Price Risk Measurement of China’s Soybean Futures Market Based on the VAR-GJR-GARCH Model

Table 6

Comparison of GJR-GARCH (p, q) model.

	GJR-GARCH (1, 1)	GJR-GARCH (1, 2)	GJR-GARCH (2, 1)	GJR-GARCH (2, 2)
r	0.371	0.371	0.369	0.371
	(192.62)	(192.95)	(246.26)	(196.42)
ARCH
L.arch	0.378	0.378	0.419	0.364
	(10.33)	(10.31)	(13.24)	(9.31)
L2.arch			0.356	0.234
			(−10.82)	(5.12)
L.gjr-garch	−0.0841	−0.0788	0.0453	−0.0529
	(−1.38)	(−1.28)	(−5.38)	(−0.84)
L2.gjr-garch		0.0133		−0.0107
		(0.59)		(−0.20)
L.garch	−0.0237	−0.0663	0.962	−0.759
	(−0.81)	(−0.92)	(153.91)	(−10.14)
Cons	0.00345	0.00361	0.00000252	0.00592
	(22.61)	(13.52)	(3.05)	(16.89)
N	1219	1219	1219	1219
Log likelihood	1589.9	1590.1	1646.4	1595.4
AIC	−3169.7	−3168.1	−3280.9	−3176.8
BIC	−3144.2	−3137.5	−3250.3	−3141.1

t-statistics are given in parentheses. , , and .