International Banking and Financial Stability in the

International Banking and Financial Stability in
the Euro Area: Do We Really Need a Federal
Institution?
Gauthier Vermandel
This version: April 8, 2014
(Preliminary draft)
Abstract
What should be the role of the federal level in the settlement of macroprudential measures? This paper evaluates the consequences of a symmetric treatment of countries regarding the implementation of macroprudential
measures in the Eurozone. In an estimated two-country DSGE model accounting for alternative macroprudential instruments and implementations
schemes, we get four main results. First, macroprudential measures lead to
an highest increase in welfare at the union level if they take into account
national features. Second, macroprudential policy implementation leads to
winners (the peripheral countries) and losers (the core countries). Third,
using a Nash bargaining game, we …nd that national-adjusted macroprudential measures need a federal institution to impose coordination between
EU members, as countries have incentives to choose the second best non
cooperative outcome. Fourth, cross-border loan ‡ows are a critical feature
to address cooperation incentives issues, we observe a Nash equilibrium reversal in banking autarky.
JEL classi…cation: E32; E44; E52; F36; F41;
Keywords: Banking Globalization, Monetary and Macroprudential policy,
Banking Union, Euro Area, Financial Accelerator, DSGE Two-Country Model,
Bayesian Estimation
CREM, UMR CNRS 6211, Université de Rennes I, Rennes,
[email protected]
Prepared for the XI Macro….
1
France.
E-mail:
This paper analyses how national and federal concerns should be balanced in
designing the macroprudential mandate1 in a monetary union such as the Eurozone.
As recommended by the IMF (IMF, 2013), the legislation regarding national
macroprudential systems should include adequate provisions regarding the objective, the functions and the powers of the macroprudential authority. Namely,
clear objectives (with explicit targets) should guide the decision-making process
and enhance the accountability of authorities. The key macroprudential functions
should include the identi…cation of systemic risks, the formulation of the appropriate policy response and the implementation of the policy response through adequate rulemaking. Finally, the macroprudential authority should be empowered
to issue regulations, collect information, supervise regulated entities and enforce
compliance with applicable rules.
The institutional organization of macroprudential policy is original in the Eurozone, as it is the only supranational setup yet to be observed2 . As they constitute an integrated …nancial area, European countries have taken into account the
fact that …nancial stability should be treated as a public good and that isolated
national actions may undermine the conduct of an appropriate macroprudential
policy. A range of new institutions have been established following the report
of De Larosière (2009). The European Financial Stability Framework accounts
for a macroprudential considerations through the European Systemic Risk Board
(ESRB). This institution is intended to encourage appropriate macroprudential
actions by the national regulators to overcome or mitigate systemic risk. The
role of this institution is still under de…nition, as the recent evolution towards the
banking union reinforces the role of the ECB in the supervision of this policy.
Furthermore, more symmetric practices between union members are encouraged.
One of the aspects of the recent institutional evolution regarding the role of
macroprudential is to consider the degree of symmetry that should be imposed
in coordinated decisions. This question arises since the other two main macroeconomic policy tools are applied symmetrically to Eurozone members: monetary
policy reacts to union wide aggregates while the Stability and Growth Pact (SGP)
imposes common rules on national debts and de…cits.
The aim of this paper is to evaluate the consequences of imposing symmetric
1
Broadly de…ned, the …nal objective of macroprudential policy is to prevent or mitigate
systemic risks that arise from developments within the …nancial system, taking into account
macroeconomic developments, so as to avoid periods of widespread distress. The main threat
of this policy is to assess and control risk at the aggregate level, in contrast to microprudential
measures that concentrate on individual actors.
2
For a synthetic presentation of the various institutional situations observed in real life practices regarding the de…nition of the macroprudential mandate, see (Nier et al., 2011). In some
cases, it involves a reconsideration of the institutional boundaries between central banks and
…nancial regulatory agencies (or the creation of dedicated policymaking committees), while in
other cases, e¤orts are made to favor the cooperation of authorities within the existing institutional structure. (Nier et al., 2011) …nd that the vast majority of arrangements that are in
place or are being developed across countries can be organized in seven models, which in turn
form three broad groups of models that di¤er in the degree of institutional integration between
central bank and regulatory agencies.
2
practices in a macroprudential mandate that is already based on the cooperation
between agents. We develop a two-country DSGE model that accounts for some
major features of the Eurozone regarding the problem at hands (key role of the
banking system, cross-border bank loans, heterogeneity credit and business cycles)
to provide a quantitative evaluation of welfare gains coming from a more symmetric
treatment of countries in the implementation of macroprudential measures. In this
model, heterogeneity between the member countries of the Eurozone is accounted
for by distinguishing core countries and peripheral countries3 .
We introduce two major novelties in the analysis. First, regarding the choice of
macroprudential measures, we assume that, according to Blanchard et al. (2013)
classi…cation, the macroprudential authorities can use two instruments: to a¤ect
the lending conditions, they react to the level of bank capital requirement, while,
on the borrowing side, they react to the evolution of the loan to income ratio.
These macroprudential concerns a¤ect in a countercyclical way the loan interest
rates faced by entrepreneurs to …nance investment. Second, we contrast three
possibilities regarding the implementation of macroprudential decisions: (i) a coordinated de…nition of macroprudential measures based on the choice of optimal
national policy parameters; (ii) a coordinated de…nition of macroprudential measures based on the choice of optimal national policy parameters for borrowing
conditions and an homogenous value for the parameter related to bank capital
requirements; (iii) an homogenous de…nition of all macroprudential measures.
