Table A1 Estimated threshold regressions...

Table A1

Estimated threshold regressions using the EPU index developed by Huang and Luk (2020)

Description	CQQQ			FXI			GXC			MCHI
Description	Coeff	SE	t	Coeff	SE	t	Coeff	SE	t	Coeff	SE	t
Regime 1	EPU_t-3< 175.50, n = 130			EPU_t-11 < 132.6, n = 57			EPU_t-11 < 132.7, n = 58			EPU_t-3 < 174.2, n = 127
β¹	1.300	0.181	7.17	0.447	0.238	1.88	0.569	0.233	2.44	1.064	0.176	6.04
Regime 2	EPU_t-3 ≥ 175.5 n = 22			EPU_t-11 ≥ 132.6, n = 95			EPU_t-11 ≥ 132.7, n = 94			EPU_t-3 ≥ 174.2, n = 24
β²	0.155	0.548	0.28	1.275	0.184	6.93	1.277	0.167	7.67	−0.027	0.518	−0.05
Seasonal dummies:
January	0.038	0.014	2.62	0.011	0.014	0.79	0.014	0.013	1.01	0.023	0.016	1.38
February	0.002	0.017	0.10	−0.022	0.015	−1.47	−0.010	0.014	−0.74	−0.011	0.016	−0.66
March	−0.017	0.026	−0.67	−0.018	0.018	−1.00	−0.018	0.017	−1.04	−0.006	0.018	−0.36
April	−0.014	0.017	−0.81	−0.009	0.013	−0.73	−0.008	0.012	−0.66	−0.004	0.014	−0.26
May	−0.007	0.017	−0.42	−0.017	0.010	−1.64	−0.013	0.011	−1.24	−0.014	0.011	−1.30
June	0.013	0.015	0.85	−0.018	0.012	−1.43	−0.005	0.013	−0.42	−0.002	0.011	−0.17
July	−0.033	0.022	−1.46	−0.018	0.022	−0.81	−0.019	0.021	−0.90	−0.017	0.021	−0.81
August	0.004	0.018	0.23	−0.017	0.013	−1.36	−0.015	0.013	−1.17	−0.007	0.013	−0.53
September	0.001	0.014	0.04	−0.008	0.008	−0.91	−0.006	0.009	−0.67	0.000	0.010	−0.03
October	0.005	0.018	0.26	−0.007	0.022	−0.31	−0.006	0.019	−0.30	0.002	0.017	0.12
November	0.001	0.016	0.03	0.010	0.019	0.50	0.008	0.015	0.52	0.004	0.018	0.21
December	−0.021	0.014	−1.48	−0.014	0.016	−0.89	−0.013	0.013	−0.98	−0.012	0.013	−0.90
R²	0.444			0.399			0.448			0.452
Adj. R²	0.392			0.343			0.396			0.400
DW	1.74			2.28			2.10			2.31
AIC	−2.562			−2.814			−2.974			−2.940
SIC	−2.284			−2.535			−2.696			−2.661
HQIC	−2.449			−2.701			−2.861			−2.827

Description	CQQQ			FXI			GXC			MCHI
Description	Coeff	SE	t	Coeff	SE	t	Coeff	SE	t	Coeff	SE	t
Regime 1	EPU_t-3< 175.50, n = 130			EPU_t-11 < 132.6, n = 57			EPU_t-11 < 132.7, n = 58			EPU_t-3 < 174.2, n = 127
β¹	1.300	0.181	7.17	0.447	0.238	1.88	0.569	0.233	2.44	1.064	0.176	6.04
Regime 2	EPU_t-3 ≥ 175.5 n = 22			EPU_t-11 ≥ 132.6, n = 95			EPU_t-11 ≥ 132.7, n = 94			EPU_t-3 ≥ 174.2, n = 24
β²	0.155	0.548	0.28	1.275	0.184	6.93	1.277	0.167	7.67	−0.027	0.518	−0.05
Seasonal dummies:
January	0.038	0.014	2.62	0.011	0.014	0.79	0.014	0.013	1.01	0.023	0.016	1.38
February	0.002	0.017	0.10	−0.022	0.015	−1.47	−0.010	0.014	−0.74	−0.011	0.016	−0.66
March	−0.017	0.026	−0.67	−0.018	0.018	−1.00	−0.018	0.017	−1.04	−0.006	0.018	−0.36
April	−0.014	0.017	−0.81	−0.009	0.013	−0.73	−0.008	0.012	−0.66	−0.004	0.014	−0.26
May	−0.007	0.017	−0.42	−0.017	0.010	−1.64	−0.013	0.011	−1.24	−0.014	0.011	−1.30
June	0.013	0.015	0.85	−0.018	0.012	−1.43	−0.005	0.013	−0.42	−0.002	0.011	−0.17
July	−0.033	0.022	−1.46	−0.018	0.022	−0.81	−0.019	0.021	−0.90	−0.017	0.021	−0.81
August	0.004	0.018	0.23	−0.017	0.013	−1.36	−0.015	0.013	−1.17	−0.007	0.013	−0.53
September	0.001	0.014	0.04	−0.008	0.008	−0.91	−0.006	0.009	−0.67	0.000	0.010	−0.03
October	0.005	0.018	0.26	−0.007	0.022	−0.31	−0.006	0.019	−0.30	0.002	0.017	0.12
November	0.001	0.016	0.03	0.010	0.019	0.50	0.008	0.015	0.52	0.004	0.018	0.21
December	−0.021	0.014	−1.48	−0.014	0.016	−0.89	−0.013	0.013	−0.98	−0.012	0.013	−0.90
R²	0.444			0.399			0.448			0.452
Adj. R²	0.392			0.343			0.396			0.400
DW	1.74			2.28			2.10			2.31
AIC	−2.562			−2.814			−2.974			−2.940
SIC	−2.284			−2.535			−2.696			−2.661
HQIC	−2.449			−2.701			−2.861			−2.827

Note(s): In this table we re-estimated the results reported in Table 4 using the EPU index developed by Huang and Luk (2020). A comparison between the results in Table 4 (based on Baker et al., 2016) and those in this Appendix indicates that, according to the R2, adjusted R2, AIC, SIC, and HQC criteria, the EPU measure proposed by Baker et al. (2016) employed in the present study outperforms the alternative index by Huang and Luk (2020). It may be argued that newspapers in mainland China do not operate under the same level of press freedom as those in Hong Kong. Nevertheless, the results derived from the two indices are broadly consistent for CQQQ and MMCHI (i.e. β¹>β²), but differ for FXI and GXC, as shown in Table 4 and the above table. In this study, we have chosen to rely on the Baker et al. (2016) EPU index, as it consistently produces superior model selection statistics and stronger explanatory power

Source(s): The authors’ estimation

[ViewLarge]

Sharing Unavailable