Macroprudential Policy Evaluation using Credit Register Big Data João Barata Ribeiro Blanco Barroso Banco Central do Brasil – Research Department Based on the country contributions to the BIS CCA CGDFS Working Group on “The impact of macroprudential policies: na empirical analysis using credit register data”. The Brazilian team is Douglas Araujo, João Barroso, Carlos Cinelli, Bernardus von Doornik and Rodrigo Gonzalez.
21
Embed
Macroprudential Policy Evaluation using Credit Register Big … · Macroprudential Policy Evaluation using Credit Register Big Data João Barata Ribeiro Blanco Barroso Banco Central
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Macroprudential Policy Evaluation using
Credit Register Big Data
João Barata Ribeiro Blanco Barroso
Banco Central do Brasil – Research Department
Based on the country contributions to the BIS CCA CGDFS Working Group on “The impact of
macroprudential policies: na empirical analysis using credit register data”. The Brazilian team is
Douglas Araujo, João Barroso, Carlos Cinelli, Bernardus von Doornik and Rodrigo Gonzalez.
The views expressed in this
work are those of the author
and do not necessarily reflect
those of the Banco Central do
Brasil nor of its members.
Motivation
Financial cycles and business cycles do not necessarily coincide.
This creates opportunities to use different tools to address different
cycles (with care for possible interactions)
Macroprudential policies for the financial cycle
Monetary policy for the business cycle
With global financial integration, global liquidity and global risk
aversion are major factors in financial cycles for many economies
As a result: simultaneous macroprudential policy experiments in
several economies, around and after the global financial crisis
Evaluation is a challenging because econometric identification is
challenging: possible benefit of big credit register data.
Motivation
Credit register data allows identification of the credit operations of
several banks with the same agent (mostly firms in our case).
The common component is a proxy for the demand for credit
One can identify the effect of the policy on the supply of credit.
Computationally challenging when the time series dimension is
included in the model: several millions of observations in a dataset.
The challenge was undertaken by several Central Banks in the
Americas in a working group under the auspices of the Bank of
International Settlement, Americas Office.
Macroprudential Policies in the Americas
Canada: LTV
housing
Colombia:
Dynamic
Provisioning
Colombia:
Countercyclical
Reserve
Requirements
Colombia:
Limits on
exchange rate
risk
Argentina:
Liquidity Ratios
Canada: LTV
housing
Colombia:
Limits on
Dividend
Distribution
Colombia:
Liquidity Ratios
Brazil:
Countercyclical
reserve
requirements
Peru: Dynamic
Provisioning
US: SCAP
capital
assessement
program
Argentina: Capital
Buffer
Brazil: Risk
Weight on specific
loans
Brazil:
Countercyclical
reserve
requirements
Canada: LTV
housing
Peru:
Countercyclical
Reserve
Requirements
Peru: Limits on
exchange rate risk
Brazil:
Countercyclical
reserve
requirements
Brazil: Risk Weight
on auto loans
Canada: LTV
housing
Colombia: Limits
on derivatives
Mexico:
Provisioning on
Expected Losses
Peru:
Countercyclical
Reserve
Requirements
Peru: Limits on
exchange rate risk
Argentina:
Capital Buffer
Brazil:
Countercyclical
reserve
requirements
Canada: LTV
housing
Chile: Warning
of house prices
Peru: Liquidity
Ratios
Brazil: LTV cap
on housing
loans
Chile: Warning
of house prices
2007 2008 2009 2010 2011 2012 2013
pre-crisis buble global financial
crisis
begining of QE
policies
deepening of QE
policies
deepening of QE +
euro crisis
deepening of QE
+ euro crisis
deepening of QE
+ taper tantrum
Canada: LTV
housing
Colombia:
Limits on
exchange rate
risk
Canada: LTV
housing
Canada: LTV
housing
Peru: Limits on
exchange rate
risk
Canada: LTV
housing
Colombia: Limit
derivatives
Peru: Limits on
forex
Canada: LTV
housing
Chile: Warning
of house prices
Brazil: LTV cap
on housing
loans
Chile: Warning
of house prices
Colombia:
Countercyclical
Reserve
Requirements
Argentina:
Liquidity Ratios
Colombia:
Liquidity Ratios
Brazil:
Countercyclical
reserve
requirements
Brazil:
Countercyclical
reserve
requirements
Peru:
Countercyclical
reserve
Requirements
Brazil:
Countercyclical
reserve
requirements
Peru:
Countercyclical
reserve
Requirements
Brazil:
Countercyclical
reserve
requirements
Peru: Liquidity
Ratios
Colombia:
Dynamic
Provisioning
Colombia:
Limits on
Dividend
Distribution
Peru: Dynamic
Provisioning
US: SCAP
capital
assessement
program
Argentina:
Capital Buffer
Brazil: Risk
weight auto
loans
Brazil: Risk
Weight auto
loans
Mexico:
Provisioning on
expected
losses
Argentina:
Capital Buffer
2007 2008 2009 2010 2011 2012 2013
Asset Based Instruments
Liquidity Based Instruments
Capital Based Instruments
Capital based instruments affect risk-taking incentives, since it
impacts how much ‘skin in the game’ financial intermediaries have.
Similar risk-taking channel in monetary policy; see Barroso, Souza
and Guerra (2016) for the Brazilian case.
Liquidity/Liability based instruments affect the cost of funding of
financial intermediaries, and therefore credit supply conditions.
Monetary policy also affects the cost of funding, basically through
the same channel.
Asset based instruments affect the budget set of borrowers
Monetary policy affects the intertemporal budget constraint
Complementarity is at leas additive, possibly multiplicative
Relation with monetary policy
Initial findings of the working group based on time series
identification
This means banks are assumed to be homogeneously affected
by the policy, but heterogeneously so in time according to policy
intensity
So far only the Brazilian team presented results based on cross-
section identification (on top of baseline time series results).
This means banks are heterogeneously affected by the policy,
although homogeneously in time according to a policy elasticity
The results for both strategies are consistent in the Brazilian case,
therefore strenghening the overall results of the group.
Time Series x Cross-section identification
Findings: Capital Based Instrument
Colombia: Dynamic provisioning had a negative effect on credit growth,
with complementarities between monetary and macroprudential policy.
Argentina: Capital buffer are effective to reduce credit cycles
United States: Bank capital stress test in 2011 effective in reducing
credit in the jumbo mortgage segment
Mexico: Introduction of Provisioning based on expected loss reduced
credit growth, particularly in local currency
Peru: Dynamic provisioning has a significant effect on credit growth
Brazil*: Conditional risk weights contributed to increase the spread
charged to affected borrowers (Martins and Schechtman, 2010)
Findings: Liquidity Based Instruments
Colombia: Countercyclical reserve requirements had negative effect on
credit growth and complementarity with monetary policy.
Peru: Conditional reserve requirements on foreign currency deposits