The Chinese Economy

ISSN: 1097-1475 (Print) 1558-0954 (Online) Journal homepage: http://www.tandfonline.com/loi/mces20

Dynamics of the China-United Kingdom

Commodity Trade

Mohsen Bahmani-Oskooee & Ruixin Zhang

To cite this article: Mohsen Bahmani-Oskooee & Ruixin Zhang (2014) Dynamics of the China-
United Kingdom Commodity Trade, The Chinese Economy, 47:2, 75-93

To link to this article: http://dx.doi.org/10.2753/CES1097-1475470204

Published online: 05 Dec 2014.

Submit your article to this journal

Article views: 9

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=mces20

Download by: [Tulane University] Date: 01 April 2016, At: 21:13

 The Chinese Economy, vol. 47, no. 2, MarchApril 2014, pp. 7593.
2014 M.E. Sharpe, Inc. All rights reserved. Permissions: www.copyright.com
ISSN 10971475 (print)/ISSN 15580954 (online)
DOI: 10.2753/CES1097-1475470204

Mohsen Bahmani-Oskooee and Ruixin Zhang

Dynamics of the ChinaUnited

Downloaded by [Tulane University] at 21:13 01 April 2016

Kingdom Commodity Trade

Is There an S-Curve Effect?

Abstract: The S-curve phenomenon postulates that while past values of the trade
balance and current exchange rates are negatively correlated, future values of the
trade balance and current exchange rates are positively correlated. We investigated
this pattern between China and the United Kingdom. When we use the bilateral trade
flows between the two countries, we find no support for the S-curve. However, when
we disaggregate the bilateral trade data by industry and construct the S-curve for
47 industries that trade between the two countries, we find support for the S-curve
in 12 cases. These 12 industries, which conduct 30 percent of the trade (including
the largest industry), could benefit from currency depreciation.

Ever since China decided to adopt trade liberalization policies, it has enjoyed
increased exports and significant economic growth. While the United States is
Chinas major trading partner, other countries have begun to enjoy similar stand-
ing over time. Table 1 identifies Chinas export markets using export shares for
the year 2011.
China is often accused of manipulating the exchange value of the renminbi
(RMB) in order to keep Chinese goods competitive in the world market and con-
tinue to enjoy increased exports. Indeed, there is ongoing political pressure by the
United States to convince China to float its currency, with the expectation that if it
does so the dollar will depreciate and thus the U.S. trade deficit with China will be
reduced. Clearly, if the U.S. dollar depreciates against the RMB due to arbitrage

Mohsen Bahmani-Oskooee is a professor at the Center for Research on International Eco-

nomics and the Department of Economics, University of WisconsinMilwaukee, WI; email:
bahmani@uwm.edu. Ruixin Zhang is a professor at the Center for Research on International
Economics and the Department of Economics, the University of WisconsinMilwaukee.
Valuable comments from an anonymous referee are greatly appreciated.

75
 76 The Chinese Economy

activities, other currencies will also depreciate, which can affect trade flows between
China and other partners.1
There are now two approaches to assess the impact of currency depreciation on
a countrys trade flows or trade balance. One approach relies upon a reduced-form
model and regression analysis; the other concentrates on cross-correlation coef-
ficients between the trade balance and the real exchange rate. The first approach
is referred to as the J-curve, the second as the S-curve. The S-curve basically
postulates that while past values of the trade balance and current exchange rates
are negatively correlated, future values of the trade balance and current exchange
rates are positively correlated. The theory behind both approaches is the same and
is mostly based on the adjustment lags between exchange rate changes and the
Downloaded by [Tulane University] at 21:13 01 April 2016

trade balance (Bahmani-Oskooee and Hegerty 2010).

