Retail Trade: Predictor Of Change In Total GDP?

Reports of retail trade in Spain, Sweden and Hong Kong were among today's statistical releases.  Retail sales are sometimes good indicators of trends in total GDP.  In order to test this hypothesis, we computed the correlation coefficients between the year over year percent change in the quarterly values of the value index of retail sales and the year over year percent change in the value of total GDP over the past ten years*.  The first three charts show these correlations.

The highest correlation was in Spain where the correlation coefficient was .94.  Hong Kong was next at .87 and Sweden was the lowest at .64.  The correlation coefficient squared, R², is a measure of the explanatory power of one variable on the other when all other factors are held constant.  Thus, in the case of Spain, based on the experience of the last ten years, the year to year change in retail sales explains about 88.4% of the variation in total GDP when all other factors are held constant.  In the case of Hong Kong, the year to year change retail sales explains 75.7% of the change in total GDP and in the case of Sweden, only 41%. The precise values are less important than the order of magnitude.  Thus, we should expect the trend in retail sales in Spain to have a greater impact on the trend in total GDP than that in Hong Kong or Sweden.

The fourth chart shows the year over year changes in monthly retail sales in the three countries. On the basis of the past relationship between retail sales and total GDP the year to year changes in the value of retail sales in October and November of 9.9% and 11.7% for Hong Kong suggest there is a good chance that the year to year change in the fourth quarter total GDP will increase from -1.99% in the third quarter to perhaps a positive value. The continued negative, though diminishing, yearly comparisons of retail sales in Spain suggest that any improvements in year to year comparison for total GDP in the fourth quarter is likely to be small. For Sweden, the year to year increases in retail sales of 5.8% and 5.7% in October and November may be large enough to have a positive effect on the increase in total GDP in spite of the fact that the yearly change in retail sales has less than a major effect on the change in total GDP. *(Computation of the correlation coefficient can be done in DLXVG3 by clicking on "Correlation" in the TOOLS menu.)