Are Weekly Inflation Forecasts Informative?*
| Date | 01 April 2009 |
| DOI | http://doi.org/10.1111/j.1468-0084.2008.00523.x |
| Author | Andreas M. Fischer,Marlene Amstad |
| Published date | 01 April 2009 |
237
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford, 2008. Published by Blackwell Publishing Ltd,
9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 71, 2 (2009) 0305-9049
doi: 10.1111/j.1468-0084.2008.00523.x
Are Weekly Inflation Forecasts Informative?Å
Marlene Amstad† and Andreas M. Fischer‡
†Swiss National Bank, Postfach, 8022 Zurich, Switzerland
(e-mail: marlene.amstad@snb.ch)
‡Swiss National Bank and CEPR, Postfach, 8022 Zurich, Switzerland
(e-mail: andreas.fischer@snb.ch)
Abstract
Are weekly inflation forecasts informative? Although several central banks review
and discuss monetary policy issues on a bi-weekly basis, there have been no attempts
by analysts to construct systematic estimates of core inflation that supports such
a decision-making schedule. The timeliness of news releases are recognized to be
an important information source in real-time estimation. We incorporate real-time
information from macroeconomic releases and revisions into our weekly updates of
monthly Swiss core inflation using a common factor procedure. The weekly estimates
for Swiss core inflation show that it is worthwhile to update the forecast at least twice
a month.
I. Introduction
Are weekly inflation forecasts informative? Inflation forecasts play a fundamental
role in monetary frameworks that declare price stability as the primary goal of mon-
etary policy. For many central banks, price stability is defined quantitatively and the
monthly Consumer Price Index (CPI) is the measure of reference. Despite the recog-
nition that daily releases of macroeconomic information are important for real-time
forecasting, inflation forecasting is conducted mostly at the quarterly and on rare
occasions at the monthly frequency. Although several central banks, including the
European Central Bank and the Swiss National Bank, review and discuss monetary
*Thispaper was previously titled ‘Sequential Information Flow and Real-Time Diagnosis of Swiss Inflation:
Intra-Monthly DCF Estimates for a Low Inflation Environment,’ seeAmstad and Fischer (2004). The authors
would like to thank two anonymous referees for helpful comments and Domenico Giannone and Lucrezia
Reichlin for their tireless support. Tobias Gr¨assli provided valuable assistance in data support. The views
expressed here are those of the authors and do not necessarily reflect the position of the Swiss National Bank.
JEL Classification numbers: E52, E58.
238 Bulletin
policy issues on a bi-weekly basis, there have been no attempts by analysts to construct
inflation forecasts that supports such a decision-making schedule.1
The aim of this paper is to produce real-time estimates of monthly core inflation
at a frequency higher than monthly and show that intra-monthly updates improve the
information content. The concept of exploiting asynchronous data releases for updat-
ing the forecast is not new. It has been used by Evans (2005) and Giannone, Reichlin
and Small (2007) to explain information shocks to quarterly GDP nowcasts.2How-
ever, the idea of weekly inflation forecasting as a policy tool has not been previously
considered. The empirical framework applies the dynamic common factor method-
ology of Forni et al. (2000, 2005) to Swiss inflation from 1994 to 2004: a period
where Swiss (annual) inflation averaged less than 1.0%.3Our challenge is thus to
show that it is possible to construct a useful measure of core inflation for real-time
policymaking even in a low inflation environment (see Stock and Watson, 2007).
Successive releases of weekly information are found to improve the monthly now-
cast of Swiss core inflation. Empirical tests show that it is worthwhile to update the
nowcast at least twice a month. This evidence says that policymakers should make full
use of the real-time flow of information stemming from data releases and revisions
to economic series: a concept we ‘call sequential information flow’.
The paper is organized as follows. The estimation framework is presented in
section II. Next, the main features of our data set and the decisions motivating our
choices are discussed in section III. The empirical estimates of the monthly Swiss
coincident index for core inflation are presented in section IV. The same section
includes a discussion of the properties of smoothness, stability and forecasting. This
is done to show that the information from the intra-monthly updates stem from sequen-
tial information flow and not from model instability. Weekly nowcasts of monthly
core inflation are presented in section V. Final remarks on the importance of sequential
information flow are offered in section VI.
II. The estimation framework
The forecasting model relies on data reduction techniques that can handle real-time
panels updated on a weekly basis.4We follow the estimation procedures of Forni
et al. (2000), Cristadoro et al. (2005), and Altissimo et al. (2001). Below, we offer
1Independent of their scheduled press releases and monetary policy statements, the Board of the European
Central Bank and the Swiss National Bank hold regularly scheduled meetings that discuss the stance of
monetary policy on a bi-weekly basis. The Board of the Bank of Japan meets twice in April, June and
October. The Board of the Riksbank also meets twice a month on rare occasions.
2See also Altissimo et al. (2007). They consider a monthly data set to make statements about quarterly GDP
for the euro area.
3Coincident indexes based on monthly panels have established themselves as popular measures of core infla-
tion. Empirical estimates of monthly core inflation include Stock and Watson (2002) for the United States,
Cristadoro et al. (2005) for the euro area, Gosselin and Tkacz (2001) for Canada and Camacho and Sancho
(2002) for Spain.
4In a similar exercise, Amstad and Fischer (2005) consider daily updates to examine the impact of import
prices on inflation.
©Blackwell Publishing Ltd and the Department of Economics, University of Oxford 2008
To continue reading
Request your trial