Volume 16, Number 1, June 2021
| Stock Indices Forecast by Hybrid Model from GARCH Families: Evidence from Global Markets |
Abstract
Recently, the rising of the Covid-19 pandemic has delivered severe panic to the whole world and enhanced the caution of stockholders in the financial market. To alleviate the negative consequence of the epidemic and certain investment decisions, this paper aims to investigate the appropriate forecasting methods to capture the volatility of the equity market in a time-varying setting. Thus, we apply two comparative models, namely Box-Jenkins ARIMA-GARCH (1,1), and VAR (4)-BEKK-GARCH in six international markets. Finally, we found that two of these forecasting models can reflex the movement of the equity market sharply, even the period is hardly predictable. Furthermore, although the results indeed confirm the goodness of VAR (4)-BEKK-GARCH in predicting stock, we then interestingly indicate that ARIMA-GARCH outperforms its competitive model in practice.
Keywords: Covid-19, ARIMA-GARCH, VAR-BEKK-GARCH, Black Swans, Box-Jenkins
JEL Classification: C22, C53, D53