Predicting the returns of the US real estate investment trust market: evidence from the group method of data handling neural network

Financial Innovation

Table 9 Coefficients of these scenarios based on the GARCH model

	\({\mathrm{r}}_{\mathrm{t}}=\mu +\alpha {r}_{t-1}+{\varepsilon }_{t}\)		\({\upsigma }_{\mathrm{t}}^{2}=\omega +\beta {\varepsilon }_{t-1}^{2}+\gamma {\sigma }_{t-1}^{2}\)
Scenarios	\(\upmu\)	\(\mathrm{\alpha }\)	\(\upomega\)	\(\upbeta\)	\(\upgamma\)	MSE
G-300/60	0.0710***	0.0691**	0.0493***	0.1311***	0.7613***	1.9317
G-600/30	0.0711***	0.0693***	0.0491***	0.1307***	0.7615**	1.6315
G-300/30	0.0703**	0.0652***	0.0474***	0.1265***	0.7521***	1.8723
G-200/30	0.0723***	0.0673***	0.0511**	0.1329***	0.7631***	2.2102
G-100/30	0.0812***	0.0686***	0.0525***	0.1341**	0.7671**	3.7935
G-30/30	0.0896**	0.0690**	0.0562***	0.1368**	0.7713***	4.9346