Banca d’Italia
SUERF/Deutsche Bundesbank/IMFS Conference
The ESRB at 1
Berlin, 8-9 November 2011
The usual disclaimer applies
1.
The financial crisis and the debate on modelling
2.
Towards a new framework?
3.
Monetary and macroprudential policies in a model with financial intermediation
Pre-crisis consensus on macro modelling
•
New Keynesian framework
Representative agent cash-less economy
nominal rigidities → role for monetary policy
no financial frictions and no intermediaries
• “ The paradigm that has emerged […] is one that is clearly applicable to normal times […] in developed, stable economies ”, J. Galí (interview for EABCN, 2009)
•
Estimated models (e.g. Smets and Wouters, 2007) used for policy analysis (e.g. RAMSES at Riksbank )
The financial crisis
•
The crisis showed how
important are the links between financial markets and the real economy
many of the assumptions that characterized the new Keynesian framework were wrong
financial markets are far from being efficient
financial markets matter in originating and propagating shocks
Intense debate on the lessons of the crisis for economics and economic modelling
•
Buiter, Goodhart, Cecchetti, Spaventa and De
Grauwe to mention some of critics of DSGE models
•
Main missing elements :
financial intermediation
insolvency and default
liquidity
regulation of intermediaries and markets
booms and busts in asset markets
The crisis as an opportunity
•
Crisis as opportunity to modify current framework
•
Since Kiyotaki and Moore (1997) and Bernanke et al . (1999), few papers have considered financial intermediation in general equilibrium
•
Intensive research ongoing since 2009
Angeloni and Faia, Curdia and Woodford, Gertler and
Kiyotaki, Bianchi and Mendoza, Jeanne and Korinek,…
•
They all fall short of modelling systemic risk
Towards a new framework?
•
Setting up a new framework that takes into account the critiques that have been raised will require time
•
Policy-makers are confronted with questions that require timely answers
•
Researchers in both academia and central banks to cooperate and develop new models
•
In the meantime, one possibility is to modify current models and use them for policy analysis
Monetary and macroprudential policies in a model with financial intermediation
I use the model developed in Gerali et al . (2010) and modified in Angelini et al.
(2011) to answer questions related to monetary and macroprudential policies
1) What was the impact of the financial crisis on economic activity in the euro area?
2) Should monetary and macroprudential policies co-operate? (BI Discussion paper, no. 801, 2011)
3) Should macroprudential policy lean against financial cycles?
The model
•
Based on Gerali, Neri, Sessa and Signoretti (2010)
“ Credit and Banking in a DSGE model of the euro area ”
•
Medium-scale model with:
real and nominal rigidities (Smets and Wouters, 2007)
financial frictions à la Kiyotaki and Moore (1997)
monopolistic competition in banking sector
slow adjustment of bank rates to policy rate
role for bank capital
time-varying risk weights in bank capital regulation
policy rule for bank capital requirement
The model (cont’d)
•
Project started in September 2007
•
Model has been estimated using Bayesian methods and data for the euro area over the period 1998-2009
•
The model has also been used to study:
impact of a credit crunch on euro-area economy
impact of higher capital requirements (Basel 3)
•
Model has some of the limitations that I have discussed
Modelling monetary and macroprudential policies
•
Monetary policy:
interest rate rule à la Taylor
R t
1
R
1
R
t
y
y t
y t
1
R
R t
1
•
Macroprudential policy:
bank capital requirements rule
t
1
1
X t
t
1
X t can be any macroeconomic variable relevant for macroprudential authority
What was the impact of the financial crisis on economic activity in the euro area?
GDP (% dev. from steady state) Investment (% dev. from steady state)
•
The recession in
1.5
4
3
2 2009 was almost 1
1
0.5
entirely caused by 0
0
-1 adverse shocks to
-0.5
-2
-3
-1 banking sector
•
The sharp
-1.5
07Q1 07Q3 08Q1 08Q3 09Q1 09Q3 10Q1 10Q3
-4
-5
07Q1 07Q3 08Q1 08Q3 09Q1 09Q3 10Q1 10Q3 reduction of policy
Consumption (% dev. from steady state) Eonia rate (p.p dev. from steady state) rates attenuated
1.5
2
1.5
the strong and
1.0
1
0.5
negative effect of the crisis on the euro-area economy
0.5
0.0
-0.5
-1.0
07Q1 07Q3 08Q1 08Q3 09Q1 09Q3 10Q1 10Q3
0
-0.5
-1
-1.5
-2
07Q1 07Q3 08Q1 08Q3 09Q1 09Q3 10Q1 10Q3 macroeconomic monetary policy banking total
Should monetary and macroprudential policies co-operate?
• In “normal” times macroprudential policy yields small benefits
•
If two authorities do not cooperate, policy tools are extremely volatile
•
Benefits are sizeable when economy is hit by financial shocks and when two authorities cooperate
0.2
0.0
-0.2
-0.4
-0.6
0
9.1
9.0
8.9
8.8
8.7
0
9.00
8.75
8.50
8.25
8.00
0
Capital requirements
10 20
Capital/assets ratio
30
10 20
Loans-to-output ratio
30
40
40
0.02
0.00
-0.02
-0.04
-0.06
-0.08
0
0.0
-0.1
0.1
0.0
0.3
0.2
-0.1
0
Policy rate
10 20
Loan rate
30
10 20
Output
30
-0.2
10 20 quarters after shock
30
Cooperative
40
-0.3
0
Non cooperative
10 20 quarters after shock
30
Only monetary policy
40
40
40
Should macroprudential policy lean against financial cycles?
•
Agents expect a reduction in aggregate risk in one year time
•
For a given target for leverage, this provides an incentive for banks to increase lending
•
Shock does not materialize
•
Tighter capital requirements can be effective in containing expansion of lending
9.50
9.40
9.30
9.20
9.10
9.00
8.90
0
0.25
0.20
0.15
0.10
0.05
0.00
0
5
Capital requirements
10 15
Output (% dev. from steady state)
5 10 quarters after shock
15
20
20
0.2
0.0
-0.2
-0.4
-0.6
0
Policy rate (dev. from steady state)
5 10 15 20
2.0
Loans-to-output ratio (dev. from steady state)
Active macroprudential policy
No macroprudential policy
1.5
1.0
0.5
0.0
0 5 10 quarters after shock
15 20
Implications for monetary and macroprudential policies
Aggressive monetary policy can help mitigating negative impact of shocks to banking sector
Monetary and macroprudential policies should closely co-operate
Benefits of macroprudential policy can be sizeable when economy is hit by financial shocks
Risk of coordination failure
Macroprudential policy can be effective in leaning against financial cycles
Conclusions
•
DSGE models have undergone severe criticism
•
No doubt that models must be improved, but working alternative missing
•
Intensive research ongoing
•
Modeling systemic risk is key
•
Meanwhile, we can adapt current models with a role for financial intermediation to address questions related to monetary and macroprudential policies