Different researchers have tested the S-curve using trade data at different ag-
gregation levels. Initially, when the concept was first introduced, the pattern was
tested by using the aggregate trade flows of each of the 11 OECD countries with
the rest of the world (Backus, Kehoe, and Kydland 1994). Bahmani-Oskooee,
Gelan and Ratha (2008), Bahmani-Oskooee, Kutan, and Ratha (2008), Parikh and
Shibata (2004), and Senhadji (1998) followed the same approach and tested the
curve for many more developing countries, including those in Africa. In countries
where the S-curve was not supported, Bahmani-Oskooee and Ratha (2007) argued
for aggregation bias and introduced the S-curve at the bilateral level between the
United States and its major trading partners. Finally, when the S-curve failed to
receive support at the bilateral level, Bahmani-Oskooee and Ratha (2008) argued
that an S-curve based on bilateral trade data suffers from another aggregation bias.
They then provided more support for the S-curve by disaggregating the trade flows
between the United States and the United Kingdom by commodities. While the
S-curve was not supported in the trade between the United States and the United
Kingdom, it was supported in 36 of the 52 industries that engaged in most of the
trade between the two countries.
As for Chinas experience with the S-curve, none of the studies reviewed above
included China, neither those that used aggregate trade flows nor those that used
bilateral trade flows. However, Bahmani-Oskooee and Ratha (2010) considered
bilateral trade flows between China and its major trading partner, the United States.
The S-curve was not supported using total bilateral trade flows between the two
countries, which implied that manipulating the RMB against the dollar will have
no effect on Chinese net exports to the United States. However, when the authors
disaggregated the trade flows and considered net exports of 100 industries engaged in
trade, the S-curve received support in 50 percent of the industries. More important,
the four largest industries, which conduct 50 percent of the trade, were shown to
conform to the S-curve and enjoyed increased net exports due to real depreciation
of the RMB today. These four industries were: office and data-processing machines,
office machines, miscellaneous manufactured articles, and baby carriages, toys,
games, and sporting supplies.
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 1

Export Shares of Chinas Trading Partners

Trading partner Exports ($billions) Share (%) Trading partner Exports ($billions) Share (%)

United States 324.86 17.878 Cuba 1.04 0.057

Hong Kong 267.52 14.722 Bulgaria 1.01 0.055
Japan 147.29 8.106 Oman 1 0.055
South Korea 82.92 4.564 Dominican Republic 0.97 0.053
Germany 76.43 4.206 Ethiopia 0.9 0.049
Netherlands 59.48 3.274 Azerbaijan 0.89 0.049
India 50.49 2.779 Costa Rica 0.88 0.049
United Kingdom 44.11 2.428 Bahrain 0.88 0.048
Singapore 35.3 1.943 Cameroon 0.87 0.048
Australia 33.91 1.866 Democratic Rep. of congo 0.83 0.045
Italy 33.71 1.855 Turkmenistan 0.79 0.043
Brazil 31.85 1.753 Georgia 0.76 0.042
France 30.24 1.664 Brunei Darussalam 0.74 0.041
Indonesia 29.26 1.61 Libya 0.72 0.04
Vietnam 29.09 1.601 Belarus 0.7 0.039
Malaysia 27.9 1.536 Mozambique 0.7 0.038
marchapril 2014 77

(continues)
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 1 (continued)

Trading partner Exports ($billions) Share (%) Trading partner Exports ($billions) Share (%)

United Arab Emirates 26.82 1.476 Senegal 0.68 0.037

Thailand 25.7 1.414 Antigua and Barbuda 0.66 0.036
Canada 25.25 1.39 Guinea 0.63 0.035
Mexico 23.98 1.32 Botswana 0.62 0.034
78 The Chinese Economy

Spain 19.72 1.085 Zambia 0.62 0.034

Belgium 18.97 1.044 Bahamas 0.55 0.03
Turkey 15.62 0.86 Cote dIvoire 0.54 0.03
Saudi Arabia 14.85 0.817 Djibouti 0.51 0.028
Iran 14.8 0.814 Madagascar 0.5 0.028
Panama 14.6 0.804 Mauritius 0.5 0.027
Philippines 14.26 0.785 Congo Republic 0.49 0.027
South Africa 13.37 0.736 Laos 0.47 0.026
Poland 10.94 0.602 Papua New Guinea 0.45 0.025
Chile 10.82 0.596 El Salvador 0.45 0.025
Kazakhstan 9.56 0.526 Nicaragua 0.42 0.023
Nigeria 9.2 0.506 Honduras 0.42 0.023
Argentina 8.5 0.468 Zimbabwe 0.41 0.022
Pakistan 8.44 0.464 Serbia 0.4 0.022
Bangladesh 7.81 0.43 Mauritania 0.39 0.021
 Downloaded by [Tulane University] at 21:13 01 April 2016

Czech Republic 7.67 0.422 Bolivia 0.38 0.021

Egypt 7.28 0.401 Jamaica 0.37 0.02
Ukraine 7.15 0.394 Uganda 0.36 0.02
Hungary 6.81 0.375 Bermuda 0.35 0.019
Israel 6.74 0.371 Haiti 0.3 0.017
Finland 6.63 0.365 Mali 0.3 0.016
Sweden 6.56 0.361 Gambia 0.29 0.016
Venezuela 6.52 0.359 Trinidad and Tobago 0.29 0.016
Denmark 6.45 0.355 Albania 0.28 0.015
Colombia 5.84 0.321 Namibia 0.28 0.015
Liberia 4.97 0.273 Gabon 0.27 0.015
Kyrgyz Republic 4.88 0.268 Equatorial Guinea 0.27 0.015
Myanmar 4.82 0.266 Afghanistan 0.23 0.013
Peru 4.66 0.256 Sierra Leone 0.23 0.012
Algeria 4.47 0.246 Fiji 0.17 0.009
Greece 3.95 0.217 Eritrea 0.15 0.008
Iraq 3.83 0.211 Barbados 0.14 0.008
Norway 3.79 0.209 Niger 0.14 0.008
New Zealand 3.74 0.206 Guyana 0.14 0.008
Switzerland 3.71 0.204 Suriname 0.14 0.007
Romania 3.45 0.19 Armenia 0.14 0.007
Ghana 3.11 0.171 New Caledonia 0.14 0.007

(continues)
marchapril 2014 79
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 1 (continued)

Trading partner Exports ($billions) Share (%) Trading partner Exports ($billions) Share (%)

Morocco 3.04 0.167 Vanuatu 0.13 0.007

Sri lanka 2.99 0.165 Netherlands Antilles 0.12 0.007
Benin 2.87 0.158 Malawi 0.11 0.006
Portugal 2.8 0.154 Moldova 0.1 0.005
80 The Chinese Economy

Angola 2.78 0.153 Maldives 0.1 0.005

Mongolia 2.73 0.15 Chad 0.09 0.005
Jordan 2.52 0.139 Macedonia 0.09 0.005
Slovak Republic 2.51 0.138 Somalia 0.09 0.005
Syria 2.43 0.134 Montenegro 0.09 0.005
Kenya 2.37 0.13 St. Vincent & Grenadines 0.08 0.004
Macao 2.35 0.129 Iceland 0.08 0.004
Malta 2.33 0.128 Lesotho 0.07 0.004
Cambodia 2.32 0.128 Timor-Leste 0.07 0.004
Austria 2.23 0.123 Rwanda 0.07 0.004
Ecuador 2.22 0.122 Burkina Faso 0.06 0.003
Ireland 2.15 0.118 Belize 0.05 0.003
Kuwait 2.13 0.117 Cape Verde 0.05 0.003
Uruguay 2 0.11 West Bank and Gaza 0.05 0.003
Tajikistan 2 0.11 French Polynesia 0.04 0.002
 Downloaded by [Tulane University] at 21:13 01 April 2016

Sudan 1.99 0.11 Bosnia & Herzegovina 0.04 0.002

Togo 1.83 0.101 Burundi 0.04 0.002
Slovenia 1.68 0.092 Samoa 0.04 0.002
Tanzania 1.66 0.091 Seychelles 0.04 0.002
Luxembourg 1.6 0.088 Swaziland 0.03 0.002
Croatia 1.54 0.085 Solomon Islands 0.03 0.002
Lebanon 1.46 0.08 Gibraltar 0.03 0.002
Uzbekistan 1.36 0.075 Dominica 0.03 0.001
Lithuania 1.34 0.073 Aruba 0.02 0.001
Guatemala 1.25 0.069 Bhutan 0.02 0.001
Paraguay 1.25 0.069 Guinea-Bissau 0.01 0.001
Qatar 1.2 0.066 Tuvalu 0.01 0.001
Latvia 1.19 0.066 Tonga 0.01 0.001
Nepal 1.18 0.065 Kiribati 0.01 0.001
Estonia 1.13 0.062 Central African Republic 0.01 0.001
Cyprus 1.12 0.062 St. Lucia 0.01 0.001
Tunisia 1.11 0.061 Comoros 0.01 0
Yemen 1.1 0.061 Grenada 0.01 0

Source: Direction of Trade Statistics by the International Monetary Fund.

marchapril 2014 81
 82 The Chinese Economy

In this article, we expand the literature by testing the S-curve between China and
another major partner, the United Kingdom. From Table 1 it is clear that the United
Kingdom ranks seventh among Chinas export markets. Which Chinese industries
would benefit from a real depreciation of the RMB against the British pound? Fol-
lowing Bahmani-Oskooee and Ratha (2010), we tested the S-curve not only using
aggregate bilateral trade flows between China and the United Kingdom, but also
for 47 industries that trade between the two countries. These are the industries for
which continuous time series data for their exports and imports from 1978 to 2011
were available. These industries engaged in 72 percent of the trade.

Method
Downloaded by [Tulane University] at 21:13 01 April 2016

The main approach to detecting the S-curve is based on the correlation function
between a measure of the trade balance and the real exchange rate. Since data at the
commodity level are reported by the United Kingdom, we define these two variables
from the British perspective. We define the net exports of an industry by TBi = (Xi
Mi)/GDP, where Xi represents exports of industry i from the United Kingdom to
China, and Mi represents imports of industry i by the United Kingdom from China.
The net export in nominal terms is deflated by the U.K.s nominal GDP.2 Next we
define the real bilateral exchange rate as REX = (PC.NEX/PUK), where NEX is the
nominal bilateral exchange rate, defined as number of British pounds per Chinese
renminbi; PC is the Chinese price level, and PUK is the price level in the United
Kingdom. Based on this definition of the real exchange rate, an increase in REX,
reflecting a real depreciation of the pound, is expected to increase the net exports
of a British industry. In order to learn about the paths of the nominal exchange rate
and the real rate, we plot their inverses in Figures 1 and 2.3 As can be seen from the
two figures, the renminbi has clearly depreciated in nominal as well as real terms.
More details about the data are provided in the Appendix.
As mentioned before, the S-curve posits that the cross-correlation coefficients
between the past values of the net exports and the current value of the real exchange
rate are negative. However, the same correlations between the future values of
the trade balance and the current real exchange rate are positive. Thus, following
Bahmani-Oskooee and Hegerty (2010) and other studies, we define the cross-
correlation coefficient between our two variables TB and REX by Equation (1).

___
( REX REX )(TB
t t+k TB )
k = (1)
___
( REX REX ) (TB
t
2
t+k TB )2

where the bar above each variable reflects the mean of that variable, t denotes time,
and k denotes the number of lags or leads.4 By allowing k to take negative values such
 marchapril 2014 83

Figure 1. Plot of Nominal Exchange Rate (pounds per renminbi)

0.35
Nominal Exchange Rate

0.3
Nominal Exchange Rate

0.25
0.2
0.15
0.1
0.05
Downloaded by [Tulane University] at 21:13 01 April 2016

Year
Source: Authors own calculations.

