Modeling Markov Switching ARMA-GARCH Neural Networks Models and an Application to Forecasting Stock Returns

Table 2

Baseline volatility models.

Baseline GARCH models, single regime
(1) RW	C							log𝐿
	0.001873***; (0.000391)							12241.56
(2) GARCH	ARCH	GARCH	C					log𝐿
	0.1591259***; (0.0068625)	0.8280324***; (0.0056163)	0.0000206***; (1.80𝑒−06)					13174.66
(3) APGARCH	APARCH	APARCH_E	PGARCH	POWER	C			log𝐿
	0.2153934***; (0.0076925)	−0.0319876***; (0.0160775)	0.7801279***; (0.0074049)	1.354999***; (0.0707303)	0.0004055***; (0.0001074)			13111.96
(4) FIAPGARCH	ARCH; (Phi1)	GARCH; (Beta1)	D-FIGARCH	APARCH; (Gamma1)	APARCH; (Delta)	C; (MEAN)	C; (VAR.)	log𝐿
	0.250877***; (0.096563)	0.449631***; (0.10367)	0.417296***; (0.040388)	0.027382; (0.033565)	2.020946***; (0.096663)	0.001635***; (0.00028929)	8.054283; (6.6685)	13196.30