William R. Bell

New Capabilities and Methods of the X-12-ARIMA Seasonal-Adjustment Program

X-12-ARIMA is the Census Bureaus new seasonal-adjustment program. It provides four types of enhancements to X-ll-ARIMA—(1) alternative seasonal, trading-day, and holiday effect adjustment capabilities that include adjustments for effects estimated with user-defined regressors; additional seasonal and trend filter options; and an alternative seasonal-trend-irregular decomposition; (2) new diagnostics of the quality and stability of the adjustments achieved under the options selected; (3) extensive time series modeling and model-selection capabilities for linear regression models with ARIMA errors, with optional robust estimation of coefficients; (4) a new user interface with features to facilitate batch processing large numbers of series.

Issues Involved With the Seasonal Adjustment of Economic Time Series

In the first part of this article, we briefly review the history of seasonal adjustment and statistical time series analysis in order to understand why seasonal adjustment methods have evolved into their present form. This review provides insight into some of the problems that must be addressed by seasonal adjustment procedures and points out that advances in modern time series analysis raise the question of whether seasonal adjustment should be performed at all. This in turn leads to a discussion in the second part of issues invloved in seasonal adjustment. We state our opinions about the issues raised and renew some of the work of our authors. First, we comment on reasons that have been given for doing seasonal adjustment and suggest a new possible justification. We then emphasize the need to define precisely the seasonal and nonseasonal components and offer our definitions. Finally, we discuss our criteria for evaluating seasonal adjustments. We contend that proposed criteria based on empirical comparisons of estimated components are of little value and suggest that seasonal adjustment methods should be evaluated based on whether they are consistent with the information in the observed data. This idea is illustrated with an example.

A Note on Overdifferencing and the Equivalence of Seasonal Time Series Models With Monthly Means and Models With (0, 1, 1)12 Seasonal Parts When ⊖ = 1

Two general models for monthly seasonal time series are considered, one in which seasonality is modeled with monthly means and another in which seasonality is modeled with a (0, 1, 1)12 ARIMA structure. The models are shown to be equivalent if the seasonal moving average parameter (⊖) is 1 and if the same assumptions about the 12 initial observations are made for both models. The role of the assumptions about the initial observations is analyzed, and it is argued that for practical purposes the two models can be regarded as equivalent when ⊖ = 1. It is observed that the result extends easily to more general models involving overdifferencing.