Measuring Exchange Rate, Price, and Output Dynamics at the Effective Lower Bound
| Date | 01 December 2018 |
| DOI | http://doi.org/10.1111/obes.12260 |
| Author | Gregor Bäurle,Daniel Kaufmann |
| Published date | 01 December 2018 |
1243
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 80, 6 (2018) 0305–9049
doi: 10.1111/obes.12260
Measuring Exchange Rate, Price, and Output
Dynamics at the Effective Lower Bound*
Gregor B ¨
aurle† and Daniel Kaufmann‡,§
†Swiss National Bank, P.O. Box, CH-8022 Zurich, Switzerland
(e-mail: gregor.baeurle@snb.ch)
‡University of Neuchˆatel, Rue A.-L. Breguet 2, CH-2000 Neuchˆatel, Switzerland
(e-mail: daniel.kaufmann@unine.ch)
§KOF Swiss economic Institute ETH Zurich, Leonhardstrasse 21, CH-8092 Zurich,
Switzerland
Abstract
New Keynesian models with sticky prices make stark predictions about how the economy
responds to shocks under different monetary policy regimes when short-term interest rates
are constrained by an effective lower bound. We use the Swiss case as a laboratory to find
evidence in favour of these predictions. We propose a Bayesian VAR to estimate impulse
responses to risk shocks for short periods with a binding effective lower bound and with
a publicly announced minimum exchange rate. In line with predictions from theory, we
find that with a binding effective lower bound, the responses of the exchange rate, prices,
and output become more persistent. However, the minimum exchange rate attenuates this
adverse impact.
I. Introduction
New Keynesian models with sticky prices make stark predictions about how the economy
responds to shocks when the central bank faces an effectivelower bound (ELB) on nominal
interest rates (see e.g. Krugman, 1998; Eggertsson and Woodford, 2003; Adam and Billi,
2007; Kiley, 2016).1After a contractionary shock, agents expect prices to fall. Therefore,
JEL Classification numbers: C11, C32, E31, E52, E58, F31.
*We thank two anonymous referees, Christiane Baumeister, Luca Benati, Fabio Canetg, Nicolas Cuche-Curti,
Harris Dellas, Alain Galli, Sylvia Kaufmann, Miles Kimball, Leo Krippner, Mariano Kulish, Stefan Leist, Carlos
Lenz, Matthias Lutz, James Morley, Alexander Rathke, Alex Perruchoud, Barbara Rudolf, Jonas Stulz, Rodney
Strachan, Dominik Studer, C´edric Tille, as well as seminar participants at the University of Bern, the Swiss National
Bank, the Bank of Lithuania, the Bank of Canada, and seminar participants at the SSES Annual Meeting, the SNB-
BIS-Dallas Fed-CEPR conference on ‘Inflation Dynamics in a Post-CrisisGlobalised Economy’, the Jahrestagung of
theVerein f¨ur Socialpolitik, the EABCN conference on ‘Inflation Developments after the Great Recession’, the RBNZ
conference on ‘Monetary Policyin Open Economies’, the AEA annual meeting, for helpful comments and discussions.
The views expressed in this paper are those of the authors and not necessarily those of the Swiss National Bank. A
previous version of this paper has circulated as SNB Working Paper titled ‘Exchange Rate and Price Dynamics in a
Small Open Economy – The Role of the Zero Lower Bound and Monetary Policy Regimes’.
1As long as it is possible to hold currency that earns a zero nominal interest rate, there is an effective lowerbound
on nominal interest rates, which is – due to storage costs, transaction costs or legal impediments to holding large
1244 Bulletin
the real interest rate increases for a given nominal interest rate stuck at the ELB, which
exacerbates the initial impact of the shock. An open economy additionally suffers from
a strong and immediate real and nominal appreciation of the currency because the real
interest rate increases relative to the real interest rate abroad (see Cook and Devereux,
2013, 2016).
Although the theoretical predictions are bleak under standard Taylor-type policy rules
and inflation targeting, it is well known that there exist alternative commitments – such
as history-dependent policy rules, as well as setting a price-level or exchange-rate target
– that ameliorate the adverse impact of the ELB (see Reifschneider and Williams, 2000;
Svensson, 2001; Eggertsson and Woodford, 2003; McCallum, 2006; Fujiwara et al., 2013).
The powerof such commitments is controversial, however,and has been called into question
in theoretical work by Del Negro, Giannoni and Patterson (2012), McKay, Nakamura and
Steinsson (2016), Carlstrom, Fuerst and Paustian (2015) and Levin et al. (2010). Moreover,
empirical work by Gar´ın, Lester and Sims (2018) raises doubt about key implications of
the New Keynesian model when the ELB is binding.2
Ultimately, whether these theoretical predictions are accurate is an empirical question.
However, such an assessment is hampered by the fact that episodes with a binding ELB
are intermingled with additional non-conventional monetary policy actions. Because these
non-conventional actions influence expectations about the future behaviour of the central
bank, constant-parameter models are subject to the Lucas critique (see also Weale and
Wieladek, 2016). Regrettably, from a strictly empirical point of view, episodes with a
binding ELB but without nonconventional policy actions are rare and of short duration.
Central banks usually takeadditional measures once their interest rate instr uments approach
the ELB.3Tothe extent that those interventions (explicitly or implicitly) shape expectations
on the future evolution of the short-term interest rate, most episodes are not informative
about the behaviour of the economy at the ELB under an unchanged monetary regime.
Todisentangle the impact of the ELB from non-conventional policy actions, Switzerland
proves to be an interesting case to study. We can identify two short episodes with a binding
ELB and little non-conventional policy (i.e. no level target, no expectations management,
and no permanent balance sheet expansions). Moreover, weobser ve an unexpectedregime
change when the SNB introduced an explicit minimum level for the CHF/EUR exchange
rate in September 2011. Therefore, we can measure the dynamics of the Swiss economy
with a binding ELB in addition to a publicly announced nominal level target.
Weestimate a structural vector autoregressive model (SVAR)and show how the impulse
responses to contractionary risk shocks, i.e. shocks that increase uncertainty in Switzer-
land’s main trading partners, differ among the regimes. To tackle the small number of
amounts of currency – potentially lowerthan zero. As Swanson and Williams (2014b) emphasise, this effectivelower
bound can also be positive for institutional reasons.
2Moreover, Keen, Richter andThrockmorton (2017) show that the effectiveness of forward guidance differs in a
nonlinear model.
3We find many examples of a binding ELB and simultaneous non-conventional actions. Most of the large-scale
asset purchases and forward guidance by the Federal Reserve were introduced at the same time as interest rates
were close to zero. Moreover, the Swiss National Bank (SNB) conducted large-scale exchange rate interventions
simultaneously with corporate bond purchases, while its policy instrument was effectivelyconstrained. Similarly, the
Czech National Bank adopted a commitment to maintain the exchange rate close to koruna 27 to the euro in autumn
2013, after short-term interest rates fell technically to zero at the end of 2012.
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial