Transcript PowerPoint-Präsentation

The Transmission of Shocks to the Chinese Economy in a
Global Context: A Model-Based Approach
by
J. Baillu, P. Blagrave, O. Kamenik and D. Laxton
Discussion by Pascal Jacquinot (ECB)
Jerusalem, 27 October 2009
Why China does matter?
• China is the world’s second largest economy, still
behind the US but ahead the euro area and Japan.
• Its GDP has grown by 8.9% in 2009q3 (on yearly basis).
• Its investment has grown by around 25% over the past
ten years and 33% in 2009q3.
• China saves over 50% of its national income.
• Growth sustainability: China that constitutes an
unprecedented story of economy development is
foreseen regaining its pre-industrial revolution status as
the first largest economy!
Why China does matter?
• China is playing a key role in the most recent policy
debates:
 Increases in the prices of energy and other
commodities: the strong growth in China (together
with other emerging economies) may be source of
upward pressure on inflation.
 Risks of outsourcing manufacturing activities in China.
 International capital flows through China’s large scale
purchases of US Treasury bonds.
Why China does matter?
 Global Imbalances: China’s large current account
surplus and accumulation of foreign asset reserves.
Even if the US deficit started to correct, partly in
response to the turmoil in the sub-prime mortgage
market, the size of these imbalances remains large.
The key role of exchange rate
 Straub & Thimann (2009) using an global DSGE model
stressed:
 The model-based results indicate that persistent current
account surpluses in China cannot be rationalized by the
occurrence of permanent technology and labour supply
shocks.
 The understanding of the macroeconomic adjustment
process in China requires to mimic the effect of potential
inefficiencies.These inefficiencies can be potentially seen
as a by-product of fixed exchange rate regime…
The key role of exchange rate
 A large number of contributions address the issue of
the valuation of the RMB, mostly pointing to a large
undervaluation.
 The challenge (already considerable for advanced
economies with deep and liquid markets): assessing
China’s equilibrium exchange rate (economy in a
transition process with short data record).
 Some economists are arguing that an exchange rate
change will not have an impact on the current account
surplus, which is essentially driven by high domestic
savings.
The key role of exchange rate
 The exchange rate regime is related to inefficiencies in
the financial sector (interest rate regulation, massive
requirement reserves or obligation to bank to purchase
sterilised bills) and thus contributes to current account
surpluses.
 More flexible exchange rate regime would allow to
advance financial liberalisation and development that
might contribute to higher domestic absorption and
thus a less inflated current account surplus.
The modelling strategy
 DSGE models vs. more reduced form models.
 The strategy largely depends on the use of the model:
forecasting or policy analysis. Fitting the data or telling
a nice story.
 Projection exercises need estimation and regarding the
quality of the Chinese data: Bayesian techniques are the
most appropriated. And the model should be easy to
handle (small size).
 The problem is the quite limited number of variables
and shocks.
The modelling strategy
 For policy analysis (structural reforms) a more detailed
model is needed: the global DSGE strategy.
 Advantages: some important features that may be
important for emerging economies such as the
tradable sector, the share of agents that do not have
access to financial markets can be easily introduced.
 Problem: difficult (if not impossible) to estimate:
calibration.
 Financial channels are still inexistent in global DSGE.
 Alternative solution: taking parameters for US, EA and
Japan from estimated DSGE and imposing his own view
on China’s parameters.
The modelling strategy
 GPM is closer to the first strategy but has the
undeniable advantage of being simultaneously a global
model and estimated.
 It’s a nice compromise between sophisticated microfounded DSGE models and very simple semi-reduced
form models (where some ad-hoc features can also be
added for the financial aspects).
 GPM has four behavioural equations where all variables
are defined in deviation from their steady-states.
 Estimated using Bayesian techniques.
 This version: GPM-G3 (US, EA and JP) expended with
China
GPM-China
 Compared to the benchmark model:
 IS curve in open economy: with foreign partners output
gaps and bilateral exchange rates.The same in both.
 Phillips curve: extra oil-price inflation term.
 Monetary policy reaction function: deviations in the
nominal exchange rate from its time-varying target.
 UIP: an additional term to capture the effects of
movement in the equilibrium exchange rate on current
and expected exchange rate.
GPM-China
 Compared to the benchmark model (cont.):
 No labour market (Okun’s law / NAIRU) nor financial
market (Bank Lending Tightening variables): no reliable
data.
 Exogenous stochastic processes driving unobserved
(equilibrium) variables: the rate of change of the
equilibrium exchange rate depends on its own lagged
value and the deviation of the growth rate of potential
output from its steady state (to mimic productivity
growth in the tradable sector). Instead of a random walk
for the advanced economies.
GPM-China
 Questions (or clarifications):
 How do you capture (if possible) some stylised facts such
as: a very high saving ratio due to precautionary motives
(high unemployment rate, no social security) or financial
regulations (the absence of financial markets)?
 How do you deal with the change of regime in the
monetary policy moving from a dollar pegging
(abandoned in July 2005) to a controlled exchange rate
strategy? Via the extra exchange rate term?
 But how did you set the exchange rate target?
 The oil price is a random walk. Does it imply that a boom
in China has no impact on the oil market?
Estimation
 Estimations are rather convincing:
 More persistence in the output gap, less forwardlookingness in the Phillips curve.
 The difficulty to estimate the monetary policy rule.
 But are results very sensitive to:
 The choice of the priors (a comparison of prior and
posterior distributions would also be informative)?
 The data de-trending?
 Strong assumption: independence of shocks?
Results
 Impulse responses are also rather convincing. Even if
the behaviour of the Chinese exchange rate is
sometime difficult to understand.
 A couple of extra simulations would have also been
quite interesting: the impact of a Chinese demand (or
potential output) shock on the US economy.
Conclusion
 A very appealing paper that I liked very much
regarding the challenge to model (in a very convincing
way) China.
 Thank you very much…