Table 4.

Investor behavior on financial contagion under different market conditions.

Dep. Variable	Contagion${}_t$
Variable	Model (1)	Model (2)	Model (3)	Model (4)	Model (5)
Panel A. In the market melt-down period
Constant	0.177${}^{\ast \ast \ast }$	0.444${}^{\ast \ast \ast }$	0.233${}^{\ast \ast \ast }$	2.026${}^{\ast \ast \ast }$	2.026${}^{\ast \ast \ast }$
	(15.93)	(15.56)	(19.36)	(19.09)	(16.16)
Contagion${}_{t-1}$	0.935${}^{\ast \ast \ast }$	0.937${}^{\ast \ast \ast }$	0.910${}^{\ast \ast \ast }$	0.910${}^{\ast \ast \ast }$	0.652${}^{\ast \ast \ast }$
	(225.33)	(227.41)	(199.86)	(195.17)	(30.45)
Country attention${}_t$	−0.029${}^{\ast \ast \ast }$
	(−4.44)
Local attention${}_t$		−0.020${}^{\ast \ast \ast }$
		(−3.01)
Global attention${}_t$			−0.078${}^{\ast \ast \ast }$
			(−13.54)
Sentiment${}_t$				−0.037${}^{\ast \ast \ast }$
				(−12.26)
Fear${}_t$					−0.003${}^{\ast \ast \ast }$
					(−7.97)
Obs.	7566	7566	7566	7566	1092
Adj R-squared	0.874	0.874	0.877	0.877	0.592
Panel B. In the market melt-up period
Constant	0.222${}^{\ast \ast \ast }$	0.219${}^{\ast \ast \ast }$	0.302${}^{\ast \ast \ast }$	0.220${}^{\ast \ast \ast }$	1.367${}^{\ast \ast \ast }$
	(34.14)	(33.85)	(40.43)	(33.91)	(69.91)
Contagion${}_{t-1}$	0.886${}^{\ast \ast \ast }$	0.888${}^{\ast \ast \ast }$	0.844${}^{\ast \ast \ast }$	0.887${}^{\ast \ast \ast }$	0.402${}^{\ast \ast \ast }$
	(265.15)	(267.41)	(218.46)	(266.31)	(49.36)
Country attention${}_t$	−0.009${}^{\ast \ast \ast }$
	(−4.44)
Local attention${}_t$		−0.002
		(−1.06)
Global attention${}_t$			−0.020${}^{\ast \ast \ast }$
			(−21.50)
Sentiment${}_t$				−0.004${}^{\ast \ast \ast }$
				(−2.28)
Fear${}_t$					−0.0005${}^{\ast \ast \ast }$
					(−6.41)
Obs.	19 110	19 110	19 110	19 110	12 636
Adj R-squared	0.789	0.789	0.794	0.789	0.167

Notes: This table reports the results of the following regression during the market melt-down and melt-up periods: $Contagio{n}_{i,t}=Constant+\beta Contagio{n}_{i,t-1}+\gamma Behavio{r}_{i,t}+{\mu }_i+{\mu }_t+{\varepsilon }_{i,t}$, where $Contagio{n}_{i,t}$ refers the pandemic-driven contagion of country $i$ at time $t$ and is defined as the sum of the tail dependence between country $i$ and its contagious countries. $Behavio{r}_{i,t}$ corresponds to the proxy variables of investor behavior including investor attention (country attention, local attention, and global attention), investor sentiment, and investor fear. ${\mu }_i$ and ${\mu }_t$ are country and daily fixed effects, respectively. $t$-statistic is reported in parentheses. ${}^{\ast \ast \ast }$ indicates significance at the 1% level. Country and daily fixed effects are included in the regression. The sample period for investor attention and sentiment is from January 1, 2019 to March 27, 2022, while for investor fear is from January 13, 2020 to March 27, 2022.