The model is estimated with Bayesian methods on Eurozone quarterly data
over the sample period 1999Q1 to 2013Q24 . The interest of DSGE models to deal
with this problem is that they provide an toolkit for deriving the optimal value
of the macroprudential stance and for performing welfare analysis to rank the
alternative macroprudential coordination schemes. We report the welfare ranking
of alternative macroprudential policy schemes in terms of an increase in permanent
consumption for the Eurozone, core countries and peripheral countries.
Our main results can be stated as follows : First, macroprudential measures
increase welfare at the union level from 0.690% to 0.886% of permanent consumption. The ranking clearly shows that a country-adjusted macroprudential mandate
o¤ers the best performance in terms of unconditional consumption. However, this
situation leads to winners (the peripheral countries) and losers (the core countries). Second, we …nd that imposing a common rule on capital requirements
between countries mitigates the welfare loss for the Core area. Third, using a simple Nash bargaining game, we …nd that without a federal institution that obliges
the euro members to coordonate macroprudential policies, the rational strategy for
national governments is to choose the non cooperative outcome. These incentives
3
The criterion to divide the Eurozone in two blocks is discussed in the following section.
Core countries: Austria, Belgium, Germany, Finland, France, Luxembourg and Netherlands.
Peripheral countries: Spain, Greece, Ireland, Italy and Portugal.
4
To our knowledge, the design of macroprudential measures has been approached by Pariès
et al. (2011), Quint and Rabanal (2013), Kannan et al. (2009), Benes and Kumhof (2011), Pariès
et al. (2011), Angelini et al. (2012), Bailliu et al. (2012), Beau et al. (2012), Collard et al. (2012),
Rubio and Carrasco-Gallego (2012), Angeloni and Faia (2013), Brzoza-Brzezina et al. (2013),
Lambertini et al. (2013), Medina and Roldós (2013), Suh (2014) and Kincaid and Watson (2013).
3
not to cooperate brings rationally the monetary union in a second best situation.
Fourth, cross-border loan ‡ows are a critical feature to address incentives issues,
we observe a nash equilibrium reversal in banking autarky.
The paper is organized as follows: section 1 describes the institutional background and some stylized facts regarding the situation of the Eurozone, section 2
outlines the model and the de…nition of macroprudential policies, section 3 presents
the results and section 4 concludes.
1
The institutional Background
Following the report of De Larosière (2009), several new institutions have been
created at the EU level since January 2011 under the European System of Financial Supervision (ESFS) to ensure supervision of the Union’s …nancial system.
Among these agencies, the European Systemic Risk Board (ESRB) is responsible for the macroprudential oversight of the …nancial system within the European
Union. The role of this independent agency is based on soft power, as it is aimed
at coordinating prudential decisions for countries participating to the Eurozone
and other European countries. It is intended to issue risk warnings and encourage
appropriate national macroprudential actions to overcome or mitigate systemic
risk. The ECB provides analytical, statistical, logistical and administrative support to the ESRB whose powers are purely advisory, including the authority to
make its advice public.
As shown in Figure 1, di¤erent policy initiatives have been taken since the
creation of the ESRB, and a road map has been set to enhance more homogenous practices among countries participating to this structure. This calendar can
be analyzed as providing a smooth transition towards a more centralized and
symmetric system. In particular, the Single Supervisory Mechanism (SSM) initiative may modify the original organization of macroprudential implementation for
countries participating to this structure. As noted by ESRB (2013, report October), it should be necessary to rediscuss the distribution of responsibility between
the European Central Bank and the National Central Agencies in the conduct of
macro-prudential policy to move towards a more centralized model: as the ECB
will be in charge of the micro-prudential instruments which will be used for macroprudential purposes it should also be in charge of macro-prudential policy. The
main task of the NCAs would be to provide information about their national business/housing cycles conditions and to make recommendations to the ECB, which
would take the …nal decision.
The move towards more symmetric practices in the conduct of macroprudential practices can be justi…ed on an economic ground as countries belonging to
the Eurozone form an integrated …nancial area. In this area, banks provide the
main liquidity to the system and cross-border banking has reached a value representing 24% of Eurozone GDP before the …nancial crisis of 2008. The transfer of
macroprudential powers to the Federal level can thus make sense as cross-border
banking activities have increased the interconnection of …nancial decisions in the
Eurozone.
4
The calendar
December 2011 : The ESRB recommends Member States to establish a legal mandate for national macro-prudential authorities by
June 2013.
December 2012 : The Council agreed on a regulation creating a single supervisory mechanism (SSM) which will be responsible for the
micro and macro-prudential supervision of all banks in the participating countries with the ECB acting as European supervisor
”responsible for the effective and consistent functioning” of the
mechanism.
April 2013 : The ESRB recommends Member States to ensure a minimum set of instruments is available, and identifies a common
benchmark for intermediate objectives and instruments.
January 2014 : The new regulatory framework for banks enters into
force (CRD4/CRR).
Summer 2014 : The SSM acquires some macro-prudential powers for
instruments included in the CRD4/CRR, namely those related to
banks.
