Synopses & Reviews
This book outlines and demonstrates problems with the use of the HP filter, and proposes an alternative strategy for inferring cyclical behavior from a time series featuring seasonal, trend, cyclical and noise components. The main innovation of the alternative strategy involves augmenting the series forecasts and back-casts obtained from an ARIMA model, and then applying the HP filter to the augmented series. Comparisons presented using artificial and actual data demonstrate the superiority of the alternative strategy.
Review
From the reviews: MATHEMATICAL REVIEWS "Altogether this book is more on the mathematical side, it is well written following the same idea throughout and contains many exercises which complete the different topics. The text concentrates on the approach of the authors...I enjoyed reading this nicely written book which can certainly be recommended to all mathematically oriented statisticians interested in the subject."
Synopsis
lengths, that could not be captured with univariate linear filters. Exam- ples of research in both directions can be found in Sims (1977), Lahiri and Moore (1991), Stock and Watson (1993), and Hamilton (1994) and (1989). Although the first approach is known to present serious limitations, the new and more sophisticated methods developed in the second approach (most notably, multivariate and nonlinear extensions) are at an early stage, and have proved still unreliable, displaying poor behavior when moving away from the sample period . Despite the fact that business cycle estimation is basic to the conduct of macroeconomic policy and to monitoring of the economy, many decades of attention have shown that formal modeling of economic cycles is a frustrating issue. As Baxter and King (1999) point out, we still face at present the same basic question "as did Burns and Mitchell fifty years ago: how should one isolate the cyclical component of an eco- nomic time series? In particular, how should one separate business-cycle elements from slowly evolving secular trends, and rapidly varying seasonal or irregular components?" Be that as it may, it is a fact that measuring (in some way) the busi- ness cycle is an actual pressing need of economists, in particular of those related to the functioning of policy-making agencies and institutions, and of applied macroeconomic research.
Synopsis
This book will be useful to economists and analysists in government and financial/commercial companies who routinely monitor the state of the economic cycle, and who produce short-term forecasts. It will also be of interest to academics who do business cycle research, and who need to extract a measure of the cycle from the data.
Table of Contents
Introduction and Brief Summary.- A Brief Review of Applied Time Series Analysis.- ARIMA Models and Signal Extraction.- Detrending and the Hodrick-Prescott Filter.- Some Basic Limitations of the Hodrick-Prescott Filter.- Improving the Hodrick-Prescott Filter.- Hodrick-Prescott Filtering Within a Model Based Approach.