On the Stability of Euro Area Money Demand and Its Implications for Monetary Policy
| Published date | 01 August 2018 |
| Author | Matteo Barigozzi,Antonio M. Conti |
| DOI | http://doi.org/10.1111/obes.12239 |
| Date | 01 August 2018 |
755
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd.
OXFORD BULLETIN OF ECONOMICSAND STATISTICS, 80, 4 (2018) 0305–9049
doi: 10.1111/obes.12239
On the Stability of EuroArea Money Demand and Its
Implications for Monetary Policy*
Matteo Barigozzi† and Antonio M. Conti‡, §
†London School of Economics and Political Science, London, UK
‡Banca d’Italia, Economic Outlook and Monetary Policy Department, Roma, Italy
(e-mail: antoniomaria.conti@bancaditalia.it)
§ECARES, Universit´e libre de Bruxelles, Bruxelles, Belgium
Abstract
We employ a recent time-varying cointegration test to revisit the usefulness of long-run
money demand equations for the ECB, addressing the issue of their instability by means
of a model evaluation exercise. Building on the results, we make a twofold contribution.
First, we propose a novel stable money demand equation relying on two crucial factors:
a speculative motive, represented by domestic and foreign price-earnings ratios, and a
precautionary motive, measured by changes in unemployment. Second, we use the model
to derive relevantpolicy implications for the ECB, since excess liquidity looks more useful
for forecasting stock market busts than future inflation. Overall, this evidence points to (i)
a possible evolution of the monetary pillar in the direction of pursuing financial stability
and (ii) the exclusion of a sudden liquidity–driven inflationary burst after the exit from the
prolonged period of unconventional monetary measures.
I. Introduction
Monetary aggregates have been largely neglected by the prevailing macroeconomic con-
sensus until the Global Financial Crisis of 2008–09 triggered a renewedinterest for money’s
role in the business cycle and the financial cycle. Following the rising and diffusion of the
New Keynesian model as a tool for policyanalysis and forecasting, money has been indeed
gradually disappearing from monetary analysis (Nelson, 2003).1Such theoretical devel-
opments, together with empirical evidence on the instability of the relationship between
inflation and money growth, had a substantial impact on the conduct of monetary policy.
For instance, the relative weights of the two pillars on which the strategy of the European
Central Bank (ECB) is based – one resting on monetary analysis the other one on economic
analysis – have changed over time, with the latter taking centre stage in 2003 (European
JEL Classification numbers: E41, E52, C32.
*The views here expressed are those of the authors and do not necessary reflect those of the Banca d’Italia or the
Eurosystem.
1In contrast, Ireland (2004) has challenged this view and a number of more recent studies have included money
in models aimed to monetary policy analysis (Zanetti, 2012; Benati et al., 2017).
756 Bulletin
Central Bank, 2003). Accordingly, the great emphasis put on the estimation of money de-
mand equations decreased, also in relation to their instability over time (see Papademos
and Stark, 2010, for a recent survey). On the other hand, a number of contributions has
recommended not to neglect money’s role in macroeconomics, both on theoretical and em-
pirical grounds, because of its impact on economic activity and inflation (Reynard, 2007;
Beck and Wieland, 2008; Nelson, 2008; Favara and Giordani, 2009; Canova and Menz,
2011).
Given this debate, in this paper we re-evaluate the relevance of monetary aggregates
for price and financial stability. In particular, by means of a newly developed econometric
method, we propose a new stablemoney demand equation from which we derive a measure
of excess liquidity (with respect to its long-run equilibrium) that provides a valuable signal
for both future financial imbalances and inflationary pressures. Financial intermediaries’
balance sheets are under the constant monitoring of policy makers and the literature has
suggested that monetary aggregates, which are the counterpart of bank lending, convey
information on the stage of the financial cycle (Shin and Shin, 2011). Indeed, one can look
at the demand for money, i.e. banks’ liabilities, as supplyof funding to banks. Therefore, a
novel empirical analysis of the determinants of the long-run stock of money is required for
studying the role of monetary aggregates in shaping policies oriented to financial stability
(European Central Bank, 2012; Kim, Shin and Yun, 2012; Allen, Carletti and Gale, 2014).
In order to accomplish this task we first have to address the well-known issue of the
instability of money demand (De Santis, Favero and Roffia, 2013), which in turn requires
comparing different existing models. While the emergence of unstable money demand
behaviour in the euro area (EA) seems to originate in the early 2000s, the exact date of
such break remains a matter of discussion. Hence, we investigate the existence of a stable
long-run money demand equation in the EA by using a time-varying cointegration test
proposed by Bierens and Martins (2010), a generalization of the standard methodology by
Johansen (1996) which avoids the need for specifying the exact timing of the structural
break.
