Resumen de Adapting seasonal adjustment procedures in uncertain times

Jorge Aramendi Rique, Haritz Olaeta Goiriena

• The abrupt COVID-19 outbreak in March 2020 implied an unquestionable need for changing the strategy of calendar and seasonal adjustment in short term business statistics in official statistical offices. Constant revision of models, treatment of outliers and the construction of intervention variables for concurrent periods have been a major challenge in 2020. In this work we describe the main unexpected issues we have faced in the past months in some of our statistics to provide the most accurate adjusted estimation of the monthly activity using JDemetra+. The derived shape of the impact of the COVID-19 is of special relevance in some of the estimated series. It is time to review all the intervention variables and specified outliers in the last year and to redefine a new calendar and seasonal adjustment strategy.