Do Central Banks Respond Timely to Developments in the Global Economy?
| Published date | 01 April 2020 |
| DOI | http://doi.org/10.1111/obes.12335 |
| Author | Sepideh Khayati Zahiri,Leif Anders Thorsrud,Hilde C. Bj⊘rnland |
| Date | 01 April 2020 |
285
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 82, 2 (2020) 0305–9049
doi: 10.1111/obes.12335
Do Central Banks Respond Timely to Developments
in the Global Economy?
Hilde C. Bjørnland, Leif Anders Thorsrud and Sepideh Khayati
Zahiri
BI Norwegian Business School, Nydalsveien 37 0484, Oslo, Norway
(e-mail: hilde.c.bjornland@bi.no, s.zahiri@gmail.com,
leif.a.thorsrud@bi.no)
Abstract
Our analysis suggests; they do not! We arrive at this conclusion by showing that revisions
to the published interest rate path projections from the central banks in New Zealand, Nor-
way and Sweden can be predicted by timely and forward-looking international indicators.
Furthermore, using individual country and Panel VARs, identified with an external instru-
ment method, we show that the policy surprises induced by the predictablerevisions likely
contain information about how the central banks assess past, current and future economic
conditions and thereby leads to a positive co-movement between the interest rate and both
financial markets and the macroeconomy.
I. Introduction
Much applied research has shown that global developments play a large role in explain-
ing business cycles and inflation in small and open economies, see, e.g. Kose, Otrok and
Whiteman (2003); Ciccarelli and Mojon (2010); Mumtaz, Simonelli and Surico (2011). At
the same time, the structural, small open-economy models used by many central banks to
analyse and predict macroeconomic outcomes cannot account for the substantial influence
of foreign-sourced disturbances identified in the numerous reduced-form studies.1Accord-
ingly, model-implied cross-correlation functions between the small open economies and
global economies are small, while data suggest that they are positive and large.
JEL Classification numbers: C11, C53, C55, E58, F17.
*We thank the editor Anindya Banerjee, three anonymous referees, Efrem Castelnuovo, Renee Fry-McKibbin,
James Morley, Adrian Pagan, Francesco Ravazzolo and seminar and conference participants at Norges Bank, the
Federal Reserve Bank of St. Louis, University of Melbourne, the Central Bank Macro Modeling Workshop at the
Reserve Bank of New Zealand and the CEF 2016Annual Meeting in Bordeaux for constructive comments and fruitful
discussions. This work is part of the research activities at the Centre forApplied Macroeconomics and commodity
Prices (CAMP) at the BI Norwegian Business School. The usual disclaimers apply.
1See Justiniano and Preston (2010) and the references therein. Recent advances in the theoretical business cycle
literature have tried to bridge this gap between the empirical findings and theory, with Bergholt and Sveen (2013)
being one example among others.
286 Bulletin
In this paper, we hypothesize that this discrepancy matters for how monetary policy is
conducted, and ultimately for how central banks make revisions to their predicted interest
rate paths, i.e. their announced policy intentions. Furthermore, if policy makers make
revisions to the interest rate path based on delayed responses to development about the
global economy, this is consistent with central banks revealing information about their
current and future view about the state of the domestic economy. After all, central banks
announce their intentions to influence (i.e. give forward guidance to) the market. If central
bank’s policy surprises include delayed responses to international developments, these
responses may also have a marked effect on the economy.
Toexamine these issues, we first construct a real-time data set of interest rate projections
from the central banks in New Zealand (Reserve Bank of New Zealand), Norway (Norges
Bank) and Sweden (Sveriges Riksbank). We focus on these three countries as they are
small and open, and because they were the first three countries adopting the practice of
communicating their policy intentions explicitly by publishing their forecasts of future
interest rates.2We then examine two questions in particular: (i) whether international
versus domestic indicators can predict the forecast revisions in the central bank’s policy
rate?, and (ii) whether fundamental variables versus forward looking variables matter?
To avoid look-ahead-biases when running the predictive regressions, we take care to use
information that was actually available to the policy makers at the time of making their
initial interest rate projections. That is, we use real-time data, that is data that are not
revised, and data that were published prior to the publication of the initial interest rate
path, and refer to them in short as timely data.
And indeed, running a battery of predictive regressions with domestic and foreign
indicators we find that the forecast revisions have a high degree of persistence, and that
there is a systematic role for international indicators in predicting the revisions to the
policy rate. Most notably is the role of forward-looking foreign indicators. In contrast,
related indexes for the domestic economy tend to be of less importance or insignificant. In
addition, at least for New Zealand and Sweden, there is a close correspondence betweenthe
information sets explaining inflation and output forecast revisions and those that explain
the interest rate forecast revisions.
While we do find forecast revisions (errors) to be predictable by foreign variables, this
is not necessarily suggesting monetary policy is inefficient or sub-optimal. In fact, there
could be many reasons for our findings, like strategical considerations (Walsh 2007; Van
der Cruijsen, Eijffinger and Hoogduin 2010), an explicit desire for forecast smoothing
(Mirkov and Natvik 2016), or inherit uncertainties related to how international develop-
ments transmit to the domestic economies going forward and noisy economic signals, like
in theories considering information rigidities (Mankiw and Reis 2002; Sims 2003).3Still,
2The Reserve Bank of New Zealand was the first central bank to publish own forecast of the interest rates in
1997, followed byNorges Bank in 2005 and Sveriges Riksbank in 2007. Since then, several other central banks have
followed, including the Central Bank of Iceland in 2007, the Czech National Bank in 2008 and most recently, the
Federal Reserve in 2012. Accordingly, the results in this paper should be of relevance for an increasing number of
central banks.
3Using data on forecasts from professional forecasters, and surveys among households, information rigidities have
been well documented empirically, see, e.g. Coibion and Gorodnichenko (2015); Dovern et al. (2015); Dr¨ager and
Lamla (2017).
©2019 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial