Detection and Forecasting of Islamic Calendar Effects in Time Series Data

Abstract

As the behavior of Islamic societies incorporates seasonal effects of both Gregorian and Islamic calendars, the conventional seasonal adjustment methods will yield distorted results. In this paper we extend the standard ARIMA modeling to incorporate Islamic calendar effects. This methodology is applied to the monthly data of currency in circulation. The presence of Islamic calendar effects is strongly pronounced in the results. These results are useful for policy makers of the central banks in tracking the path of currency in circulation. Furthermore, this methodology can be used for computing seasonal adjusted series and factors in Islamic countries.