Financial Innovation

Table 3 Cascade-correlation network results

From: Volatility contagion between cryptocurrencies, gold and stock markets pre-and-during COVID-19: evidence using DCC-GARCH and cascade-correlation network

	Panel A: The entire sample							Panel B: Pre-COVID-19							Panel C: During-COVID-19
	XAU	JPN	SPX	UK	CHN	DEU	CAC	XAU	JPN	SPX	UK	CHN	DEU	CAC	XAU	JPN	SPX	UK	CHN	DEU	CAC
Model parameters
Input layer
Number of neurons	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Hidden layer
Number of neurons	25	26	27	16	30	22	15	19	20	24	18	18	23	18	10	7	8	6	8	11	6
Sigmoid neurons	3	5	4	3	2	4	1	4	2	6	5	3	3	2	3	0	1	1	3	1	1
Gaussian neurons	22	21	23	13	28	18	14	15	18	18	13	15	20	16	7	7	7	5	5	10	5
Minimum weight	− 9893	− 3426	− 929	− 3346	− 512	− 3863	− 3567	− 1225	− 1137	− 484	− 5702	− 692	− 1292	− 30	− 134	− 37	− 20	− 2505	− 781	− 175	− 288
Maximum weight	437	660	118	34	152	114	9.247	97,453	11,767	1151	394	1298	23,876	1345	1215	18	268	11	247	48	22
Output layer
Number of neurons	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1	1
Minimum weight	− 0.395	− 0.330	− 0.333	− 0.265	− 0.299	− 0.282	− 0.312	− 0.208	− 0.403	− 0.196	− 0.424	− 0.264	− 0.407	− 0.154	− 0.400	− 0.445	− 0.066	− 0.259	− 0.174	− 0.427	− 0.075
Maximum weight	0.62	0.414	0.444	0.666	0.281	0.393	0.598	0.381	0.473	0.381	0.385	0.394	0.459	0.574	0.612	0.581	0.523	0.514	0.35	0.424	0.666
Analysis of variance
R²	0.836	0.969	0.970	0.674	0.891	0.943	0.887	0.863	0.957	0.955	0.890	0.792	0.940	0.924	0.712	0.935	0.900	0.840	0.611	0.86	0.855
CV	0.068	0.056	0.065	0.055	0.092	0.059	0.067	0.049	0.059	0.057	0.032	0.104	0.054	0.047	0.026	0.032	0.049	0.036	0.077	0.044	0.056
NMSE	0.163	0.030	0.029	0.325	0.108	0.056	0.112	0.136	0.042	0.044	0.109	0.207	0.059	0.075	0.287	0.064	0.099	0.159	0.488	0.13	0.144
MAPE	4.887	4.436	4.828	4.042	7.420	4.494	5.478	3.631	4.874	4.243	2.568	8.741	4.317	3.702	2.104	2.529	3.601	2.873	5.419	3.932	4.260

Our data consists of 2,958 daily observations from January 2, 2011, to June 2, 2022, and divided into two sub-samples. The first sub-sample covers the period from January 2, 2011, to February 23, 2020, which is pre-COVID-19. The second sub-sample covers the period from February 24, 2020, to June 2, 2022, which is during COVID-19. This Table shows Cascade-Correlation, network models. This Table shows that gold (XAU) and the Japanese stock market (JPN) have the lowest MAPE values of 2.104 and 2.529, respectively which indicate a short-term volatility contagion between Bitcoin and both gold and the stock market of Japan

Back to article page