The X-II-ARIMA seasonal adjustment method

Written by:

Introduction

The majority of seasonal adjustment methods so far developed are based on univariate time series models. They are selected mainly for their simplicity and can be applied without specialized knowledge in a subject matter field.

A few attempts have been made to estimate seasonals based on causal explanations but none of them reached further than the experimental stage. Mendershausen (1939), for instance, tried to regress the seasonal for each month on a set of exogenous variables (meteorological and social variates) in order to build an explanatory model for seasonality but his empirical results were inconclusive.

Page Last Revised - October 8, 2021