December 2014 : Deadline ESRB recommendation on intermediate
objectives and instruments.
December 2015 : Deadline ESRB recommendation on strategy.
Figure 1: Road map of the newly created institutions in the European Union
Setting homogenous macroprudential rules at the union level can be considered
as a solution to a problem of externalities. Externalities arise from the fact that
national authorities do not internalize their contribution to federal …nancial instability. Two macroprudential functions outlined by the IMF legal de…nition are
concerned. First, regarding the identi…cation of systemic risks, where …nancial institutions have a¢ liates in multiple jurisdictions, this complicates the assessment
of systemic risk and can lead to con‡icts between home and host authorities. According to the home country principle, national authorities are only able to assess
the …nancial disequilibrium of their national intermediaries.
Secondly, the formulation of the appropriate policy response should be made
at the federal level, as cross-border spillovers cannot be ignored. Indeed, lack of
forceful macroprudential action in one country can increase the likelihood of crises,
imposing negative externalities on other countries. Furthermore, national policies
to contain risks from a rapid build-up of domestic credit can lead to an increase in
the provision of cross-border credit. Finally, policies to strengthen the resilience of
systemic institutions in one country can cause their activities to migrate to other
countries.
However, the choice of homogenous values for the macroprudential instruments
may a¤ect the implementation of the appropriate policy response. To be e¤ective
these policies should also be tailored to the situation of economies. The problem
5
500
450
(a) Corporate Credit
per capita millions euro
(b) Corporate Credit Spread
Credit minus refi rate
(c) Current Account
% of GDP
4%
0%
400
350
300
3%
2%
2006 2008 2010 2012 2014
−5 %
2005
2010
Core countries
2000
2005
2010
Peripheral countries
Figure 2: Regional divergences in the UEM
is that national situations regarding …nancial disequilibrium are heterogeneous.
In Figure 2.c), we separate the Eurozone in two groups according to their status
in terms of surplus or de…cit of their current account. As noticed, situations are
di¤erent. A descriptive exercise regarding the attribution to the ECB of a simple
rule that reacts to the growth rate of loans clearly shows the problem associated to
the conduct of an homogenous decision at the federal level. The fact that …nancial
cycles are mainly national may overcome the gain to conduct symmetric macroprudential policies, as it should be optimal to target in a speci…c way national
development in a tailored way that may di¤er from countries to countries. More
particularly, the recent …nancial turmoil highlights very important di¤erences in
the credit cycles. Particularly, Figure 2.b) shows that the transmission of monetary
policy via the credit channel remains di¢ cult as the wedge between the corporate
credit rate and the re…nancing rate is still widening in periphery. According to
Figure 2.a) and Figure 2.b), Periphery may require a di¤erent implementation of
macroprudential policy as peripheral countries experienced an explosive growth of
credit followed by a sharp drop as well as a rise of the corporate bond yield.
Thus, the main question that arises from this short description of the main stylized fact is to evaluate wether an institutional evolution leading to more symmetric
practices could be an e¢ cient solution to conduct macroprudential measures in
the Eurozone.
2
The Analytical Framework
This section introduces a two-country DSGE model that accounts for the main
speci…cities of the Eurozone for the question at hands (namely …nancial heterogeneities and cross-border loans). In this setting, macroprudential policies are
implemented using two instruments that are set cooperatively between union members.
6
2.1
A Description of the model
Our model5 describes a monetary union made of two asymmetric areas i 2 fh; f g
(where h is for home and f for foreign parts) of relative sizes n and 1 n. As
shown in Figure 3, each part of the monetary union is populated by consumers,
intermediate and …nal producers, entrepreneurs, capital suppliers and a banking
system. Regarding the conduct of macroeconomic policy, we assume national
…scal authorities and a common central bank. We present the model anticipating
the symmetric equilibrium across households, …rms and banks that populate the
economy.
d
Deposits Dh,t
Bank
Central Bank
Refinancing
Rate Rt
Bank
Lsh,t
Production
Ch,t
Cross-Borders
Loans
Investment
Flows
Consumption
Flows
Lsf,t
Production
Cf,t
Household
Household
d
Deposits Df,t
Figure 3: The model of a two-country monetary union with international bank
loan ‡ows
d
Households The representative household supplies Hi;t hours of work, saves Di;t
P1
e"i;t+ U (Ci;t+ ; Hi;t+ ), where
and maximizes utility intertemporally Et
=0
Ci;t is the consumption,
2 (0; 1) is the subjective discount factor and "i;t is
an exogenous shock to preferences. The period utility function takes the form
1+ L
1 c
c
i
0 is the curvature
= 1 + Li where Li
U (Ci;t ; Hi;t ) Ci;t i = (1
i Hi;t
i)
c
coe¢ cient in the disutility of labor and i 0 is the risk aversion coe¢ cient. The
consumption basket of the representative household is composed of home and for(
1)=
( 1)=
C 1=
1=
C
eign goods Ci;t = ((1
Chi;t
+( C
Cf i;t ) =( 1) where 1
i )
i )
i > 1=2
is the home bias in consumption and
0 is the elasticity of substitution between
home and foreign goods.
Firms There is a continuum of monopolistically competitive …rms, each producing di¤erentiated goods using hours of work and capital inputs Ki;t and set
production prices Pi;t according to the Calvo model. Output supplied by …rms is
A
1
Yi;t = e"i;t Ki;t Hi;t
where "A
2 [0; 1]
i;t is an innovation to the productivity and
5
The whole model is presented in appendix.
