Table 4.

Results from a VAR model. This table reports results from a VAR model in which stock price returns and exchange rate returns are endogenous while all control variables are exogenous including the contemporaneous exchange rate returns, $\beta$. Two lags are used in the VAR model based on the schwarz information criterion. We only report the null hypothesis that $\beta =0$—both the slope coefficient and the t-statistic are reported for the three sample periods, namely the COVID-19 sample (31/12/2019 to 17/8/2020); the pre-COVID-19 Sample A (04/1/2010 to 30/12/2019); and the pre-COVID-19 Sample B (31/12/2018 to 16/8/2019). A total of three models are estimated: (a) model with no controls, (b) model with day-of-the-week (DOW) controls only, and (c) model with both the DOW and oil price growth (GOIL) controls.

Panel A: Post-COVID-19 period
	No controls	DOW controls	DOW$+$GOIL controls
$\beta =0$	1.1092⁎⁎⁎ (4.7385)	1.0356⁎⁎⁎ (4.6026)	1.1068⁎⁎⁎ (4.8141)
$R^2$	16.5%	24.6%	20.4%
Panel B: Pre-COVID-19 Sample A
	No controls	DOW controls	DOW$+$GOIL controls
$\beta =0$	1.2532⁎⁎⁎ (33.8591)	1.2297⁎⁎⁎ (33.2836)	1.2538 (33.8447)
$R^2$	31.28%	32.44%	31.32%
Panel A: Pre-COVID-19 Sample B
	No controls	DOW controls	DOW$+$GOIL controls
$\beta =0$	1.2492⁎⁎⁎ (9.1889)	1.2031⁎⁎⁎ (8.6716)	1.2414⁎⁎⁎ (9.1407)
$R^2$	34.52%	36.98%	35.89%

^***Denotes statistical significance at the 1% level.