Table 6.

Determinants of net directional connectedness.

$${NC}_{i,t}={\alpha }_0+{\alpha }_1\bullet \left({TA}_{i,t}-{TA}_t^{US}\right)+{\alpha }_2\bullet {TR}_{i,t}+{\alpha }_3\bullet {VIX}_{i,t}+{\alpha }_4\bullet {Illiquidity}_{i,t}+{\alpha }_5\bullet {Volatility}_{i,t}$$

$$+{\alpha }_6\bullet {COVID\_\mathit{19}}_{i,t}+{\alpha }_7\bullet {New\_case}_{i,t}+{\alpha }_8\bullet Ne{w\_case\_per\_million}_{i,t}+{\alpha }_9\bullet {Yield}_t^{High}$$

$$+{\alpha }_{10}\bullet {Yield}_t^{Low}+{\alpha }_{11}\left(Ne{w\_case\_per\_million}_{i,t}\bullet {Yield}_t^{High}\right)$$

$$+{\alpha }_{12}\left(Ne{w\_case\_per\_million}_{i,t}\bullet {Yield}_t^{Low}\right)+e_{i,t}$$

	(1)	(2)	(3)	(4)	(5)	(6)
Constant	−1.29 (4.85)	−1.50 (2.92)	−9.58* (5.16)	−8.15 (6.19)	−30.88*** (10.20)	−35.12*** (10.55)
TAi,t－TAUSt	0.11*** (0.01)	0.11*** (0.02)	0.11*** (0.02)	0.11*** (0.02)	0.11*** (0.02)	0.11*** (0.02)
TRi,t	−0.15 (0.44)	−1.33 (1.12)	−1.47 (1.90)	−1.48 (2.23)	−4.90*** (1.33)	−5.81*** (1.30)
VIXi,t	0.06 (0.05)	0.09*** (0.03)	0.07* (0.04)	0.08** (0.04)	0.10** (0.05)	0.10** (0.05)
Illiquidityi,t	2.29*** (0.90)	2.32*** (0.84)	2.16** (1.07)	2.03** (2.00)	2.53** (1.02)	2.79** (1.30)
Volatilityi,t	1.70*** (0.10)	1.81*** (0.23)	1.58*** (0.10)	1.60*** (0.10)	1.38*** (0.09)	1.29*** (0.11)
COVID_19i,t		2.24* (1.20)
New_casei,t			0.86*** (0.05)
New_case_per_milloni,t				0.12*** (0.03)	0.18*** (0.03)	0.04*** (0.01)
$Yiel{d}_t^{High}$					7.16*** (0.62)	6.99*** (0.56)
$Yiel{d}_t^{Low}$					−1.79** (0.88)	−1.86** (0.91)
$({\text{New}\_\text{case}\_\text{per}\_\text{millon}}_{\mathrm{i},\mathrm{t}}\cdot Yiel{d}_t^{High})$						0.51*** (0.12)
$({\text{New}\_\text{case}\_\text{per}\_\text{millon}}_{\mathrm{i},\mathrm{t}}\cdot Yiel{d}_t^{Low})$						−0.07 (0.19)
Fixed year effects	Yes	NO	NO	NO	NO	NO
Fixed currency effects	Yes	Yes	Yes	Yes	NO	NO
A. R2	0.60	0.61	0.61	0.61	0.25	0.25

Note: This table reports the estimation of the linear models for net directional connectedness from i exchange rate (${NC}_{i,t}$).

where ${TA}_{i,t}-{TA}_t^{US}$ represents the difference in total assets scales in balance sheets between country i and the U.S. (divided by 10⁷). ${TR}_{i,t}$ represents the federal funds target rate (upper bound). VIX_t denotes the CBOE volatility index. Illiquidity_i,t is the percentage bid-ask spread for exchange rate i (multiplied by 10²), which is calculated by dividing the bid-ask spread by the midpoint. Volatility_i,t denotes one-month option's implied volatility for exchange rate i. COVID_19 is a dummy equal to one during the period from January 23, 2020, to June 17, 2021, and zero otherwise. To analyze the impact of COVID-19 infection cases on net directional connectedness, we adopt the shorter sample period from January 23, 2020, to June 17, 2021, in models (3)–(6). New_case_i,t denotes the number of new COVID-19 infection cases of country i relative to that of the U.S. (divided by 10⁴). New_case_per_million_i,t denotes the new COVID-19 infection cases per one million population of country i relative to that of the U.S. ${Yield}^{High}$ represents the dummy variable that equals one if the currency i belongs to the high-yield group and zero otherwise, and ${Yield}^{Low}$ represents the dummy variable that equals one if the currency i belongs to the low-yield group and zero otherwise. Robust standard errors (Newey and West, 1987) are in parentheses. ***, **, and * indicate significance at the 1%, 5%, and 10% levels, respectively.where ${TA}_t^{US}$ represents the total assets in the Fed's balance sheet (divided by 10⁶). VIX_t denotes the CBOE volatility index. FXCTG10_t represents the difference in Bloomberg Carry Trade Index and denotes the carry trade returns computed from G10 currency baskets. ∗∗∗, ∗∗, and ∗ indicate significance at the 1%, 5%, and 10% levels, respectively