7
is the share of capital services in the production. According to the Calvo mechanism, each period …rms are not allowed to reoptimize the selling price with probability pi but price increases of pi 2 [0; 1] at last period’s rate of price in‡ation,
p
Pi;t = i;ti 1 Pi;t 1 where i;t = Pi;t =Pi;t 1 . Under this setting, it is possible to derive the aggregate in‡ation rate of production goods, it is de…ned by the function,
i;t = f (Et i;t+1 ; i;t 1 ; mci;t ) where mci;t is the marginal cost of production.
Entrepreneurs We add a borrowing constraint to the production sector to implement a banking sector in the model. We standardly introduce an entrepreneurial sector that buys capital at price Qt in t and uses that capital in the
production in period t + 1. Under this assumption, the capital arbitrage equak
=
tion implies that the expected rate of return on capital is given by 1 + Ri;t+1
Et [Zi;t+1 + (1
) Qi;t+1 ] =Qi;t where 2 [0; 1] and and Zi;t are respectively the
depreciation rate and the marginal product of capital. Assuming that entrepreneurs are credit-constrained, they …nance capital by their net wealth Ni;t
k
and lending LH
i;t+1 subject to external habits hi 2 [0; 1]. The balance sheet
of the entrepreneur then writes, Qi;t Ki;t+1 = LH
i;t+1 + Ni;t+1 . The entrepreneur
has access to domestic and foreign banks to meet its balance sheet, Li;t+1 =
( 1)=
( 1)=
L 1=
((1
Lhi;t+1 + ( Li )1= Lf i;t+1 ) =( 1) where Li represents the percentage
i )
L
, is
of cross-border loan ‡ows in the monetary union. The total cost of loans, Pi;t
L
L
L 1
L
L 1
1=(1 )
L
thus de…ned according to, Pi;t = ((1
+ i (Rf;t ) )
where Ri;t
i )(Rh;t )
denotes the credit rate set by bank in country i.
The representative entrepreneur conducts a mass ! 2 [! min ; +1) of heterogenous investment projects drawn from Pareto distribution. The rentability of the
k
! th investment project is, ! 1 + Ri;t+1
. There is a critical project ! ci;t that determines the threshold of pro…tability of the …rm. Aggregating projects above the
threshold, we can compute the share of pro…table projects i;t in the economy i.
Supposing that entrepreneurs are pessimistic as De Grauwe (2010) regarding their
expected aggregated return ! on investment projects, we …nd the …nancial acceleration equation as in Bernanke et al. (1999). The external …nance premium drives a
wedge between the expected return on capital and the expected return demanded
k
L
by banks and takes the form, Ri;t
=Pi;t
= {i f (Qi;t Ki;t+1 =Ni;t+1 (e)) where {i > 0
measures the elasticity of the premium with respect to leverage.
Banks The representative banks collects monopolistically deposits Di;t from
households, borrows funds LRF
i;t from the central bank and monopolistically pros
vide loans Li;t to entrepreneurs. The balance sheet writes, Lsi;t+1 = Di;t+1 +LRF
i;t+1 +
BKi;t+1 . Following Pariès et al. (2011), we measure the pass-through of interest
rates by supposing that the representative bank sets the deposit and credit rates in
staggered basis à la Calvo. Letting Li ( D
i ) denotes the country speci…c probability
of the bank not being able to reset it credit (deposit) interest rate. The aggregate
D
D
D
deposit rate writes, Ri;t
= f Et Ri;t+1
; Rt ; "D
i;t where "i;t is a markup shock and
Rt is the (taylored) ECB re…nancing rate. The New Keynesian Phillips Curve for
deposit rates implies that the expected future rate depends on the re…nancing rate
markup and exogenous shock. Similarly, the aggregate credit rate is de…ned by,
8
L
L
L
Ri;t
= f Et Ri;t+1
; Et i;t+1 ; Rt . Solving forward Ri;t
, one can see that current and
expected future ECB rate Rt and …rms pro…tability Et i;t+1 drive today’s credit
rates.
Capital Suppliers The representative capital producer buys deprecated capital
stock (1
) Ki;t and investment goods Ii;t and produces new capital goods Ki;t+1
at a price Qi;t . Capital supplier buys home and foreign investment goods, Ii;t =
1=
( 1)=
( 1)=
I 1=
I
Ihi;t
If i;t
( 1
+ Ii
) =( 1) where 1
i
i > 0:5 is the home bias
in its consumption basket.
Monetary Policy Finally, monetary authorities choose the nominal interest
rate according to a standard Taylor rule Rt = f cu;t ; Yu;t ; "R
where cu;t and
t
Yu;t are the growth rates of price and GDP of the monetary union and "R
t is an
exogenous monetary policy shock.
2.2
Macroprudential Policy
In this paper, we take into account the classi…cation initially introduced by Blanchard et al. (2013). We assume that macroprudential policy is implemented
through two instruments: one directed toward the …nancial stability of the lender,
the other towards the borrower.
