Forecasting performance of asy metric GARCH models

الملخص EN

Considerable research effort has focused on the forecasting of asset return volatility.

Debate in this area centers a round the performance of time series models, in partioular GARCH, relative to implied volatility from observed option premiums.

Existing literature suggests that the performance of any volatility forecast is sensitive.

This paper rigorously examines the forecasting performance of four GARCH(1, 1) models (GARCH, EGARCH, GJR and APARCH) used with three distributions (Normal, Student-t and Skewed Student-t).

We explore and compare different possible sources of forecasts improvements', asymmetry in the conditional variance, fat-tailed distributions and skewed distributions.

The Jordanian stock index are studied using daily data over a 12-years period from 1992-2003.

The results provide considerable insight into the performance of these alternative volatility forecasting procedures over forecasts improvement.

The evidence suggests that improvements of the overall estimation are achieved when asymmetric GARCH are used and when, fat-tailed distributions are taken into account in the conditional variance.

Moreover, it is found that GJR and APARCH give better forecasts than symmetric GARCH.

Finally increased performance of the forecasts is not clearly observed when using non-normal distribution.