Forecast Performance of the Taiwan Weighted Stock Index: Update and Expansion

This research introduces the following to establish a TAIEX prediction model: intervention analysis integrated into the ARIMA–GARCH model, ECM, intervention analysis integrated into the transfer function model, the simple average combination forecasting model, and the minimum error combination forecasting model. The results show that intervention analysis integrated into the transfer function model yields a more accurate prediction model than ECM and intervention analysis integrated into the ARIMA–GARCH model. The minimum error combination forecasting model can improve prediction accuracy much better than non-combination models and also maintain robustness. Intervention analysis integrated into the transfer function model shows that the TAIEX is affected by external factors, the INDU, the exchange rate, and the consumer price index; therefore, facing the different issues of the TAIEX, the government could pursue some macroeconomic policies to reach the goals of policies.