First regarding lenders, we assume that macroprudential policy account for a
ratio related to the Basel I-like capital requirement of the banking system (Pariès
et al., 2011)
BKi;t+1
KRBi;t = f
;
(1)
Lsi;t+1
where KRBi;t implies that the bank must pay a quadratic cost whenever the
capital-to-assets ratio BKi;t+1 =Lsi;t+1 moves away from a calibrated optimal target6 .
Second, regarding borrowers, we assume that macroprudential policy accounts
for the evolution of the loan-to-income ratio (Gelain et al., 2012),
LT Ii;t =
Ldi;t+1
Yi;t
Ldi
:
Yi
(2)
where Ldi;t+1 denotes the lending demand from the home private sector and Yi;t
is the output. Thus when LT Ii;t increases, it is interpreted by macroprudential
authorities as an excessive growth to credit in comparison to activity.
As noticed, in these two expressions ((1)) and ((2)), the value of loans that is
taken into account di¤ers in the two instruments, given the possibility of national
banks to engage in cross border lending. Thus, the instrument directed toward
6
Following Gerali et al. (2010) and (Pariès et al., 2011), the quadratic cost added in the bank
2
1
B
s
maximization program writes, ACi;t
BKi;t+1 =Lsi;t+1 BKi;t+1 . After maxi2 BKi =Li
L
mizing, the marginal cost of loans takes the form, M Ci;t
= f Et i;t+1 ; Rt + i KRBi;t where
B
KRBi;t = @ACi;t
=@Lsi;t+1 = BKi =Lsi
BKi;t+1 =Lsi;t+1
9
BKi;t+1 =Lsi;t+1
2
the lending side of the economy accounts for the supply of loans in the country,
while the instrument directed towards the borrowing side accounts for the fact that
national agents can borrow from di¤erent national sources. However, neglecting
the possibility for banks to engage in cross-border activities would make both
instruments substitutable in the evaluation of systemic risk at the national level.
As in Quint and Rabanal (2013), macroprudential instruments a¤ect the general equilibrium of the model through the lending conditions of commercial banks.
A tightening of credit conditions due to macroprudential measures will thus increase the interest rate faced by borrowers. In the presence of macroprudential regulations, the marginal cost of loans mcLi;t faced by borrowers is altered by KRBi;t
and LT Ii;t at a degree i and 'i respectively,
mcLi;t = f Et
i;t+1 ; Rt
+
i KRBi;t
(3)
+ 'i LT Ii;t :
The marginal cost is the main determinant of the credit rate in the economy7 .
As shown in Table 1, our analysis of the optimal level of coordination of macroprudential policies contrasts three situations: in the …rst situation (a), we allow
countries to set optimally all parameters, in the second situation (b), we impose a
symmetric value on bank capital ratios in the monetary union and in the third situation (c), we impose symmetry on all parameters of the macroprudential scheme.
All parameters are set to maximize the joint welfare of union participants. The
details of the scenarios are set in Table 1.
Scheme
Instrument
a ) Total National
b ) Partial Symmetric
c ) Total Symmetric
KRB h;t =
6 KRB f;t ; LT I h;t =
6 LT I f;t
KRB h;t =
6 KRB f;t ; LT I h;t =
6 LT I f;t
KRB h;t =
6 KRB f;t ; LT I h;t =
6 LT I f;t
Penalization
h
h
h
6=
=
=
f
f
f
; 'h =
6 'f
; 'h =
6 'f
; 'h = ' f
Table 1: Di¤erent Implementation Scenarios of Macroprudential Policy
3
The Suboptimality of Homogenous Macroprudential Decisions
This section summarizes the results of the analysis. These results are obtained
given the estimation of the model presented in appendix B. We present some clear
evidence of the sub optimality of macroprudential decisions aimed at imposing
homogenous practices to all countries participating to the Eurozone. We then
turn to the discussion of the enforceability of the optimal cooperative situation
using a Nash bargaining game.
7
As described in section 2, the new keynesian Phillips curve for credit rates implies that the
L
L
L
credit rates are determined by Ri;t
= f Et Ri;t+1
; mcL
i;t where mci;t is the marginal cost of
loans.
10
3.1
Winners and losers of di¤erent policy regimes in the
EMU
The results are reported in Table 2. The …rst part of this table evaluates the welfare
gains - in terms of permanent consumption - computed in a situation where banks
can engage in cross-border banking. The second part of this table reports …gures
for a benchmark situation that does not account for cross-border bank ‡ows. All
results are evaluated in terms of permanent consumption gains, with respect to
a benchmark situation where the central bank implements an optimal monetary
policy rule8 .
First, neglecting macroprudential decisions, the value of parameters obtained
for the optimal monetary rule gives high value to the parameter related to in‡ation
in the interest rate rule while it ignores output stabilization. This is coherent with
the …ndings of Schmitt-Grohé and Uribe (2007); Pariès et al. (2011); Quint and
Rabanal (2013). The estimated interest rate rule leads to a welfare loss that
represents -0.056% of permanent consumption in the monetary union.
Optimal Parameters
Taylor
y
Macroprudential
'h
'f
h
f
Optimal Rule
Empirical rule
2.88
1.96
0.00
-
-
0.16
-
-
With Banking Flows
a) total national
b) partial symmetric
c) total symmetric
2.88*
2.88*
2.88*
0*
0*
0*
2.59 30
30
18.60
0
0
Without Banking Flows
d ) total national
e) partial symmetric
f ) total symmetric
2.88*
2.88*
2.88*
0*
0*
0*
2.43 30
30
30
0
0
-
-
Unconditional
Consumption Gains (%)
Union Core Periph.