Our contribution is twofold. First, we run a model evaluation exercise of different
specifications of EA money demand proposed by the literature (Carstensen, 2006; Dreger
and Wolters, 2010a, and many others; seeTable 1 for a complete list), in order to determine
which ones, if any, are stable and therefore effective for policy analysis. We consider a
quarterly data set spanning the sample from 1980:Q1 to 2016:Q3 and we find that most of
the considered models are indeed unstable after 2001, although improvementsare observed
when including either (i) the spread between EA and US price-earnings ratios (De Santis
et al., 2013), or (ii) the changes in EA unemployment rate (De Bondt, 2010). Second, we
develop further this result by proposing, estimating, and identifying a new stable model
for money demand relying on both those motives. Finally, we derive useful implications
for monetary policy. Specifically, we find that: (i) the monetary overhang computed from
our model is a leading indicator of the probability of stock market busts and (ii) it provides
some incremental predictive content for forecasting inflation, although this tends to vanish
when excess liquidity is considered together with real GDP growth,proxying the economic
pillar of the ECB.
To our knowledge, this is the first contribution proving that excess liquidity measures
derived from estimated long-run money demand equations may be helpful for predicting
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
On the stability of euro area money demand 757
TABLE 1
Models for euro area moneydemand. Specifications considered
References Acronym Xtr Sample k
Coenen and Vega (2001) CV `t,st,t3 1980:Q1–1998:Q4 5
Calza et al. (2001) CGL1 (st−ot), (`t−ot) 1 1980:Q1–1999:Q4 4
Calza et al. (2001) CGL2 (st−ot) 1 1980:Q1–1999:Q4 3
Gerlach and Svensson (2003) GS (`t−st) 2 1980:Q1–2001:Q4 3
Vlaar (2004) V (`t−st), t3 1980:Q1–2000:Q4 4
Carstensen (2006) C (st−ot), (et−ot), vt1 1980:Q1–2003:Q4 5
Beyer (2009) B1 ot,st,t,ht2 1980:Q1–2008:Q4 6
Beyer (2009) B2 ht2 1980:Q1–2008:Q4 3
Beyer (2009) B3 ht,t2 1980:Q1–2008:Q4 4
Dreger and Wolters (2010a) DW1 t1 1983:Q1–2004:Q4 3
Dreger and Wolters (2010b) DW2 st,lt,t,ft2 1983:Q1–2010:Q2 6
Dreger and Wolters (2010b) DW3 ft,t1 1983:Q1–2010:Q2 4
Dreger and Wolters (2010b) DW4 ft,t,(lt−st) 2 1983:Q1–2010: Q2 5
De Bondt (2010) D1 wt,ot,et,ut1 1983:Q1–2007:Q2 6
De Bondt (2010) D2 wt,ot,et1 1983:Q1–2007:Q2 5
De Bondt (2010) D3 ot,ut2 1983:Q1–2007:Q2 4
De Santis et al. (2013) DFR1 `t,`*
t,qt,q*
t,ot3 1980:Q1–2007:Q3 7
De Santis et al. (2013) DFR2 `t,`*
t,qt,q*
t3 1980:Q1–2007:Q3 6
De Santis et al. (2013) DFR3 (`t−`*
t), (qt−q*
t) 3 1980:Q1–2007:Q3 4
Barigozzi and Conti BC ut,(qt−q*
t), (`t−ot), t3 1980:Q1–2008:Q4 6
Barigozzi and Conti BC ut,(qt−q*
t), (`t−ot), t3 1980:Q1–2014:Q2 6
Notes: The time-varying long run money demand equation considered is (mt−pt)=0+y
tyt+X
tXt, where: (mt−
pt)=log-real balances, yt= log-real income, `t=long-term interest rate, st=short-term interest rate, ot=own rate,
t=q-o-q inflation rate (in Beyer (2009) y-o-y inflation rate), et=q-o-q equity returns, vt=log-volatility of equity
returns, ht=y-o-y housing wealth growth rate, ft=log-real financial wealth, wt=log-real wealth, ut=y-o-y
differences of unemploymentrate, `*
t=US long-term interest rate, qt=log-price to earnings ratio, q*
t=US log-price
to earnings ratio. The last three columns report the cointegration rank r, the sample, and the number of variables k,
used in the original paper.
financial crises. Moreover, this finding is still valid when controlling for the slope of the
yield curve, the credit-to-GDP ratio and non-core banks liabilities, considered among the
most relevantpredictors of financial stress (Bank of Inter national Settlements, 2010; Hahm,
Shin and Shin, 2013). Hence, we add to the increasing debate on central banks financial
stability mandate (Hahm et al., 2013; Kim, Shin andYun,2013) suggesting that infor mation
from the monetary pillar, in particular from excess liquidity measures, should mainly be
used to assess the risk of financial crises in the EA.
Results are robust to the estimation of our model on different time spans, i.e. both in
‘normal times’ (1980:Q1–2008:Q4) and when extending the sample to most recent data
(1980:Q1–2016:Q3), covering the Global Financial Crisis, the Sovereign Debt Crisis and
the unconventional measures implemented by the ECB. Furthermore, our model helps in
predicting M3 growth in the EA both in ‘normal times’ and in times of financial stress.
Overall, our findings support the usefulness of money demand equations as tools for
monetary policy, provided they are properly specified, and suggest that they matter more
for financial stability than for controlling inflation.
©2018 The Department of Economics, University of Oxford and JohnWiley & Sons Ltd
To continue reading
Request your trial