International Journal of Statistics and Applied Mathematics
2017, Vol. 2, Issue 6, Part B
A time series evaluation of the asymmetric nature of heteroscedasticity: an EGARCH approachAuthor(s):
Imoh U Moffat, Emmanuel A Akpan and Ubon A AbasiekwereAbstract:
Symmetric GARCH models have provided a rich and useful approach to capturing the conditional variances in stock returns, thereby filling the gap created by the inability of ARIMA models to account for the presence of volatility clustering that leads to the violation of assumption of constant variance. However, symmetric GARCH models are deficient in capturing the leverage effects and their parameters are restricted to ensure that the conditional variance is positive. In order to salvage these deficiencies in the symmetric GARCH model, we employed the Exponential GARCH (EGARCH) model that can capture the asymmetry property of the stock returns (leverage effects) and remove restrictions on parameters by introducing natural logarithm to the conditional variance. We considered the time series on the closing share prices obtained from the Nigerian Stock Exchange on Zenith bank plc spanning from January 4, 2006 to December 30, 2015. ARIMA(1,0,3) model was fitted to the return series. Based on the residuals of the fitted model, the ARCH effect was detected and captured by GARCH(0,1) model while the leverage effect in the return series was captured by EGARCH(0,1) model. Further findings from the EGARCH (0,1) model revealed the asymmetric nature of heteroscedasticity as indicated by the significance of the negative coefficient of the asymmetric parameter. The implication of the findings is that an unexpected decrease in price increases the predictable volatility more than an unexpected increase in price of similar magnitude.Pages: 111-117 | Views: 606 | Downloads: 13Download Full Article: Click Here
How to cite this article:
Imoh U Moffat, Emmanuel A Akpan, Ubon A Abasiekwere. A time series evaluation of the asymmetric nature of heteroscedasticity: an EGARCH approach. Int J Stat Appl Math 2017;2(6):111-117.