PowerPointPrsentation

1
The equilibrium value of the world interest rate
Singapore July 2007 Domenico Giannone, ECB and
CEPR Michele Lenza, ECB Lucrezia Reichlin, ECB
and CEPR
2
Some facts and some questions

• Since the early 80s historical decline of
  long-run interest rate in all G7 countries, both
  nominal and real
• Less synchronization in the short term rate
  (policy rate)
• Recent years conundrum

3
Real and nominal rates 1970-2006
4
Ex-ante 10-year real interest rates (survey
based) a zoom since 1990
5
Nominal yields - 10 year a zoom since 1990
6
Nominal yields short and long- a zoom since 1990
7
Conundrum

• Global Factors Panel regression on OECD
  countries
• Baseline
• Controlling for a rough measure of global long
  term interest rate

Long term country j
Short term country j
Long term country j
Short term country j
Long term OECD
8
Conundrum
Long term country j
Short term country j
Long term country j
Short term country j
Long term OECD
9
This Paper

• Can the historical decline of the long-term
  interest rate be explained by an historical
  decline of the world equilibrium interest rate?
• World eq. interest rate the rate that clears the
  world capital market

10
A popular explanation

• Saving glut
• ? equilibrium world interest rate has declined
  to keep the world saving and investment in
  equilibrium

11
Problems

• The world interest rates is not observable
• Expectation of inflation
• Risk premia
• Exchange rate expectation
• Where is the real interest rate heading to?

12
Problems

• Difficult to assess cant know unless we are
  able to identify the demand and supply of capital
• Difficult in a standard situation
• In the present case even more difficult since we
  do not even observe the price (real rate)

13
Our Solution

• Can we quantify the importance of real factors in
  world interest rate dynamics?
• Idea
• extract information on the equilibrium real
  rate from the intetemporal consumption choices in
  different countries (consumption depends on the
  real rate)

14
The real story
All movements in the world real rate should
be reflected in expected consumption in the
world Why? If the real rate is decreasing, so
it is expected consumption growth ?people will
tend to consume more today than tomorrow when the
gains from postponing consumption are smaller,
that is when the real rate is low
15
Measurement extracting the real rate from data
on expected consumption growth

• We can exploit the fact that, in each country i
  we
• must have
• Extract rtw from this relationship, using data
  from
• G7 countries since 1970
• The model implies the restriction that the world
  rate is common across countries (common component
  in expected consumption growth)

16
Measurement

• Steps
• Estimate expected consumption using a model
  containing current account data for the G7
  imposing the Euler equation restrictions
• Validate the model out-of-sample and compare it
  with an unconstrained model
• Extract that component of expected consumption
  that is common to all countries
• ? ?irtw
• ? this is proportional to the world
  interest rate

17
Measurement
Identification problem we can only identify the
world interest rate up to a scaling parameter ?
which is the elasticity of the intertemporal
substitution for the average consumer in the G7
? Assume ? is equal to one
18
The world interest rate

19
Observations

• The worldwide decline in consumption points to a
• decline in the global real equilibrium interest
  rate
• Is this quantitatively relevant? It depends on
  the value of the G7 average of the elasticity of
  intertemporal substitution
• 1.5 decline if elasticity 1
• 3.0 if elasticity .5

20
Real interest rates and the world
equilibriumaverage elasticity of intertemporal
substitution 1
21
Real interest rates and the world
equilibriumaverage elasticity of intertemporal
substitution .5
22
Does the model capture consumption dynamics?
The forecasting evaluation shows that the model
works well, especially for Japan and continental
Europe
23
Consumption growth

24
And expected consumption growth

25
Country heterogeneity

• For each country, the equilibrium rate depends on
  its own elasticity ?this is estimated
• We can estimate the difference sensitivity of
  expected consumption growth to the world interest
  rate

26
Elasticities

27
Comments

• 2 Clusters
• Canada, the UK and the US low elasticity
• Japan and continental Europe high elasticity

28
Do relative prices matter?
They dont Out-of sample performance
evaluation of the model show that adding
relative prices does not matter
29
Conclusions

• Historical trends in the G7 interest rates are
  explained in part by historical decline of the
  equilibrium world real rate
• This information can be extracted from
  intertemporal consumption choices
• Some countrys heterogeneity on how expected
  consumption reacts to the world real interest rate