Norges Bank Output Gap Estimates: Forecasting Properties, Reliability, Cyclical Sensitivity and Hysteresis

Published date01 February 2023
AuthorFrancesco Furlanetto,Kåre Hagelund,Frank Hansen,Ørjan Robstad
Date01 February 2023
DOIhttp://doi.org/10.1111/obes.12524
OXFORD BULLETIN OF ECONOMICS AND STATISTICS, 85, 1 (2023) 0305-9049
doi: 10.1111/obes.12524
Norges Bank Output Gap Estimates: Forecasting
Properties, Reliability, Cyclical Sensitivity
and Hysteresis
FRANCESCO FURLANETTO,†,‡ K˚
ARE HAGELUND,† FRANK HANSEN†,§
and ØRJAN ROBSTAD†
†Norges Bank, Bankplassen 2, PB 1179 Sentrum Oslo, 0107, Norway (e-mail:
francesco.furlanetto@norges-bank.no)
‡BI Norwegian Business School, Nydalsveien 37, Oslo, 0484, Norway (e-mail:
kare.hagelund@norges-bank.no)
§Gard AS, Kittelsbuktveien 31, Arendal, 4836, Norway (e-mail: frank.hansen@gard.no)
Abstract
This paper documents the suite of models (SoMs) used by Norges Bank to estimate the
output gap. The models are estimated using data on GDP, unemployment, inflation, wages,
investment, house prices and credit. We evaluate the estimated output gap series in terms
of its forecasting properties, its reliability and its cyclical sensitivity to various measures
of demand and supply shocks. A simple equally weighted average of estimates from
different models features a better forecasting performance than each individual model. In
addition, it helps predicting inflation in pseudo real time and exhibits limited variations
when new data become available. The summary measure of potential output responds
strongly and rapidly to permanent shocks and to narrative measures of technology shocks
but, although to a more limited extent, also to demand shocks, thus partly capturing
hysteresis effects.
I. Introduction
The output gap, defined as the difference between actual output and its potential level, is
a key variable for central banks, in particular for those operating in a flexible inflation
targeting regime, such as the Central Bank of Norway (Norges Bank henceforth). In fact,
having an accurate measure of the state of the business cycle is crucial to appropriately
evaluate trade-offs between different objectives and to fulfil the central bank’s mandate.
JEL Classification numbers: C38, E17, E32.
This paper should not be reported as representing the views of Norges Bank. The views expressed are those of the
authors and do not necessarily reflect those of Norges Bank. We would like to thank the editor Francesco Zanetti,
two anonymous referees and one anonymous referee for the Norges Bank working paper series for constructive
comments. We also thank Knut Are Aastveit, Ida Wolden Bache, Paul Beaudry, Julien Champagne, Tord Krogh,
Per Espen Lille˚
as, Elmar Mertens, Alexander Rathke, Samad Sarferaz, Rodrigo Sekkel, Ingrid Solberg and
seminar participants at KOF ETH in Zurich, CEF 2019 Conference in Ottawa, SNDE 2019 conference at Dallas
Fed, CIRET/KOF/OECD/INSEE 2019 Workshop in Paris and Forecasting at Central Banks 2019 Conference at
Bank of Canada.
238
©2022 The Authors. Oxford Bulletin of Economics and Statistics published by Oxford University and John Wiley & Sons Ltd.
This is an open access article under the terms of the Creative Commons Attribution-NonCommercial-NoDerivs License, which permits use and
distribution in any medium, provided the original work is properly cited, the use is non-commercial and no modifications or adaptations are made.
Estimation of output gap in Norges Bank 239
Norges Bank bases its assessment of the output gap on a broad set of indicators and models
that are revised and expanded over time. The objective of this paper is to document the
suite of models (SoMs) used as a key input to produce output gap estimates at Norges
Bank. In particular, we evaluate its forecasting performance and its reliability when new
data become available. In addition, we also provide a novel analysis on the cyclical
sensitivity of the output gap to various measures of supply and demand shocks identified
with different methods, along the lines of Coibion, Gorodnichenko and Ulate (2018).
The output gap is not observable and its measurement is obviously challenging.
Traditionally, simple univariate statistical filters, such as the Hodrick– Prescott (HP) filter,
have been prominent. With sufficient information about the future, these simple methods
provide a good measure of capacity utilization. Unfortunately, as emphasized in the
seminal paper by Orphanides and van Norden (2002), they are less reliable in real time as
they are often subject to substantial revisions. This is primarily because trend extraction
becomes problematic at the end of a given sample period (see Hamilton (2018)fora
comprehensive discussion of the HP filter and Canova (2020) for a broader discussion
on the properties of several filters and their ability to recover the output gap generated
by structural macroeconomic models). Despite these well-known issues, according to
Coibion et al. (2018) simple univariate statistical filters still seem to play a central role in
the production of output gap estimates at several policy institutions.
In order to estimate the output gap, Norges Bank uses multivariate models, which, in
addition to GDP data, use data on other variables such as unemployment, wages, inflation,
investment, credit growth and house price growth. Over time, a suite of multivariate models
composed of unobserved component (UC) models and structural vector-autoregressive
(SVAR) models has been developed. A simple equally weighted average of the estimates
obtained from the different models included in the suite tracks well the official Norges
Bank’s output gap series, which is a judgemental measure representing the institutional
view on capacity utilization in the Norwegian economy. Norges Bank’s approach builds
on Fleischman and Roberts (2011), Basistha and Startz (2008) and Jarocinski and
Lenza (2018), among others, who find that using multivariate models is key to obtaining
more precise and reliable estimates of the output gap for the United States and the euro
area. The use of data on inflation in combination with GDP follows Kuttner (1994), who
initiated the literature on the output gap and Phillips curve estimation in UC models.
Clark (1989), Apel and Jansson (1999) and Sinclair (2009) model the relationship between
the output gap and unemployment according to Okun’s law. The use of data on investment
follows Domenech and Gomez (2006), who emphasize the importance of the investment
share in the economy as a key variable for identifying cyclical developments. Furthermore,
Borio, Disyatat and Juselius (2017) argue that it is important to take into account data on
financial variables when assessing potential output, a point also stressed by Furlanetto,
Gelain and Taheri Sanjani (2021a) in the context of an estimated macroeconomic model
with financial frictions and by Berger, Richter and Wong (2022b) in a Bayesian Vector
Autoregression.
We show that the model suite delivers estimates of the output gap that perform
relatively well in terms of forecasting and reliability when new data become available.
Notably, a simple equally weighted average of the nine output gap estimates features
better forecasting properties than the individual models. In addition, such a summary
©2022 The Authors. Oxford Bulletin of Economics and Statistics published by Oxford University and John Wiley & Sons Ltd.

Get this document and AI-powered insights with a free trial of vLex and Vincent AI

Get Started for Free

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex

Start Your Free Trial of vLex and Vincent AI, Your Precision-Engineered Legal Assistant

  • Access comprehensive legal content with no limitations across vLex's unparalleled global legal database

  • Build stronger arguments with verified citations and CERT citator that tracks case history and precedential strength

  • Transform your legal research from hours to minutes with Vincent AI's intelligent search and analysis capabilities

  • Elevate your practice by focusing your expertise where it matters most while Vincent handles the heavy lifting

vLex