Table 3 Estimates of global herding around the market average using the baseline models.

(1) CSAD unconditional				(2) CSAD conditional on up and down				(3) CSAD conditional on US regimes
Variable	Parameter	Est. Coeff	T-Stat	Variable	Parameter	Est. Coeff	T-Stat	Variable	Parameter	Est. Coeff	T-Stat
Constant	α	1.514	0.000	Constant	α	1.481**	21.496	Constant	α	1.632**	18.861
\|r_g\|	β	−0.096	0.354	\|r_g\|D_d	β_d	1.796**	3.808	\|r_g\|A₁	β_g1	−0.142	−0.845
R_g²	γ	0.076	0.008	\|r_g\|D_u	β_u	−0.017	−0.183	\|r_g\|A₂	β_g2	−0.138	−1.112
				R_g²D_d	γ_gd	−0.436**	−2.647	R_g²A₁	γ_g1	0.179**	2.925
				R_g²D_u	γ_gu	0.106**	3.938	R_g²A₂	γ_g2	0.104**	3.381
				R_US²D_d	γ_usd	−0.087*	−2.295	R_US²A₁	γ_us1	−0.135**	−5.248
				R_US²D_u	γ_usu	−0.066**	−5.618	R_US²A₂	γ_us2	−0.037*	−2.280
Adj R²		0.187		Adj R²		0.399		Adj R²		0.302
χ² test for exclusion of global drivers				γ_gd= γ_gu= 0		27.16**		γ_g1= γ_g2= 0		11.22**
χ² test for exclusion of US drivers				γ_usd= γ_usu= 0		36.85**		γ_us1= γ_us2= 0		33.03**
F test for symmetric herding		by global driver		γ_gd= γ_gu		11.22**		γ_g1= γ_g2		1.61
F test for symmetric herding		by US driver		γ_usd= γ_usu		0.29		γ_us1= γ_us2		10.09**

T-Stat represents t-statistics based on HAC standard errors. Coefficients estimates in boldface indicate the presence of significant herding at conventional significance levels. **, *, and + denote significance at 1%, 5%, and 10%, respectively.

Quick links

Search