Figure 2. Plot of Real Exchange Rate (pounds per renminbi)

Real Exchange Rate

Source: Authors own calculations.

as 5, 4, 3, . . . and 0, we calculate the cross-correlation coefficients between past

values of the trade balance and the current real exchange rate. By allowing k to take
positive values such as +1, +2, . . . +5, we calculate cross-correlation coefficients
between future values of the trade balance and the current real exchange rate. Then,
by plotting the constructed values of k against negative as well as positive values
of k, we produce the S-curve. Note that in order to avoid spurious correlation, we
follow other studies and use the Hodrich-Prescott (HP) filter to estimate the trend
path of each series. Then deviation of each variable from its filtered trend is used
in Equation (1) to calculate the correlation coefficients.5
 84 The Chinese Economy

Figure 3. S-Curve at Bilateral and at Industry level

 
Downloaded by [Tulane University] at 21:13 01 April 2016

 

 

 

 

 




 marchapril 2014 85

Table 2

Industries Studied and Their Share of Trade

Code Product 2008 2009 2010 Support

112 Alcoholic beverages 0.0016 0.0017 0.0017

262 Wool and other animal hair 0.0006 0.0007 0.0006
276 Other crude minerals 0.0004 0.0004 0.0004
292 Crude vegetable materials,
n.e.s. 0.0004 0.0004 0.0003
512 Organic chemicals 0.0084 0.0087 0.0083
Downloaded by [Tulane University] at 21:13 01 April 2016

513 Inorganic chemicals 0.0034 0.0023 0.0028

514 Other inorganic chemicals 0.0014 0.0009 0.0009
531 Synthetic & organic dyestuffs 0.0010 0.0006 0.0008
533 Pigments, paints, varnishes 0.0014 0.0017 0.0015
541 Medical & pharmaceutical
products 0.0095 0.0127 0.0130
551 Essential oils, perfume, and
flavor 0.0005 0.0005 0.0006
554 Soaps, cleansing, & polishing
preparations 0.0011 0.0012 0.0011 Yes
581 Plastic materials, regenerated 0.0069 0.0079 0.0070
599 Chemical materials and prod-
ucts, n.e.s. 0.0050 0.0047 0.0047
613 Fur skins, tanned or dressed 0.0001 0.0001 0.0001
641 Paper and paperboard 0.0020 0.0029 0.0030 Yes
642 Articles of paper, pulp, paper-
board 0.0067 0.0068 0.0065
651 Textile yarn and thread 0.0010 0.0008 0.0010
653 Text fabrics woven excluding
narrow fabric 0.0025 0.0025 0.0030
655 Special textile fabrics and
related 0.0017 0.0017 0.0018 Yes
663 Mineral manufactures, n.e.s. 0.0023 0.0022 0.0021 Yes
665 Glassware 0.0033 0.0035 0.0035
666 Pottery 0.0043 0.0041 0.0041 Yes
689 Miscellaneous ferrous base
metals 0.0025 0.0015 0.0025
695 Hand tools or machine tools 0.0057 0.0047 0.0048
698 Manufactures of metal, n.e.s. 0.0172 0.0151 0.0145 Yes
711 Power-generating machinery 0.0204 0.0206 0.0177

(continues)
 86 The Chinese Economy

Table 2 (continued)

Code Product 2008 2009 2010 Support

714 Office machines 0.0641 0.0752 0.0873

715 Metalworking machinery 0.0031 0.0029 0.0027 Yes
718 Machines for special industries 0.0085 0.0068 0.0086 Yes
719 Machinery and appliances non
electrical 0.0390 0.0339 0.0375 Yes
722 Electric power machinery and
switch 0.0198 0.0184 0.0208
Downloaded by [Tulane University] at 21:13 01 April 2016

724 Telecommunications appa-

ratus 0.0644 0.0768 0.0687 Yes
726 Electrical apparatus for
medical purposes 0.0011 0.0016 0.0018
729 Other electrical machinery
and apparatus 0.0325 0.0277 0.0318
732 Road motor vehicles 0.0246 0.0231 0.0459
812 Sanitary, plumbing, heating,
& light 0.0141 0.0127 0.0133
841 Clothing except fur clothing 0.1410 0.1485 0.1264 Yes
861 Scientific, medical, optical,
means 0.0150 0.0176 0.0180
863 Developed cinematographic
film 0.0000 0.0000 0.0000
891 Musical instruments, sound
recorder 0.0343 0.0412 0.0346 Yes
892 Printed matter 0.0080 0.0094 0.0088
893 Articles of artificial plastic
make 0.0188 0.0194 0.0187
894 Perambulators, toys, games,
and sporting goods 0.0746 0.0686 0.0643
895 Office and stationery supplies,
n.e.s. 0.0038 0.0042 0.0030
897 Jewelry and gold-/silversmiths
watches 0.0068 0.0071 0.0067
899 Manufactured articles, n.e.s. 0.0116 0.0117 0.0105
Source: Authors calculation.
Note: n.e.s.= not elsewhere specified.
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 3

