Title

Constructing a Coincident Index of Business Cycles without Assuming a One-Factor Model

Publication Type

Conference Paper

Publication Date

8-2004

Abstract

The Stock–Watson coincident index and its subsequent extensions assume a static linear one-factor structure for the component indicators. Such assumption is restrictive in practice, however, with as few as four indicators. In fact, such assumption is unnecessary if one defines a coincident index as an estimate of latent monthly real GDP. This paper considers VAR and factor models for latent monthly real GDP and other coincident indicators, and estimates the models using the observable mixed-frequency series. For US data, Schwartz’s Bayesian information criterion selects a two-factor model. The smoothed estimate of latent monthly real GDP is the proposed index.

Discipline

Econometrics

Research Areas

Econometrics

Publication

Econometrics Programme on Forecasting and High Frequency Data Analysis (International Conference)