U.S. flag

An official website of the United States government

Skip Header


On the Spectrum Diagnostics Used by X-12-ARIMA to Indicate the Presence of Trading Day Effects after Modeling or Adjustment

Written by:
Working Paper Number RR99-03

Introduction

Most seasonal adjustment programs for monthly and quarterly time series include a capability for estimating and removing trading day effects, repetitive effects associated with the seven days of the week. In monthly series, such effects produce periodic movements, primarily at the frequency associated with the fractional part of the number of weeks (seven day cycles) in an average month. Since spectral analysis can be used to detect the presence of periodic components, it is a natural diagnostic tool for detecting trading day effects as well as seasonal effects. The spectrum is not as powerful a diagnostic as likelihood-ratio based procedures for detecting trading day effects when a good model for the possible trading day effect and the series itself is known, based on fitted regression and ARIMA models. It is more versatile because it does not depend upon a correct model specification.

Page Last Revised - October 28, 2021
Is this page helpful?
Thumbs Up Image Yes Thumbs Down Image No
NO THANKS
255 characters maximum 255 characters maximum reached
Thank you for your feedback.
Comments or suggestions?

Top

Back to Header