Improving Forecasts of the EGARCH Model Using Artificial Neural Network and Fuzzy Inference System

Table 4

The EGARCH-ANN (1, 2, 3, 1) model fitted to the given data.

Parameters	Coefficients	Std. errors	t-stats
K	−3.7087	0.5434	−6.8248
ARCH1	−0.1665	0.0696	−2.3914
GARCH1	0.2668	0.0860	3.1021
GARCH2	0.6137	0.2389	2.5683
Leverage1	−0.4321	0.2121	−2.0373
LeveragebyANN1	−0.3534	0.1003	−3.5246

With addition information like log likelihood: 831, Akaike information criterion: −1649, Bayesian information criterion: −1620, and Hannan–Quinn information criterion: −1638.