Skip to main content

Table 8 Estimation results of the benchmark models during Bitcoin crashes

From: The linkage between Bitcoin and foreign exchanges in developed and emerging markets

 

G7 basket

BRICS basket

 

2013 crash

2018 crash

COVID-19

2021 crash

2022 crash

2013 crash

2018 crash

COVID-19

2021 crash

2022 crash

 

Panel A: multivariate \(t_\nu\) copula

dof

4.91

4.92

4.52

4.15

3.67

7.38

6.09

5.76

4.96

5.01

Log-likelihood

112.20

119.80

108.50

180.20

220.80

80.33

179.60

175.70

88.85

82.71

AIC

− 202.46

− 217.70

− 195.03

− 338.47

− 419.67

− 128.66

− 363.16

− 319.38

− 145.71

− 133.41

BIC

− 163.99

− 175.77

− 154.51

− 305.96

− 383.44

− 72.70

− 302.18

− 260.44

− 98.41

− 80.72

 

Panel B: DCC model

Log-likelihood

73.98

88.62

− 22.05

146.76

41.42

− 132.91

− 9.04

− 62.07

25.71

17.20

AIC

− 141.96

− 171.24

50.10

− 287.51

− 76.84

271.82

24.09

130.14

− 45.41

− 28.40

BIC

− 131.47

− 159.80

61.15

− 278.65

− 66.98

282.31

35.52

141.19

− 36.54

− 18.53

  1. This table reports the selection criteria of the multivariate \(t_\nu\) copula (Panel A) and DCC model (Panel B) as benchmarks during the Bitcoin crashes. The G7 basket includes Bitcoin and G7 currency returns, and the BRICS basket includes Bitcoin and BRICS currency returns. The reported log-likelihoods refer to the dependence structures excluding the marginals