EMs and
Banking
L.Rossi
Introduction
The model
Banking Frictions and Endogenous Firms Exit
Conclusions
Future
Research
Lorenza Rossi
University of Pavia
MACFINROBODS First Consortium Scienti…c Workshop
London - 15-16 January 2015
Aim of the Paper
EMs and
Banking
L.Rossi
A DSGE model with:
endogenous …rms’entry and exit
monopolistically competitive banking sector
Introduction
The model
Conclusions
Future
Research
Study:
1
Model Dynamics (Today Presentation)
the relationship between …rms dynamics and banking
Endogenous versus Exogenous exit model
2
Macroprudential Policies and Model Estimation (Next
Step)
Basel II versus Basel III
the e¤ects of alternative TR targeting also capital
requirements
Motivation
EMs and
Banking
L.Rossi
Introduction
STARTING POINTS. Recent empirical evidence
The model
Conclusions
Future
Research
1. The role played by …rms dynamics in the business
cycle: propagating real shocks.
2. The role played by the banking sector: loans to …rms
increase in booms and decrease in downturns
3. The literature on …rms dynamics:
The exit rate is exogenous and constant =) The number
of …rms exiting the market is acyclical or mildly procyclical!
Financial market are perfect
Evidence on Firms Dynamics
EMs and
Banking
L.Rossi
CORRELATION between GDP and FIRMS ENTRY and
EXIT
Introduction
The model
1
Firm entry is procyclical. Jaimovich and Floetotto
(2008), Bilbiie, Ghironi, and Melitz (2012), Lewis V. and
Poilly C. (2010 JME).
2
Firm exit is countecyclical. Campbell (1998), Vilmi
(2009), Totzek (2009), Rossi L. (2014)
3
Campbell (1998): correlation between GDP growth and
exit rates is negative and stronger than that with entry
rates (0.28 versus -0.51)
4
Recent …nancial crisis: …rms’net change became
NEGATIVE from 2008 to 2010 (peak 2009)
Conclusions
Future
Research
Evidence on Bank Lending to Firms
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
FRED data on C&I loans:
about 15% reduction in between 2001 and 2003
25% reduction in between 2008 and 2010
C&I loan rate spread:
Countercyclical Bank Spread. (Hannan and Berger, 1991,
Leuvenstein et al. 2008)
Recent …nancial crisis: the C&I loan rate spread about 66
basis points higher (or 23% higher) than its long-term
average. Kwan (2010):
Bank Markup:
Countercyclical Bank markup (Dueker and Thornton 1997,
Cetorelli 2003, Aliaga-Diaz & Olivero 2006, Claessens et
al. 2000)
Main Results of the model: it replicates all these
stylized facts
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
The model with both endogenous …rms entry and exit
shows
Countercyclical exit (both of rates and n of …rms), in
line with the empirical evidence.
Countercyclical banks’markups, in line with the
empirical evidence.
Countercyclical loan spread, in line with the empirical
evidence. The loan spread is procyclical or acyclical in the
model with exogenous exit
Firms net change becomes negative in response to a
…nancial shock (bank capital shock)
Stronger propagation mechanism than the model with
exogenous exit
The Model
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Four agents:
1
Firms: Intermediate sector ! Endogenous …rms’exit
and entry
2
Households
3
Monopolistic Banking: Markup and Markdown
4
A Monetary Authority: set the interest rate
Future
Research
The Intermediate good-producing …rms
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
N …rms producing an intermediate good, i 2 (0, N )
Monopolistic competition with sticky prices
Two types of …rms:
1
2
New entrants
Incumbent …rms
The Intermediate good-producing sector
EMs and
Banking
L.Rossi
Introduction
The production function of …rm i is
The model
Conclusions
Future
Research
yiI,t = At zi ,t li ,t
At : aggregate productivity
li ,t : labor hours employed by …rm i;
zi ,t is a …rm speci…c productivity, as in Ghironi and
Melitz (2005) .
Entry and Exit
EMs and
Banking
L.Rossi
Introduction
Timing:
The model
Conclusions
Future
Research
New Entrants
Entry condition:
ṽt = f E
(1)
Upon entry, …rms draw their productivity zi from the
common Pareto distribution with CDF
zmin ξ
G (zi ) = 1
with support zi 2 [zmin , ∞).
zi
Entry and Exit
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
Timing:
Incumbents
borrow bi ,t from the banking sector to pay the …xed cost
fF
Firms exit: before producing, if zi ,t < z̄t
z̄t is cut-o¤ level: the one that equals pro…ts to zero.
Average productivity:
1
G (z̄t )
Z ∞
1
θ 1
ziθ,t 1 g
(zi ) dzi
Endogenous death probability: η t = 1
zmin
zt
z̃t
with:
1
z̄t
d z̃t
>0
d z̄t
ξ
Firms Dynamics
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
NtE : new entrants at t
Nt : …rms already producing at t
Time to build. Entrants in period t start producing in
period t + 1
The law of motion of …rms is:
) Nt + 1 = 1
with η t +1 = 1
zmin
z t +1
ξ
η t +1
Nt + NtE
Households
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
Builds on BGM (2012). Households maximize:
"
#
1 +φ
∞
Lt
t P
max E0 ∑ β εt ln Ct
1+φ
t =0
Households: consume, work, decide how much to invest in
new …rms and in the shares of incumbent …rms and how
much to deposit to the banking sector.
