Let's understand the Stata code to run the autocorrelation test first. To find autocorrelation, we draw correlograms. The command to draw correlograms is corrgram, for example, take a look at the following:

corrgram fridge_sales, lags(12)

You will get the following output when you run this command:

To explore the different relationships between two given time series, you can leverage the xcorr command.

The following graph depicts the existing correlation between the fridge_sales quarterly growth rate and volume. When you use the xcorr command, you need to type all the independent variables first, followed by the dependent variable:

xcorr fridge_sales volume, lags(10) xlabel(-10(1)10,grid)

At lag 0, you can see that there is a nicely pointed out negative correlation between fridge_sales and the volume of the given model. This implies that the drop in volume causes a continuous and immediate increase in fridge_sales, as shown in the following table:

Now...