EN
The aim of this paper is to construct a parametric method in a Bayesian framework to identify trading-day frequency for monthly data. The well-known visual spectral test (implemented, for example, in X-12-ARIMA) is a popular tool in the literature. In the article’s proposed method, the assumption concerning the almost periodicity of the mean function plays a central role. We use a set of frequencies that corresponds to the trading-day effect for monthly data. As an illustration, we examine this effect in production in industry in European economies for data adjusted by working days and for gross data.