-0.056
-0.072
-0.036
0.05
0.05
0.02
0.886
0.821
0.690
-0.185
-0.176
-0.214
2.281
2.12
1.871
0.03
0.03
0.03
0.646
0.588
0.583
-0.045
-0.093
-0.087
1.550
1.479
1.460
Note: (*) denotes calibrated parameters on the optimal rule.
Table 2: Welfare ranking of macroprudential policy
Turning to the implementation of macroprudential policy measures, we compute the value of parameters i and 'i by maximizing the joint welfare of the two
8
In the quantitative simulation, we …rst search for weights attached to in‡ation
and GDP
y
growth
in the Taylor rule that gives the highest unconditional welfare of households. Here,
we maintain the autoregressive parameter of the policy rule R at its estimated value since it
has low e¤ects on welfare. Based on the grid search by 0:01 unit, we limit our attention to
y
policy coe¢ cients in the interval (1; 3] for , [0; 3] for
as Schmitt-Grohé and Uribe (2007),
and in the interval [0; 30] for macroprudential instruments i and 'i . Our grid search interval
for macroprudential instruments is consistent with the …ndings of Gerali et al. (2010). The
size of this interval is arbitrary, but but policy coe¢ cients larger than 30 would be di¢ cult to
communicate to policymakers or the public Schmitt-Grohé and Uribe (2007).
11
parts of the monetary union.
First in situation (a), imposing no homogenous decision on these parameters,
we …nd that there is a clear di¤erence between the two parts of the Eurozone,
as the optimal value for the parameters is much lower for the core country group
with h = 2:59 while f hits the higher bound. The capital constraint faced
by peripheral banks tends to be e¤ective at diminishing the explosive growth of
credit. Capital requirements helps in dampening the …nancial distress emanating
from …nancial shocks as they are a major source of macroeconomic ‡uctuation in
Periphery. Conversely, banks populating the core area do need important capital
constraint to mitigate the systemic risk.
With an homogenous treatment of countries regarding bank capital requirements in case (b), the parameter imposed to core counties is the same as the optimal national weight imposed to peripheral countries, which is clearly penalizing.
Finally, imposing homogenous macroprudential decisions on the two instruments
a¤ects the weight for both counties: it clearly increases the value of the parameter for core counties (that should get higher interest rate than needed, following
the …nancial developments in this part of the monetary union) while correcting
measures are lighter for the peripheral measures (the weight associated to both
instruments is almost half of the value that was required on national grounds),
thus leading to movements in the interest rates served to borrowers that are insu¢ cient to dampen all the …nancial disequilibrium in this part of the monetary
union.
In the last situation (c), unconditional consumption gains are the lowest: a
common rule imposed by the …nancial regulator is even more penalizing than in
the situation (b). As observed, the best outcome for the monetary union is the situation that account for national speci…cities in the setting of both lender and borrower parameters. This situation combines the coordination of decisions required
to dampen the externality coming the implementation of a national macroprudential decision with the fact that the decision takes into account national …nancial
developments in each part of the monetary union.
However, as reported, the national consequences of macroprudential measures
is heterogenous: peripheral countries are better o¤ (up to a gain of 2.281% of
permanent consumption in the case of a national setting of parameters) while
welfare diminishes in core countries (between -0.176% and -0.214% of permanent
consumption loss).
To evaluate the consequences of cross-border loans on these result, the last part
of Table 2 reports the increase in permanent consumption that would be obtained
in a situation of complete segmentation of the loan market. As noticed, the welfare
gains are reduced. This feature may be explained by the combination of two
factors: First, there is one supplementary …nancial friction in the monetary union;
Second, as underlined in the previous section, both macroprudential instruments
would account for the same loan value, which would make them closer substitutes.
It should be noted, that even if the relative ranking of the implementation
schemes is not a¤ected with respect to the situation with cross-border loans (imposing more symmetry in macroprudential measures is suboptimal), the two sit-
12
uation with symmetric features lead to very close results. Concentrating on the
situation of core countries (that are looser for all macroprudential policy experiments), we clearly observe that the optimal situation for core country in the
situation of cross-border loan is a situation where the parameter on bank capital
requirement is set symmetrically for all countries. This feature disappears once
neglecting cross-border banking.
3.2
Are Countries Encouraged to Coordinate Naturally?
A Nash bargaining game
As the optimal situation incurs losers and winners, a natural question arises
whether it can be reached following national incentives. To evaluate the enforceability of the optimal cooperative solution we use a Nash bargaining game. Since
our aim is to focus on how the macroprudential mandate should be shared among
EMU members, we propose to borrow a convenient concept from game theory
proposed by Nash Jr (1950, 1953). Following the bargaining theory, we examine
how the surplus generated by macroprudential policy will be split between EMU
participants when national governments may have incentive to set unilaterally
macroprudential instruments.
There are two participants in the game i = h; f (h for core and f for periphery).
Each player considers two possible strategies: coordination and no coordination.
If both players choose to coordinate macroprudential instruments, they maximize
the welfare index of the monetary union Wu . If one country i chooses to deviate
singly, it maximizes the welfare index of its country Wi to the detriment of the
monetary union. Payo¤s are in terms of households unconditional consumption9 .
