and thus, can suitably be modeled using EWMA and GARCH family
models. Empirical investigation was conducted in order to model price volatility using EWMA
and GARCH family models. Among the GARCH family models considered in this study,
ARMA (0, 1)-GARCH-M (2, 2) model with Student’s t-distributional assumption of residuals
and ARMA (1, 3)-EGARCH (3, 2) model with normal distributional assumption of residuals
were found to be better fit for price volatility of gold and silver, respectively.
Saving interest rate, exchange rate and price of crude oil were found to have statistically
significant effect on monthly price volatility of gold. On the other hand, saving interest rate and
general inflation rate have statistically significant effect on monthly price volatility of silver. The
risk premium effect for GARCH-M (2, 2) model was positive and statistically significant. This
implies that an increase in volatility would increase the mean return. The asymmetric term was
found to be positive and significant in EGARCH (3, 2) volatility model for sliver. This is an
indication that unanticipated increase in price had larger impact on price volatility than
unanticipated decrease in the price of silver.
A comparison was made between GARCH family models and exponentially weighted moving
average (EWMA) model. The study suggests that GARCH class of models appear to be better in
volatility forecasting than EWMA model as judged by RMSE and MAE criteria.
Key words: EWMA model, GARCH model, Precious metals and Volatility
Description
Keywords
Ewma Model, GARCH Model, Precious Metals, Volatility