Credit frictions and optimal monetary policy C

1
Credit frictions and optimal monetary
policyCúrdia and Woodford
Discussion Frank Smets
Towards an integrated macro-finance framework for
monetary policy analysis Brussels, 16-17 October
2008
2
Summary

• Nice, elegant (once you get through several pages
  of algebra) extension of the basic New Keynesian
  (NK) model
• Includes heterogeneous consumers (with different
  saving behaviour) which gives rise to lending and
  borrowing in equilibrium
• Financial intermediation is costly, giving rise
  to an external finance premium possibility of
  financial mark-up.
• As a result, the basic NK IS and Phillips curves
  are amended with a term (financial wedge) that
  acts like a cost-push shock.

3
Policy implications

• The principles of optimal monetary policy are not
  changed much.
• The NK targeting criteria is still very much in
  place identical with exogenous finance premium
• What has changed is the transmission process
  financial frictions do affect how shocks and
  policy affect output and inflation
• However, in the benchmark calibration these
  effects are quite small.
• So, more or less business as usual.

4
Modification of the IS and Phillips curves

• A higher wedge (lower deposit rate, higher
  borrowing rate) leads to more consumption by
  lender and less consumption by borrower. As
  lender consumes less, the aggregate effect on
  consumption is typically negative (but small).
• A higher wedge leads to lower labour supply by
  lenders and more work by borrower. As lenders
  work relatively more, the lower labour supply by
  lenders will dominate in the aggregate (but
  small).

5
Responses to a financial shock
6
What about the role of financial frictions?
7
What about the role of financial frictions?
8
What about the role of financial frictions?

• Two observations
• The variation in the interest rate spread is
  minimal in response to all aggregate shocks (only
  a few basis points). Is this just a calibration
  issue and thus solvable or a more fundamental
  problem? In reality, the external finance premium
  is much more volatile.
• There is a clear positive correlation between the
  interest rate spread and credit. In the data the
  finance premium is countercyclical, there is a
  negative correlation. This will be difficult to
  solve with different calibration.

9
Two main comments

• The CW form of financial friction does not appear
  compatible with the countercyclical external
  finance premium in the data?
• The premium is partly modelled as an exogenous
  mark-up and partly as covering an ad-hoc
  cost-function. The intermediation (monitoring)
  costs are a function of the amount of lending.
• What financial frictions may be appropriate?
• The financial sector is not explicitly modelled.
  There are no explicit banks that maximise profits
  and face asymmetric information problems, the
  possibility of defaults or capital adequacy
  constraints,

10
Some evidence (Christiano et al, 2008)

• External finance premium in the euro area
• Low in booms, high in recessions

11
Alternative financial frictions

• Financial accelerator and procyclical credit
  worthiness broad balance sheet effect
• Bank lending channel liquidity and capital of
  banks
• Risk-taking channel the attitude towards risk is
  procyclical.

12
Broad balance sheet effects

• Most modelling has focused on broad balance sheet
  effects (BGG, KM, Iacoviello, )
• Financial accelerator can lead to a
  countercyclical external finance premium
• Higher productivity, higher profitability, higher
  economic activity, higher asset prices increase
  the net worth of households and firms, which
  reduces the external finance premium,
• Pro-cyclical creditworthiness
• A booming economy implies lower risks of default,
  lower risk premia, higher asset prices, and
  lower risk premia.

13
Banks versus firms/households?

• Normal times limited bank lending channel
  securitisation further reduces this channel.
• Recently, most of the variation has been in
  premium paid by banks

14
Risk-taking channel
Maddaloni, Peydró-Alcalde and Scopel
(2008) Based on Bank Lending Survey data.
15
Different sectors respond differently

• Giannoni, Lenza and Reichlin (2008)
• Large BVAR with money/credit aggregates and
  associated interest rates
• Some findings
• Credit to non-financial firms increases following
  a monetary policy tightening credit to
  households falls. See also De Haan et al (2007).
• Lending rates are sticky
• Why?

16
Credit Aggregates
Note Dashed lines represent the 68 confidence
interval.
17
Lending rates
Note Dashed lines represent the 68 confidence
interval.
18
Conclusions

• Elegant paper
• Financial frictions do not appear to be
  quantitatively important and do not fundamentally
  change the principles of monetary policy
• Is likely to be model-dependent
• Other financial frictions (or shocks) are
  necessary to explain the procyclicality of the
  external finance premium
• Interaction with investment (rather than
  consumption) is missing.