Transcript Systemic Risk and Macroprudential Reguatlion

Systemic Risk and
Macroprudential Regulation
Gerald P. Dwyer
Clemson University
University of Carlos III, Madrid
Acknowledgement
• John Devereux
• James Lothian
• Margarita Samartín
Systemic Risk
• The G10 Report on Consolidation in the Financial
Sector (2001) suggested a working definition:
• "Systemic financial risk is the risk that an event
will trigger a loss of economic value or confidence
in, and attendant increases in uncertainly [sic]
about, a substantial portion of the financial
system that is serious enough to quite probably
have significant adverse effects on the real
economy."
Systemic Risk
• George G. Kaufman and Kenneth E. Scott
(2003) define "systemic risk" in imprecise
terms:
• "Systemic risk refers to the risk or probability
of breakdowns in an entire system, as
opposed to breakdowns in individual parts or
components, and is evidenced by
comovements (correlation) among most or all
the parts."
Systemic Risk
• Darryll Hendricks (2009), who is a practitioner,
suggests in a missive for the Pew Financial
Trust, a more theoretical definition from the
sciences:
• "A systemic risk is the risk of a phase transition
from one equilibrium to another, much less
optimal equilibrium, characterized by multiple
self-reinforcing feedback mechanisms making
it difficult to reverse."
Systemic Risk
• George G. Kaufman and Kenneth E. Scott
(2003) define "systemic risk" in imprecise
terms:
• "Systemic risk refers to the risk or probability
of breakdowns in an entire system, as
opposed to breakdowns in individual parts or
components, and is evidenced by
comovements (correlation) among most or all
the parts."
Macroprudential Regulation
• “The objective of a macroprudential approach
is to limit the risk of episodes of financial
distress with significant losses in terms of the
real output for the economy as a whole.”
(Borio 2003)
Forecasting
• Usefulness depends on ability to forecast
when crises will occur and when not
Types of Crises
• Banking crises
• Sovereign-debt crises
• Foreign exchange crises
Banking Crisis
• Losses at banks same order of magnitude as
equity capital in banking system
Sovereign and FX Crises
• These are state-created problems!
Government Debt to GDP
2000 to 2013
200
UK
Germany
Spain
180
France
Greece
US
Italy
Portugal
Ireland
160
140
120
100
80
60
40
20
0
2000
2001
2002
Source of Data: Eurostat
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
Definition of A Banking Crisis
• Laeven and Valencia (2013)
• Significant signs of financial distress in the banking system as indicated by
– significant bank runs
– losses in the banking system
– and/or bank liquidations
• Significant banking policy intervention measures in response to significant
losses in the banking system. At least three out of following six
–
–
–
–
deposit freezes and/or bank holidays
significant bank nationalizations
bank restructuring gross costs at least 3 percent of GDP
extensive liquidity support (5 percent of deposits and liabilities to
nonresidents)
– significant guarantees put in place
– significant asset purchases (at least 5 percent of GDP).
Data
•
•
•
•
Data on real GDP and banking crises for
21 countries generally from 1870 to 2009
2,950 annual observations
91 banking crises
Another Set of Data
•
•
•
•
Data on real GDP and banking crises for
176 countries from 1970 to 2011
7,392 annual observations
150 banking crises
Real GDP per Capita and Banking Crises
Source: Dwyer, Devereux, Baier and Tamura (2013)
Definition of Recession
• General discussion
– Bry and Boschan (NBER, 1971)
• Algorithm to determine contractions in economic activity
• Peak when increase followed by five monthly decreases
• Cycle at least 15 months long
– Harding and Pagan (2002)
• Peak when increase is followed by two quarters of decline in real GDP
• Partly reflects requirement that a contraction be at least six months
long
• Annual data
– Peak when an increase followed by a decrease in real GDP
• Peak when an increase followed by a decrease of real Gross
Domestic Product (GDP) per capita
– Jordà, Schularick and Taylor (2012) and others in crisis literature
How To Measure Output Loss
• Laeven and Valencia (2013) measure output
losses as the sum of the difference between
actual and trend GDP over four periods (T to
T+3)
– Expressed as a percentage of GDP for the starting
year of the crisis
• Compute trend GDP by applying an HP filter
Examples in which This Approach
Produces Plausible Results
The approach is plausible for economies with an apparent
or seeming underlying trend growth.
This is not the case for all economies.
Source: Devereux and Dwyer (2014)
GDP losses without a GDP Contraction
Country
Lebanon
Israel
Spain
Colombia
Swaziland
Turkey
Guinea-Bissau
Start
1990
1977
1977
1982
1995
1982
1995
Laeven and Valencia (2013)
End
1993
1977
1981
1982
1999
1984
1998
Output loss Laeven
and Valencia
102
76
59
47
46
35
30
40000
35000
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
An example
Figure Two
Lebanon 1990 (loss 102 percent)
Banking
Crisis
30000
25000
20000
15000
10000
5000
0
A second Example
Israel 1977 (77 percent of GDP)
90000
80000
Trend
Crisis
70000
60000
50000
40000
30000
20000
10000
1970
1971
1972
1973
1974
1975
1976
1977
1978
1979
1980
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
0
Isolated Cases?
