Table 1.

Summary statistics of returns.

	Min	Max	Mean	Std. Dev.	Skewness	Kurtosis
Before COVID-19
OIL	−6.449	11.070	0.011	1.943	0.071	5.576
BTC	−23.874	22.512	0.004	4.875	0.052	6.664
USDT	−4.820	5.661	−0.000	0.612	0.724	24.211
USD	−1.015	1.125	0.004	0.331	−0.011	3.183
During COVID-19
OIL	−64.370	41.202	0.069	4.953	−2.787	59.050
BTC	−46.473	19.153	0.175	4.869	−1.644	17.654
USDT	−5.257	5.339	−0.001	0.399	0.484	105.144
USD	−1.697	1.589	0.015	0.414	0.173	4.420
	JB	ADF	PP	LB-$Q$(20)	LB-$Q^2$(20)	ARCH-LM(20)
Before COVID-19
OIL	155.825***	−24.010***	−23.996***	22.677	52.083***	63.200***
BTC	313.773***	−22.620***	−22.692***	13.922	92.898***	66.500***
USDT	10,504.237***	−32.500***	−39.508***	113.750***	159.180***	42.680***
USD	0.883	−23.900***	−23.922***	28.688*	29.407*	23.600
During COVID-19
OIL	87,816.382***	−26.650***	−26.722***	71.732***	191.050***	99.300***
BTC	6245.490***	−27.240***	−27.310***	39.497***	12.531	354.000***
USDT	288,800.994***	−39.200***	−57.031***	191.710***	180.440***	75.900***
USD	59.798***	−23.720***	−23.856***	29.895*	461.780***	25.300

Note: JB stands for the Jarque–Bera test statistic of normality, while ADF and PP, respectively, refer to the augmented Dickey–Fuller and Phillips–Perron test statistics of a unit root. LB-$Q$(20) and LB-$Q^2$(20) are the Ljung–Box statistics for testing serial correlation in the return and squared return series, respectively, while ARCH-LM(20) stands for the Lagrange multiplier statistic utilized to test conditional heteroscedasticity or ARCH effects up to lag 20. The asterisks * and *** indicate statistical significance at the 10% and 1% levels, respectively.