As reported in Table 3 and Table 4, we …nd that each country would choose
to follow the non cooperative strategy because incentives to deviate from the
cooperative equilibrium are too large. In summary, federal institutions should
enforce the cooperative solution, as national decisions would lead to a second best
equilibrium. To reach the Pareto optimal equilibrium, it would thus be necessary
to enforce cooperation through institutional incentives, as this situation cannot
be reached on national incentives. We observe a welfare reversal when there is no
cross-border lending in the monetary union, as nations will choose to cooperate
naturally given the optimal national outcomes of the coordination situation. The
cooperative equilibrium is the optimal one both on aggregate and on national
features. As a conclusion, the possibility of bank to engage in cross-border loans
creates a situation that requires a strong federal action to impose coordinated
decisions on macroprudential measures.
4
Conclusion
This paper has discussed the optimality of imposing homogenous macroprudential
measures to all countries belonging to the Eurozone. Assessing the fact that
9
Unconditional consumption is evaluated as steady state in…nite streams of household consumption.
13
CORE
PERIPHERY
Coordination
No Coordination
f = 30; 'f = 0:05
f = 30; 'f = 0:04
Coordination
h = 2:59; 'h = 0
( -0.185, 2.281 )
0.886
( -0.204, 2.289 )
0.879
No Coordination
h = 7:95; 'h = 0
( -0.153, 2.194 )
0.866
( -0.17, 2.205 )
0.861
Note: Nash equilibrium is bold
Table 3: Nash Equilibrium Payo¤s Matrix (in terms of unconditional consumption
gains from leaving the simple optimal policy)
CORE
PERIPHERY
Coordination
No Coordination
f = 30; 'f = 0:03
f = 30; 'f = 0:0303
Note:
Coordination
h = 2:43; 'h = 0
( -0.045, 1.55 )
0.646
( -0.045 ,1.55 )
0.646
No Coordination
h = 0; 'h = 0
( -0.133, 1.613 )
0.624
( -0.042, 0.514 )
0.198
L
h
=
L
f
= 0 (no cross-border lending).
Table 4: Nash Equilibrium Payo¤s Matrix (in terms of unconditional consumption
gains from leaving the simple optimal policy) without cross-border lending
14
the implementation of the macroprudential mandate has been set on cooperative
decisions, we have developed a two-country DSGE model that accounts for some
major features of the Eurozone (key role of the banking system, cross-border
bank loans, heterogeneity in …nancial factors) to provide a quantitative evaluation
of welfare gains coming from a more symmetric treatment of countries in the
implementation of macroprudential measures.
We separate the Eurozone in two parts (the core and the periphery) and evaluate how a macroprudential policy combining two instruments (namely, bank capital requirements and the evolution of the loan-to-income ratio) should be implemented. We more particularly contrasted three possibilities, ranking from a
coordinated de…nition of macroprudential measures based on the choice of optimal national policy parameters to a homogeneous de…nition of all macroprudential
measures.
We …nd that optimal measures should be set cooperatively and should account
for a heterogeneous treatment of countries as this policy leads to the highest increase of welfare on average in the monetary union. However, it creates losers
(the core countries) and winners (the peripheral countries). Second, focusing on
national developments, we …nd that the optimal solution that minimizes the permanent consumption loss for the core country needs to account for a symmetrical
de…nition of the macroprudential parameter related to the bank capital requirement. Third, we …nd that the enforceability of the optimal cooperative solution
needs an institutional decision, as leaving the choice of the equilibrium to national
incentives leads to the choice of a non-cooperative outcome that is optimal on
national counties but suboptimal in terms of joint welfare in the monetary union.
These elements should be useful to analyze the proposed evolution of macroprudential policy organization in the Eurozone. There are important welfare gains in
implementing national adjusted macroprudential measures, but a federal institution is necessary to oblige EU members to cooperate. Without a federal constraint
on macroprudential policy implementation, the Euro members may choose not to
coordinate which may signi…cantly reduce the welfare of the Eurosystem.
15
References
Angelini, P., Neri, S., Panetta, F., 2012. Monetary and macroprudential policies.
Angeloni, I., Faia, E., 2013. Capital regulation and monetary policy with fragile banks. Journal
of Monetary Economics 60 (3), 311–324.
Bailliu, J. N., Meh, C. A., Zhang, Y., 2012. Macroprudential rules and monetary policy when
…nancial frictions matter. Bank of Canada.
Beau, D., Clerc, L., Mojon, B., 2012. Macro-prudential policy and the conduct of monetary
policy. Banque de France.
Benes, J., Kumhof, M., 2011. Risky bank lending and optimal capital adequacy regulation.
International Monetary Fund.
Bernanke, B., Gertler, M., Gilchrist, S., 1999. The …nancial accelerator in a quantitative business
cycle framework. Handbook of macroeconomics 1, 1341–1393.
Blanchard, O. J., Dell’Ariccia, M. G., Mauro, M. P., 2013. Rethinking macro policy II: getting
granular. International Monetary Fund.
Brooks, S. P., Gelman, A., 1998. General methods for monitoring convergence of iterative simulations. Journal of computational and graphical statistics 7 (4), 434–455.
Brunnermeier, M., De Gregorio, J., Eichengreen, B., El-Erian, M., Fraga, A., Ito, T., 2012.