• Very large output losses in Laeven and
Valencia (2013) indicate this is empirically
important
Largest Output Losses
in Laeven and Valencia (2013)
Banking Crisis
Country
Kuwait
Congo, DR
Burundi
Thailand
Jordan
Ireland
Start
1982
1991
1994
1997
1989
2008
End
1985
1994
1998
2000
1991
Output loss
Laeven and Valencia
143.4
129.5
121.2
109.3
106.4
106.0
Largest Output Loss
Figure Four
Kuwait 1982 (144 percent of GDP)
Source: Devereux and Dwyer (2014)
Measurement of Output Losses
• A trend in potential output for many countries is
uninformative because a stable trend growth of GDP
does not exist for countries without modern economic
growth
• An almost intractable problem since it holds true,
almost by definition, for most of the low income
economies
• A similar point can be made regarding Reinhart and
Rogoff’s (2009) use of GDP per capita
– Modern economic growth is relatively rare
– Venezuela/Argentina where income per capita takes from
twenty or over forty years to recover after a banking crisis
Association Between Banking Crises
and Contractions in Historical Data
Banking Crisis Start
Contraction in real
GDP per Capita in
Same Year
Yes
No
Yes
43
643
No
48
1949
Source: Dwyer, Devereux, Baier and Tamura (2013)
Association Between Banking Crises
and Contractions in Historical Data
Banking Crisis Start
Contraction in real
GDP per Capita in
Same Year or
Following Year
Yes
No
Yes
67
1038
No
24
1564
Source: Dwyer, Devereux, Baier and Tamura (2013)
Association Between Banking Crises
and Contractions in Historical Data
Banking Crisis Start
Contraction in real Yes
GDP per Capita in
No
Same Year or
Following Two Years
Yes
No
69
1318
22
1284
Source: Dwyer, Devereux, Baier and Tamura (2013)
Source: Dwyer, Devereux, Baier and Tamura (2013)
Densities of Mean Growth Rate of Real GDP per
Capita Before and After Banking Crises
Source: Dwyer, Devereux, Baier and Tamura (2013)
Recessions After Banking Crisis
in Post 1970 Data
Year of Beginning of Recession After Banking Crisis
Year of Crisis
54
Year After Crisis Two Years After No Recession
Crisis
even Two Years
After Crisis
41
1
44
Source: Devereux and Dwyer (2014)
Real GDP Growth and Banking Crises
in Post 1970 Data
Source: Devereux and Dwyer (2014)
United States Banking Crises
Source: Dwyer and Lothian (2012)
Frequency of Recessions
after Banking Crises
• Many banking crises are not associated with
decreases in real GDP
Frequency of Recessions
after Banking Crises
• Many banking crises are not associated with
decreases in real GDP
• Why big differences?
– Prior conditions and severity of problems
– Policies before and after crisis
Macroprudential Regulation Itself
• “The objective of a macroprudential approach
is to limit the risk of episodes of financial
distress with significant losses in terms of the
real output for the economy as a whole.”
(Borio 2003)
Macroprudential Regulation Itself
• “The objective of a macroprudential approach
is to limit the risk of episodes of financial
distress with significant losses in terms of the
real output for the economy as a whole.”
(Borio 2003)
• Can we predict with reasonable confidence
when these relatively infrequent events will
occur?
Is Macroprudential Regulation
Likely to Work?
• Different strategies
– Discretionary regulation
• Prevent banking crises
• Limit the cost if they occur
– Rule-based regulation
• Higher capital requirements at banks
• Different strategies
– Anticipation
• Prevent banking crises
– Resilience
• Limit the cost
Is Macroprudential Regulation
Likely to Work?
• Discretionary regulation
– Down-payment requirements for houses
• Rule-based regulation
– Higher capital requirements at banks
Housing and Financial Crisis
• Common theory: Rising housing prices created
bubbles in various countries and that was a
proximate cause of the recent banking crisis
• Implication for some: Limit rises in housing
prices
– Decreases probability of crisis
– For example, change down payment requirements
when housing market is over-heated
Changing Down Payment
Requirements
• Problem: This works by locking many people out
of the housing market
– Not obviously desirable
– Not obviously feasible in a democratic country
– Certainly will have consequences for the regulator
• General problem: Changing relative prices and
creating identifiable winners and losers in the
general population
– This is quite different than most rationales for
standard macroeconomic policy
Changing Down Payment
Requirements
• Problem: This works by locking many people out
of the housing market
– Not obviously desirable
– Not obviously feasible in a democratic country
– Certainly will have consequences for the regulator
• General problem: Changing relative prices and
creating identifiable winners and losers in the
general population
– This is quite different than most rationales for
standard macroeconomic policy
• Works quite differently in practice as well
Raising Capital Requirements
• Creates winners and losers of course
– Likely to be fewer or at least smaller banks
– Fewer financial services
– Quite possibly, fewer financial services
counteracts the provision of too many financial
services due to deposit insurance and too-big-tofail
– Makes financial system more resilient to losses
seen in past banking crises
Effective Anticipation
• What would it be good to have for
anticipation to be effective?
– Foresight into cause of next banking crisis
– Given infrequency, some estimate of when more
or less likely
– A firm foundation for connection between policies
and probability of a crisis
Effective Resilience
• What would it be good to have for resilience
to be effective?
– Foresight into likely factors that would lessen
effects as commonly seen
– Some estimate of when more or less likely if a
state-contingent policy is to be adopted
– Depending on the cost of resilience, a connection
between policies and severity of a crisis
Compare and Contrast Two Policies
• Anticipation – Head off a bubble in housing
prices
• Resilience – Require banks to have more
capital
Conclusion
• There is little doubt that events such as
banking crises, sovereign-debt crises and
possibly other “crises” are important
• Do we know enough to help avoid these
events or lower their costs?
• Is macroprudential regulation, with its
constant oversight, likely to be effective for
banking crises which may not happen once in
fifty years?
Conclusion
• Regulation is different than fiscal or monetary
policy in operation
– Public-interest and private-interest theories of
regulation
– I think an independent central bank engaging in
macroprudential regulation is unlikely and
undesirable in any case
A Quest for Stability
Source: http://img2.travelblog.org/Photos2/
One Way to Avoid Banking Crises