Causal estimation of COVID-19 and SARS on China's stock market: evidence from a time series counterfactual prediction

Figures & data

Figure 1. Shanghai index during the COVID-19 (left) and SARS (right) epidemics.

Source: By authors.

Figure 2. Covariates during the COVID-19 (left) and SARS (right) outbreaks.

Source: By authors.

Figure 3. The time-varying causal effect of COVID-19 on the Shanghai index.

Note: (a) In the time series of the Shanghai index, the dotted line is the counterfactual forecast value, (b) pointwise (daily) incremental impact of COVID-19, (c) cumulative impact of COVID-19.

Source: Authors' calculations.

Figure 4. The time-varying causal effect of SARS on the Shanghai index.

Note: (a) In the time series of the Shanghai index, the dotted line is the counterfactual forecast value, (b) pointwise (daily) incremental impact of the SARS, (c) cumulative impact of SARS.

Source: Authors' calculations.

Table 1. The average causal effect of COVID-19 and SARS on the Shanghai index.

	COVID-19		SARS
	Average	Cumulative	Average	Cumulative
Actual	415	5391	185	2777
Prediction (s.d.)	438 (12)	5694 (153)	196 (3.5)	2935 (52.2)
95% CI	[425, 465]	[5525, 6041]	[189, 202]	[2832, 3037]
Absolute effect (s.d.)	−23 (12)	−303 (152)	−11 (3.5)	−158 (52.2)
95% CI	[−50, −10]	[−650, −134]	[−17, −3.7]	[−260, −55.7]
Relative effect (s.d.)	−5.3% (2.7%)	−5.3% (2.7%)	−5.4% (1.8%)	−5.4% (1.8%)
95% CI	[−11%, −2.4%]	[−11%, −2.4%]	[−8.9%, −1.9%]	[−8.9%, −1.9%]

Note: The ‘Average’ column talks about the average (across time) during the post-intervention period. The ‘Cumulative’ column sums up individual time points. The number of MCMC samples to draw is 5000.

Source: Authors' calculations.

Figure 5. Re-selection of covariates (COVID-19).

Note: (a) The time series of the Shanghai index, the dotted line is the counterfactual forecast value, (b) pointwise (daily) incremental impact of COVID-19, (c) cumulative impact of COVID-19.

Source: Authors' calculations.

Figure 6. Re-selection of covariates (SARS).

Note: (a) The time series of the Shanghai index, the dotted line is the counterfactual forecast value, (b) pointwise (daily) incremental impact of the SARS, (c) cumulative impact of SARS.

Source: Authors' calculations.

Table 2. Re-selection of covariates.

	COVID-19		SARS
	Average	Cumulative	Average	Cumulative
Actual	415	5391	185	2777
Prediction (s.d.)	462 (2.8)	6005 (36.9)	196 (3.4)	2938 (51.3)
95% CI	[456, 467]	[5933, 6077]	[189, 202]	[2835, 3035]
Absolute effect (s.d.)	−47 (2.8)	−614 (36.9)	−11 (3.4)	−161 (51.3)
95% CI	[−53, −42]	[−686, −542]	[−17, −3.9]	[−259, −58.1]
Relative effect (s.d.)	−10% (0.61%)	−10% (0.61%)	−5.5% (1.7%)	−5.5% (1.7%)
95% CI	[−11%, −9%]	[−11%, −9%]	[−8.8%, −2%]	[−8.8%, −2%]

Note: The ‘Average’ column talks about the average (across time) during the post-intervention period, the ‘Cumulative’ column sums up individual time points. The number of MCMC samples to draw is 5000.

Source: Authors' calculations.

Figure 7. Replacing the Shanghai index with the Shenzhen index.

Note: The impact of COVID-19 on the Shenzhen index is one the left. The impact of SARS on the Shenzhen index is on the right.

Source: Authors' calculations.

Figure 8. Re-selection of covariates to check the robustness.

Note: On the left is the impact of COVID-19 on the Shanghai index. On the right is the impact of SARS on the Shanghai index.

Source: Authors' calculations.

Figure 9. Using one-lagged-period covariates.

Note: The impact of COVID-19 on the Shanghai index is shown on the left. The impact of SARS on the Shanghai index is shown on the right.

Source: Authors' calculations.

Figure 10. Extending the sample date to February 28, 2020.

Note: (a) In the time series of the Shanghai index, the dotted line is the counterfactual forecast value, (b) pointwise (daily) incremental impact of COVID-19 and (c) cumulative impact of COVID-19.

Source: Authors' calculations.