CrossCorrelation Coefficients at Various Lags and Leads of Net Exports

Code Industry 6 5 4 3 2 1 0 1 2 3 4 5 6

112 Alcoholic beverages 0.08 0.38 0.39 0.23 0.04 0.24 0.30 0.29 0.18 0.00 0.12 0.10 0.00
262 Wool and other animal
hair 0.08 0.01 0.02 0.29 0.25 0.06 0.43 0.70 0.06 0.44 0.37 0.03 0.17
276 Other crude minerals 0.10 0.13 0.06 0.21 0.39 0.22 0.28 0.39 0.02 0.23 0.06 0.22 0.07
292 Crude vegetable materi-
als, n.e.s. 0.17 0.14 0.12 0.40 0.71 0.42 0.33 0.46 0.04 0.31 0.07 0.29 0.42
512 Organic chemicals 0.13 0.12 0.36 0.57 0.41 0.04 0.17 0.03 0.28 0.08 0.34 0.58 0.42
513 Inorganic chemicals
elements, oxides, 0.20 0.06 0.08 0.11 0.07 0.04 0.06 0.09 0.18 0.09 0.08 0.16 0.11
514 Other inorganic
chemicals 0.34 0.28 0.14 0.34 0.26 0.01 0.19 0.30 0.11 0.15 0.24 0.18 0.03
531 Synthetic organic dye-
stuffs, natural 0.08 0.02 0.15 0.23 0.22 0.06 0.15 0.11 0.27 0.13 0.11 0.15 0.03
533 Pigments, paints,
varnishes 0.23 0.21 0.36 0.36 0.21 0.20 0.44 0.28 0.27 0.04 0.34 0.54 0.15
541 Medicinal & pharmaceu-
tical products 0.28 0.32 0.36 0.28 0.06 0.21 0.36 0.26 0.08 0.04 0.13 0.21 0.20
551 Essential oils, perfume,
and flavor 0.31 0.21 0.14 0.07 0.00 0.14 0.32 0.26 0.04 0.03 0.04 0.07 0.04
554 Soaps, cleansing, &
polishing preparations 0.02 0.12 0.46 0.65 0.20 0.02 0.22 0.30 0.29 0.06 0.05 0.14 0.32
marchapril 2014 87

(continues)
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 3 (continued)

Code Industry 6 5 4 3 2 1 0 1 2 3 4 5 6

581 Plastic materials,

regenerated 0.47 0.34 0.17 0.08 0.01 0.23 0.05 0.09 0.05 0.20 0.21 0.11 0.20
599 Chemical materials and
products, n.e.s. 0.45 0.07 0.30 0.42 0.28 0.08 0.10 0.08 0.14 0.03 0.12 0.05 0.11
613 Fur skins, tanned or
88 The Chinese Economy

