Download Discussion by K. Tsatsaronis

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Post–World War II economic expansion wikipedia , lookup

Transcript
Liquidity, inflation and asset prices in a timevarying framework for the euro-area
Paper by C Baumeister, E Durinck and G Peersman
Discussion by
Kostas Tsatsaronis
Bank for International Settlements
Towards an integrated macro-finance framework for monetary policy
NBB Conference
Brussels, 16 October 2008
1
Overview
 Main question: Look at the dynamic links between liquidity
(money) and other macro variables from a monetary policy
perspective.
• How do prices and quantities react to money shocks?
• Do these reactions differ conditionally on the broader
macro context ?
 Basic answers: Money does matter…
• …for inflation, output growth and real asset prices
• …in particular “narrow money” and credit
• …especially during a financial boom-bust cycle
2
General comments
 An central question in macro and very important for central
bankers
 Rich implications for inputs to policy decision making
 Brings to bear useful techniques:
• Time-varying VAR
• analysis of responses within different macro context
 Provides a lot of food for further thought
3
The workhorse: VAR
Endogenous variables:
Exogenous variables:
GDP growth
Period dummies: great
moderation post 1985
Inflation
Interest Rate
Equity volatility index
(high-low split)
Real asset price growth
Liquidity growth
Estimation:
1971-2005
Three lags
Choleski identification
4
Comment of Grumpy Old Discussant
 Why use the “synthetic” euro area data for such an
investigation?
 The euro area did not exist but for six out of 35 years in
the sample period
• Data artificially biased towards an average that may
mean little for each individual economy
 Focus on financial and monetary variables while ignoring
the flexibility of European exchange rates!
 Why not look at single countries, or Germany together with
its close monetary allies?
5
Further comments on the VAR
 Asset price volatility: maybe deviation from trend?
 How important is the ordering of the first variables?
• Especially the interest rate and asset price growth
 Three lags may be an issue
• Evidence that some of the mechanisms of interest are
long-fused
• Especially the “endogenous risk” component
6
Time-varying parameter VAR
 An interesting idea to capture more subtle shifts in
mechanisms
 Results are a little puzzling:
• Recently a liquidity shock leads to stronger output and
inflation response, despite the higher interest rate
• Evidence of a change in the nature of what “liquidity”
proxies for?
• Maybe worth to look into the M1 vs M3-M1 split
7
Analysis conditional on “states”
 An interesting idea to uncover regularities
• similar to split-sample regression but more flexible
• akin to quartile regressions in some cases where the
states refer to ranges for the LHS variable
 The use of estimated residuals as RHS variables could be
problematic, but I am not a purist
8
Conditional results
 The effect of liquidity shocks on output, inflation and real
asset prices is strengthened during asset price booms and
busts
 The liquidity effects during business cycle upswings are
not too pronounced except for property prices
 In high inflation regimes liquidity boosts nominal asset
prices and real property prices
9
Conditional results (cont’d)
 Policy should be concerned with the dynamics of asset
markets in assessing the response to liquidity shocks
 Could one interpret the asset price boom periods as
supply-driven, and business cycle boom periods as
demand driven episodes of increased liquidity and credit?
 What do we know about the periods that combine
characteristics?
10
Bottom line
I like the paper because:
 It presents different facets of the interactions between
money/credit and the macroeconomy
 It provides ground for more structured analysis of these
channels
I think that authors have to look deeper in:
 Explaining the patterns they have uncovered
 Making sure that the results are not influenced by the
“synthetic” nature of the data used
11