Transcript A study of the diversification of China`s foreign reserves

A three-country stock-flow consistent model:
A study of the diversification of
China's foreign reserves
Jun Zhao and Marc Lavoie
0
Outline

Introduction

Other work on diversification

Theoretical framework

Experiments and simulation results

Conclusion
1
Introduction (1)

China and Japan are by far the two biggest holders of
foreign reserves as well as the two biggest foreign holders
of US dollar securities.

China’s total foreign exchange reserves rose to US$1.7
trillion at the end of March 2008, the largest in the world.

A change of policy by the Chinese monetary authorities is
likely to have an impact on relative exchange rates and
economic activities throughout the world.
2
Introduction (2)

Some Chinese economists (Zheng and Yi , 2007; Wang,
2006) have emphasized that China’s actual reserves have
far exceeded its normal demand.

They claim that huge foreign exchange accumulation has
created large opportunity costs and wealth losses due to a
weakening dollar. They have suggested to diversify
foreign exchange reserves.

A sudden shift away from dollar assets could destabilize
financial markets and, in their view, put upward pressure
on US interest rates. They conclude that a gradual smallscale diversification would be beneficial and preferable.
3
Introduction (3)

In order to assess the possible qualitative effects of
diversification of the Chinese foreign reserves, we
propose the use of a stock-flow consistent model,
describing three countries each with its own currency.

The three countries are presumed to be China, the USA
and the rest of the world, which we will call Euroland.

A important feature of our model is the assumption of
imperfect asset substitutability.
4
Introduction (4)

Europe is assumed to be on a floating exchange rate
regime with the US and China, while China and the USA
are tied by a fixed exchange rate (although this is only partly true,
because China moved onto a managed float exchange rate regime in July
2005, with the renminbi having appreciated by about 13 percent relative to the
dollar. So, our model could work as well with such a trend of appreciation but
we may want to underline that such would only add a small complication to
the analytics without really adding substance.)

An important feature of our model is the assumption of
imperfect asset substitutability.
5
Introduction (5)

In this paper, we present three experiments dealing with
the diversification of China’s foreign reserves – i.e., a
shift away from dollar securities to euro securities:

The first experiment presents simulation results
arising from a sudden diversification of China’s
foreign reserves.
 The second experiment includes a set of tests for
parameter sensitivity.
 The third experiment presents an alternative closure
of the model by setting a target percentage
composition of China’s foreign reserves.
6
Other work on diversification (1)

Dooley, Folkerts-Landau, and Garder (2004) argue that a portfolio shift
in the foreign reserves of Asian countries would lead to euro
appreciation.

Blanchard et al. (2005) developed a simple four-region portfolio
balance model. They discussed the impact of a change in the
composition of reserves by Asian central banks. They found that the
path of adjustment was likely to be associated primarily with a further
appreciation of the euro against the dollar. An interesting result for the
authors is that it is not by itself a catastrophe for the U.S., while the
danger is more serious for Japan and Western Europe.

Using a simple three-asset-portfolio balance model, Dullien (2007) has
also examined this issue. He concludes that if Asian central banks
move from buying dollars towards buying euros, then the euro would
strongly appreciate against the dollar and the competitiveness of
European industries in world markets would be affected negatively.
7
Other work on diversification (2)

In what ways is our work different from that of Blanchard
et al. (2005) and also of Dullien (2007)?
• It is based on a stock-flow consistent approach;
• There are five sectors;
• All relevant variables are endogenous, such as GDP,
consumption, investment, taxes, debt servicing, the
money supply, imports, exports, the current account
and the capital account, the exchange rate, … ;
• The model is solved numerically, by making simulations;
• Both the short-run and long-run impacts of internal and
external shocks are examined.
8
Theoretical framework (1)

The stock-flow consistent framework takes into account
not only the stocks of assets and liabilities of the various
sectors of the economy, but also the flows of income,
investment and saving. “This framework makes it
essential to analyze the behavior of all parties of any
international transaction” (Godley and Lavoie 2004a, p.1).