Banks and cross-border capital ‡ows: Policy challenges and regulatory responses. Brookings
Committee on International Economic Policy and Reform.
Brzoza-Brzezina, M., Kolasa, M., Makarski, K., 2013. Macroprudential policy instruments and
economic imbalances in the euro area. Tech. rep.
Christiano, L., Motto, R., Rostagno, M., 2010. Financial factors in economic ‡uctuations.
Collard, F., Dellas, H., Diba, B., Loisel, O., 2012. Optimal Monetary and Prudential Policies.
Banque de France.
De Grauwe, P., 2010. The scienti…c foundation of dynamic stochastic general equilibrium (dsge)
models. Public choice 144 (3-4), 413–443.
De Larosière, J., 2009. The High-Level Group on …nancial supervision in the EU: report. European Commission.
Eyquem, A., Poutineau, J.-C., 2010. Markets integration and macroeconomic dispersion in a
monetary union. Louvain Economic Review 76 (1), 5–30.
Gelain, P., Lansing, K. J., Mendicino, C., 2012. House prices, credit growth, and excess volatility: Implications for monetary and macroprudential policy. International Journal of Central
Banking.
Gerali, A., Neri, S., Sessa, L., Signoretti, F. M., 2010. Credit and banking in a dsge model of
the euro area. Journal of Money, Credit and Banking 42 (s1), 107–141.
Gilchrist, S., Ortiz, A., Zakrajsek, E., 2009a. Credit risk and the macroeconomy: Evidence from
an estimated dsge model. Unpublished manuscript, Boston University.
Gilchrist, S., Sim, J., Zakrajsek, E., 2009b. Uncertainty, credit spreads, and investment dynamics. Tech. rep., Boston University mimeo.
16
Hirakata, N., Sudo, N., Ueda, K., 2009. Chained credit contracts and …nancial accelerators.
Tech. rep., Institute for Monetary and Economic Studies, Bank of Japan.
IMF, 2013. Implementing macroprudential policy - selected legal issues. Tech. rep., International
Monetary Fund.
Jondeau, E., Sahuc, J.-G., de France, B., 2006. Optimal monetary policy in an estimated DSGE
model of the euro area with cross-country heterogeneity. Banque de France.
Kannan, P., Rabanal, P., Scott, A., 2009. Monetary and macroprudential policy rules in a model
with house price booms. International Monetary Fund, Research Department.
Kincaid, G. R., Watson, M., 2013. The role of macroprudential policies and implications for international policy coordination. In: a Chatham House workshop and available at http://www.
sant. ox. ac. uk/pefm/publications. html.
Kolasa, M., 2009. Structural heterogeneity or asymmetric shocks? poland and the euro area
through the lens of a two-country dsge model. Economic Modelling 26 (6), 1245–1269.
Lambertini, L., Mendicino, C., Teresa Punzi, M., 2013. Leaning against boom–bust cycles in
credit and housing prices. Journal of Economic Dynamics and Control 37 (8), 1500–1522.
Lubik, T., Schorfheide, F., 2006. A bayesian look at the new open economy macroeconomics. In:
NBER Macroeconomics Annual 2005, Volume 20. MIT Press, pp. 313–382.
Medina, J. P., Roldós, J., 2013. Monetary and macroprudential policies to manage capital ‡ows.
forthcoming IMF/WP.
Nash, J., 1953. Two-person cooperative games. Econometrica: Journal of the Econometric Society, 128–140.
Nash Jr, J. F., 1950. The bargaining problem. Econometrica: Journal of the Econometric Society,
155–162.
Nier, E., Osi´nski, J., Jácome, L. I., Madrid, P., 2011. Towards e¤ective macroprudential policy
frameworks: An assessment of stylized institutional models. Tech. rep., International Monetary Fund.
Pariès, M. D., Sørensen, C. K., Rodriguez-Palenzuela, D., 2011. Macroeconomic propagation
under di¤erent regulatory regimes: Evidence from an estimated dsge model for the euro area.
International Journal of Central Banking 7 (4), 49–113.
Quint, D., Rabanal, P., 2013. Monetary and macroprudential policy in an estimated dsge model
of the euro area. No. 13/209. IMF Working Paper.
Rubio, M., Carrasco-Gallego, J. A., 2012. Macroprudential measures, housing markets, and
monetary policy. Tech. rep., CEPREMAP.
Schmitt-Grohé, S., Uribe, M., 2003. Closing small open economy models. Journal of international
Economics 61 (1), 163–185.
Schmitt-Grohé, S., Uribe, M., 2007. Optimal simple and implementable monetary and …scal
rules. Journal of monetary Economics 54 (6), 1702–1725.
Smets, F., Wouters, R., 2003. An estimated dynamic stochastic general equilibrium model of the
euro area. Journal of the European Economic Association 1 (5), 1123–1175.
Smets, F., Wouters, R., 2007. Shocks and frictions in us business cycles: A bayesian dsge approach. American Economic Review 97 (3).
17
Suh, H., 2014. Dichotomy between macroprudential policy and monetary policy on credit and
in‡ation. Economics Letters 122 (2), 144–149.
Ueda, K., 2012. Banking globalization and international business cycles: Cross-border chained
credit contracts and …nancial accelerators. Journal of International Economics 86 (1), 1–16.
18

Download Report