Retail Branch for Loans
EMs and
Banking
L.Rossi
The optimal interest rate on Loans, under ‡exible rate
Introduction
The model
Conclusions
Future
Research
Exiting Firms do not pay back the interest rates on
loans
εbt
rtb = b
rt
ε t 1 1 η t +1
rtb is a mark-up over marginal costs, rt .
Exiting Firms do not pay back both the interest rates
and the notional value of loans
rtb =
εb
εb
1 1
η t +1
rt + η t +1
Retail Branch for Deposits
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
The optimal interest rate for deposits, under ‡exible rate is
rtd =
εd
εd
1
rt
the interest rate on deposits is mark-down over the policy
rate rt . Remember the non-arbitrage condition!
Central Bank
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
The central bank sets the interest rate prevailing in the
interbank market rt using the following Taylor rule
ln
(1
= φR ln 1 +1 +rtr 1 +
h
φR ) φπ ln ππt + φy ln
1 +r t
1 +r
Yt
Y
i
+ ln εrt
Business Cycle Analysis
EMs and
Banking
L.Rossi
Introduction
The model
Impulse response functions to:
productivity shocks
shocks to bank capital
Conclusions
Future
Research
We compare the performance of three models:
1
2
3
a standard DSGE model with a …xed number of …rms:
Constant Firms
a model with …rms exogenous exit: Exogenous Exit
a model with …rms endogenous exit: Endogenous Exit
Positive Technology Shock
EMs and
Banking
L.Rossi
Introduction
Firms (Loans)
N ew entrants
8
The model
Conclusions
Future
Research
E xiting firms
100
6
50
50
0
0
-5 0
4
2
0
0
5
10
15
20
25
30
-5 0
0
5
10
Output
15
20
25
30
-1 0 0
0
5
4
15
20
S pread r - r
0
3
10
b
Inflation
25
30
25
30
25
30
d
0
-1
-5
-2
-1 0
2
1
0
0
5
10
15
20
25
30
-3
0
5
15
20
25
30
0
5
10
15
20
0
5
25
30
-2 0
-5
-4 0
-1 0
-8 0
10
15
20
Interest R ate on Loans
0
-6 0
0
-1 5
S pread r - r
0
-1
-2
10
b
B ank Markup
1
-1 5
0
5
Ex o g e n o u s Ex i t
10
15
20
Co n s ta n t Fi rm s
25
30
-2 0
0
En d o g e n o u s Ex i t
5
10
15
20
Negative Bank Capital Shock
EMs and
Banking
L.Rossi
Introduction
Firms (Loans)
N ew entrants
0
The model
-0 .0 1
Conclusions
Future
Research
E xiting firms
0 .1
0 .3
0
0 .2
-0 .1
-0 .0 2
-0 .0 3
0
5
10
15
20
25
30
0 .1
-0 .2
0
-0 .3
-0 .1
0
5
10
Output
15
20
25
30
0
5
10
15
20
b
Inflation
0 .0 1
S pread r -r
0 .0 1
25
30
25
30
25
30
d
0 .0 6
0
0 .0 4
0 .0 0 5
-0 .0 1
-0 .0 2
0 .0 2
0
5
15
20
25
30
0
0
5
x 10
10
15
20
25
30
0
0
S pread r - r
0 .0 8
6
0 .3
0 .0 6
4
0 .2
0 .0 4
2
0 .1
0
5
10
15
20
25
30
0
0
10
15
20
Interest R ate on Loans
0 .4
0
5
b
B ank Markup
-3
8
10
0 .0 2
5
Ex o g e n o u s Ex i t
10
15
20
Co n s ta n t Fi rm s
25
30
0
0
En d o g e n o u s Ex i t
5
10
15
20
Unconditional BC Correlations with Output (US
data 1992q3-2013q4)
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
Correlations with Y
Loans
Exit rate
Exit Number
Bank Markup
Loan Spread
Data
0.25
-0.34 (-0.36)
-0.10 (-0.58)
-0.37(-0.29)
-0.20
Endo Exit
0.55
-0.30
-0.32
-0.40
-0.78
Exo Exit
0.26
0
0.26
0
-0.92
Conclusions
EMs and
Banking
L.Rossi
Introduction
The model
Conclusions
Future
Research
Models with endogenous …rms exit and monopolistic
banking:
stronger propagation mechanism in response to both real
and …nancial variables than a standard DSGE as well as a
model with exogenous exit.
endogenous countercyclical exit of both rates and number
of …rms in line with the empirical evidence;
endogenous countercyclical banks markup in line with the
empirical evidence;
endogenous countercyclical spread in line with the
empirical evidence.
Firms net change becomes negative in response to a
…nancial shock (bank capital shock)
Future Research
EMs and
Banking
L.Rossi
Introduction
The model
Macroprudential
Conclusions
Alternative borrowing mechanism
Future
Research
Comparison of model IRFs with SVAR IRFs in response to
a credit spread shock or TFP shock?
Bayesian Estimation of the Model with Endogenous versus
Exogenous Exit
Welfare: Optimal Monetary and Macroprudential Policy