Godley (1999), Izurieta (2003), Lequain (2003), Godley
and Lavoie (2005-06; 2007a; 2007b, ch. 12) applied this
framework to various open-economy models.
9
Theoretical framework (2) – the matrices
the U.S.A.
HouseItems
holds
Tangible K
Firms
Banks
Euroland
Govern-
Central
House-
ment
Bank
holds
K1
Banks
Govern-
Central
ment
Bank
Hh1
-Hs1
Deposit
M1
-M1
Bill1
Bh1,1
Bcmb1,1
Bill2
Bh1,2
Bill3
Bh1,3
House-holds
Hh2
-Hs2
M2
-B1
Bcb1,1
Bh2,1
Bcb1,2
Bh2,2
L1
Firms
-M2
Bcmb2,2
ment
Hh3
M3
-B2
-M3
SUM
0
0
0
-Vg1
0
-V2
0
0
0
0
0
Bh3,1
Bcb3,1
0
Bcb2,2
Bh3,2
Bcb3,2
0
Bcb3,3
0
L2
0
K
0
Bcmb3,3
-L3
-B3
L3
0
0
-V1
Sum
Bcb2,1
Advance
Balance
Centra Bankl
-Hs3
Bh3,3
-L2
Banks
K3
Bh2,3
-L1
Govern-
K2
Cash
Loan
Firms
China
-Vg2
0
-V3
0
0
0
0
0
A3
-Vg3
-Vcb3
0
0
0
V
0
10
Theoretical framework (3) – The Model
National income identity and trade

Nominal GDP:

Imports in each country are determined by the relevant
income and price elasticities, that is by domestic income
and the exchange rate.
11
The Model (Cont’d)
Households

Consumption is a function of households’ disposal
income and net wealth.

Households allocate their wealth among money, deposits,
and government bills.

The financial assets are imperfect substitutes. The
allocation of the assets varies in response to changes in
rates of return and risk consideration.
12
The Model (Cont’d)
Firms

The targeted capital stock depends on sales and the
target capital to sales ratio.

The target capital to sales ratio is inversely related to the
interest rate.

The net investment of firms is financed by loans from
commercial banks.
13
The Model (Cont’d)
Commercial banks





Commercial banks supply loans on demand to firms.
Commercial banks set deposit and lending rates by
marking down and marking up Treasury bill rates.
In China, in addition to the deposits from households, the
liabilities of commercial banks include advances from the
central bank of China.
In the U.S. and Euroland, besides loans to firms, the
assets of commercial banks include Treasury Bills.
In China, advances from the central bank play the role of
a buffer; in the U.S. and Euroland, Treasury bills play the
role of a buffer.
14
The Model (Cont’d)
The government

The newly issued Treasury bills are equal to government
expenditures (including debt servicing) minus
government revenues.

Pure government expenditures, that is government
expenditures excluding debt servicing, are an exogenous
variable.
15
The Model (Cont’d)
The central bank

Central banks set the interest rate on Treasury bills.

Central banks hold Treasury bills.
The central banks of China and Euroland also hold
foreign reserves in the form of U.S. Treasury bills
(initially).


Any monetization is demand-led.
16
 0
The Model (Cont’d)
China’s central bank

The main model – equations for China’s foreign reserves:
(73)
Bcb3d, 2,t    Bcb3d,1,t
Bcb3d,1,t  1  FR3,t
Where

