Table 5.

The BTKS tests of asymmetric risk spillover (CoVaR and ΔCoVaR measurements).

Direct dependency					Indirect dependency
China spillover effect	CSI300	CSI300	CSI300	CSI300	CSI300	CSI300	CSI300
	HS	SP500	N225	KS	SP500|HS	N225|HS	KS|HS
The tranquil period
CoVaR	−3.729	−2.4169	−2.1282	−2.2048	−1.2695	−1.7195	−1.4395
	(2.5121)	(2.5287)	(1.0984)	(1.4155)	(1.6543)	(1.1243)	(1.4243)
ΔCoVaR	1.2830	0.8619	0.4982	0.6458	0.3287	0.1487	0.3147
	(0.1541)	(0.0068)	(0.0065)	(0.0121)	(0.3154)	(0.0644)	(0.1846)
BTKS	0.4123	0.6854	0.5497	0.8733	0.1856	0.1486	0.3956
	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]
The 2015 crisis
	CSI300	CSI300	CSI300	CSI300	CSI300	CSI300	CSI300
	HS	SP500	N225	KS	SP500|HS	N225|HS	KS|HS
CoVaR	−3.312	−2.2592	−3.4962	−1.708	−1.2071	−2.1192	−1.0463
	(0.4102)	(0.6126)	(11339)	(0.2956)	(1.8164)	(1.1535)	(0.6916)
ΔCoVaR	0.8743	0.5137	0.8137	0.2838	−0.2361	0.0255	−0.0512
	(0.0166)	(0.0254)	(0.0254)	(0.0017)	(0.3163)	(0.0242)	(0.692)
BTKS	0.9054	0.4251	0.5252	0.7731	0.0000	0.0155	0.7938
	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[1]	[0.8570]	[1]

Direct dependency					Indirect dependency
US spillover effect	SP500	SP500	SP500	SP500	SP500	–	–
	HS	CSI300	N225	KS	CSI300|HS
The tranquil period
CoVaR	−2.8229	−3.3169	−2.7382	−2.3188	−2.0955	–	–
	(0.7977)	(1.6287)	(1.1284)	(1.6155)	(1.1243)
ΔCoVaR	1.1130	0.6919	1.4995	1.1658	0.2051	–	–
	(0.0141)	(0.0197)	(0.0172)	(0.2025)	(0.1844)
BTKS	0.7033	0.395	0.7991	0.8733	0.2031	–	–
	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]
The COVID-19 period
	SP500	SP500	SP500	SP500	SP500	SP500	SP500
	HS	CSI300	N225	KS	CSI300|HS	HS|KS	KS|N225
CoVaR	−3.012	−6.8592	−2.7462	−1.818	−4.9271	−1.5692	−1.4563
	(0.3502)	(1.4326)	(0.6839)	(0.1956)	(1.7564)	(0.5735)	(0.3916)
ΔCoVaR	1.1243	1.1237	1.1937	1.6838	1.8561	0.4105	0.7562
	(0.0176)	(0.1454)	(0.0754)	(0.087)	(2.3163)	(0.0742)	(0.3052)
BTKS	0.9154	0.941	0.9252	0.8731	0.7157	0.4155	0.7938
	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]	[0.0000]

Notes: Table 5 reports the mean values of CoVaR and ΔCoVaR. In the round bracket, we display their standard deviation. BTKS tests the null hypothesis that CoVaR in the intense shift is equal to it in regular circumstances. The alternative hypothesis is: CoVaR in the intense shift is greater than it in the regular circumstances. In square brackets, we display the p-value. The figure is underlined if the CoVaR and ΔCoVaR are larger in the direct or indirect dependence. The figure is in bold if the ΔCoVaR is larger in the tranquil period or the instability period.