Indeed, as emphasizes by Shiller and Perron (1985) it is not the frequency (number of observations) rather the span (number of years) of the data that is more important to test for random walk hypothesis of economic variables or causal relationships. Secondly, our monthly database which covers twelve years of data only includes four GCC countries out of six and doesn’t permit to draw any conclusion about Qatar and United Arab Emirates which are absent from the database. Consequently, given data availability, using simultaneously the two different datasets can be seen as test of robustness of our econometric results. .
