What i have understood from your discussion is somewhat like...When you will be doing your regression, put the dummy variables along with others.

If this is not what you meant to say you can rewrite it... Hope this helps

Yes, you can use a dummy. But I think there will be a structural break and you should perform cointegration and causality analysis in the presence of structural break.

I think that there are some problems with your approach to this

problem. There are two possible approaches, a time series approach

and 2. a structural approach.

In the time-series approach, you estimate a good forecasting model.

This can be multivariate (i,e, VAR, VECM, etc.) or univariate

(i.e. ARIMA, etc.}. You may need to estimate several models using out-of-sample forecasting performance. It is not possible to say in

advance what kind of model is best in these circumstances. You must

evaluate several models.

You can estimate such a model up to the pandemic and then forecast

through the pandemic. If the forecasts for the pandemic area are

significantly different from the outcome you may conclude that changes

in the variables involved (GDP, etc) coincided with the pandemic. You

may not conclude that the pandemic caused these changes or use these

changes as a measure of the effect of the pandemic.

An alternative method would be to include a dummy for the epidemic in

your foresting model and to measure the significance of the dummy.

I suspect that this methodology may not lead to satisfactory results

as GDP, imports and exports are impacted by many other factors other

than the pandemic and that it will be very difficult to find a time-series model with clean residuals. (I presume that in your analysis you are

using logs of volume figures for each series)

If you wish to establish causal links you must use your economic

theory to specify some form of a structural model that specifies the

causal links that you wish to test and estimate. Looking at your

series all three are simultaneous in that imports and exports are

important determinants of GDP, GDP and imports are important determinants of

exports, and GDP and exports are important determinants of imports. This simultaneity implies endogeneity which you will need to take into account when you are estimating your system. With three endogenous variables, you may also need to add exogenous variables to aid identification.

Causality established with such a model is conditional on your model. Another researcher may have another model that is also data consistent but has different causal links. Choosing between such a model is a matter for the economic evaluation of both models. This evaluation may be aided by econometric analysis but the main input is the economic one.

I am not sure how you intend to apply "Granger Causality" to your problem. The procedure that I have outlined for a forecasting model is in the spirit of this methodology apart from the possible problem of missing variables

It is not clear, what you want to show by your analysis. Do you try to find an explanation for GDP (for forecasting or?) and trying to get rid of the disturbing effect of Covid or is the main goal to estimate the Covid-effect on the economy? For me it seems that you are not interested in an economic problem, but only want to get some exercise in econometric methods and programs. Your "model" (i.e. equation to estimate) seems to be

GDP = Function(Exports, Imports, Covid19-Dummy), where, I suppose, that Function is linear, but the variables are transformed (ln, differences, exp etc.?). Whatever these transformations, this is a rather strange model. It is neither demand nor supply side, and I cannot see any other theoretical base.

As GDP is - by definition - the sum of internal demand (consumption, investment, change of stocks)+exports-imports, there is a severe endogeneity problem (see John's contribution) - especially as imports can be assumed as highly dependent on GDP (via income and consumption or because they are inputs for the domestic production).