 0
Bcb3d, 2,t   2  FR3,t

,
1
 

1  



1  
2
1 



 and



1   2  1
The alternative closure:
(73b)
 t    (  T   t 1 )
 0
17
Experiment 1:
At time T1, the U.S. increases its propensity to import goods from China, and
at time T2, China starts to diversify its foreign reserves.
Figure 1: Impact of the diversification of China’s foreign reserves on the exchange rate
1.1
1
0.9
0.8
Value of the dollar in
euro
0.7
0.6
0.5
T1
T2
18
Experiment 1:
Figure 2: Impact of the diversification of China’s foreign reserves on trade balances
T1
40
T2
China
30
20
10
Euroland
0
-10
-20
-30
the U.S.
-40
19
Experiment 1:
Figure 3: Impact of the diversification of China’s foreign reserves on GDP
760
China
660
560
Euroland
460
360
the U.S.
260
T1
T2
20
Experiment 2 – Sensitivity analysis:
At time T1, the U.S. increases its propensity to import goods from China, and
at time T2, China starts to diversify its foreign reserves.
Figure 4: Impact of changes of parameter (beta) on the value of the dollar in euro
 =0
1.0
 =0.2
0.9
 =0.4
0.8
 =0.6
 =0.8
 =1
0.7
0.6
0.5
0.4
0.3
0.2
T1
T2
21
Experiment 2 – Sensitivity analysis:
Figure 5: Impact of changes in the parameter (beta) on the trade balance of Euroland
50
 =0
0
 =0.4 
=0.6
 =0.2
 =0.8
-50
 =1
-100
-150
-200
-250
T1
T2
22
Experiment 2 – Sensitivity analysis:
Figure 6: Impact of changes in the parameter (beta) on GDP of Euroland
 =0
480
 =0.2
 =0.4
420
 =0.6
 =0.8
360
 =1.0
300
240
180
120
60
0
T1
T2
23
Experiment 3 – Alternative closure: Gradual diversification
Figure 7: Impact of the diversification on the share of the euro in the China’s reserves,
comparison to that in the main model
20
The main closure
Percentage
16
12
The alternative
closure
8
4
0
T2
24
Experiment 3 – Alternative closure: Gradual diversification
Figure 8: Impact of the diversification on the value of the dollar in terms of the euro,
compared to what occurs in the main closure
1.1
the main closure
1.0
0.9
0.8
the alternative
closure
0.7
0.6
T1
T2
25
Experiment 3 – Alternative closure: Gradual diversification
Figure 9: Impact of the diversification on the trade balance of Euroland
10
T1
T2
0
-10
-20
-30
the alternative
closure
-40
-50
the main closure
-60
-70
26
Experiment 3 – Alternative closure: Gradual diversification
Figure 10: Impact of the diversification on GDP of Euroland
520
the main closure
480
440
400
the alternative
closure
360
320
280
T1
T2
27
Conclusion (1)

The simulation results show that with the diversification
of China’s foreign reserves, the euro appreciates
against the dollar and RMB. There is an overshooting
effect on exchange rates.

China benefits from the diversification, so does the US
since there is also a mechanism by which the US
exports to China increase because of the US import
component of Chinese exports. However, the Euroland
economy slows down. These findings are consistent
with those obtained by Blanchard et. al (2005) and
Dullien (2007).

Through sensitivity analysis, we find that the larger the
diversification of China’s foreign reserves (the bigger
the share of the euro bills in China’s foreign reserves),
the larger the negative effects on the Euroland economy.
28
Conclusion (2)

Gradual diversification, in place of a sudden one-step
switch in the composition of reserves, can overcome the
overshooting effect on exchange rates. The value of the
euro against the dollar and RMB increases gradually to
a higher level.

The results on long-run exchange rates, GDP and trade
balances are qualitatively similar, but quantitatively
different in the alternative closure relative to the main
closure. In the long run, gradual diversification is more
favorable to the U.S. economy and less favorable to the
Euroland economy.
29
Conclusion (3)

What seems particularly interesting is that the models
uncover some sort of path dependence. How the
central bank of China achieves its target diversification
rate has an impact on the steady state values of the
model.

In other words, the transition path towards the
diversification target influences its long-run equilibrium.
The equilibrium levels of the key endogenous variables
(such as GDP and trade balance among others) depend
on the path taken to achieve the target diversification
rate.
30