dressed 0.00 0.31 0.13 0.32 0.35 0.12 0.18 0.07 0.04 0.23 0.24 0.01 0.05
641 Paper and paperboard 0.26 0.21 0.08 0.18 0.28 0.28 0.18 0.02 0.24 0.20 0.23 0.09 0.04
642 Articles of paper, pulp,
paperboard 0.02 0.21 0.35 0.29 0.08 0.12 0.17 0.14 0.00 0.13 0.06 0.06 0.06
651 Textile yarn and thread 0.06 0.20 0.30 0.33 0.06 0.03 0.21 0.19 0.17 0.01 0.06 0.06 0.05
653 Textile fabrics woven
excluding narrow, spec, 0.39 0.12 0.28 0.49 0.36 0.15 0.15 0.27 0.23 0.04 0.10 0.17 0.07
655 Special textile fabrics
and related 0.17 0.04 0.12 0.14 0.46 0.45 0.21 0.05 0.37 0.49 0.28 0.16 0.13
663 Mineral manufactures,
n.e.s. 0.04 0.21 0.34 0.31 0.23 0.05 0.27 0.35 0.13 0.09 0.06 0.06 0.13
665 Glassware 0.12 0.20 0.31 0.26 0.09 0.11 0.22 0.11 0.09 0.23 0.20 0.07 0.04
666 Pottery 0.24 0.07 0.28 0.29 0.22 0.15 0.02 0.21 0.26 0.17 0.02 0.01 0.17
689 Miscellaneous nonfer-
rous base metals 0.18 0.01 0.17 0.34 0.40 0.04 0.06 0.06 0.06 0.04 0.05 0.02 0.03
695 Hand tools or machine
tools 0.16 0.21 0.23 0.16 0.12 0.24 0.25 0.14 0.07 0.20 0.36 0.26 0.02
698 Manufactures of metal,
n.e.s. 0.07 0.07 0.13 0.08 0.06 0.01 0.12 0.23 0.23 0.19 0.06 0.14 0.20
 Downloaded by [Tulane University] at 21:13 01 April 2016

711 Power-generating ma-

chinery 0.18 0.40 0.16 0.08 0.07 0.09 0.07 0.07 0.43 0.26 0.01 0.02 0.23
714 Office machines 0.06 0.22 0.33 0.37 0.30 0.13 0.17 0.21 0.14 0.01 0.04 0.03 0.06
715 Metalworking machinery 0.16 0.10 0.14 0.02 0.34 0.39 0.14 0.02 0.32 0.21 0.02 0.29 0.31
718 Machines for special
industries 0.07 0.22 0.20 0.14 0.52 0.47 0.09 0.59 0.51 0.10 0.24 0.30 0.21
719 Machinery and appli-
ances nonelectronic 0.04 0.32 0.42 0.10 0.31 0.51 0.11 0.52 0.48 0.12 0.15 0.39 0.45
722 Electric power machin-
ery and switches 0.25 0.34 0.31 0.22 0.05 0.18 0.20 0.02 0.13 0.16 0.12 0.01 0.14
724 Telecommunications
apparatus 0.19 0.17 0.12 0.08 0.04 0.03 0.04 0.08 0.35 0.32 0.22 0.01 0.09
726 Electrical apparatus for
medical purposes 0.20 0.07 0.37 0.28 0.16 0.05 0.17 0.21 0.12 0.18 0.29 0.04 0.12
729 Other electrical machin-
ery and apparatus 0.17 0.08 0.36 0.23 0.10 0.23 0.11 0.11 0.12 0.01 0.14 0.14 0.00
732 Road motor vehicles 0.30 0.24 0.12 0.08 0.08 0.01 0.26 0.31 0.15 0.12 0.15 0.20 0.23
812 Sanitary, plumbing, heat-
ing, & light 0.04 0.17 0.34 0.33 0.22 0.16 0.09 0.02 0.08 0.08 0.15 0.22 0.16
841 Clothing except fur
clothing 0.09 0.08 0.12 0.17 0.06 0.19 0.23 0.02 0.19 0.16 0.07 0.11 0.19
861 Scientific, medical, opti-
cal, meas. 0.14 0.11 0.19 0.27 0.29 0.18 0.03 0.09 0.11 0.18 0.18 0.08 0.02
863 Developed cinemato-
graphic film 0.15 0.19 0.03 0.23 0.18 0.15 0.22 0.45 0.01 0.25 0.26 0.39 0.18
marchapril 2014 89

(continues)
 Downloaded by [Tulane University] at 21:13 01 April 2016

Table 3 (continued)
Code Industry 6 5 4 3 2 1 0 1 2 3 4 5 6

891 Musical instruments,

sound recorder 0.40 0.25 0.08 0.20 0.31 0.08 0.05 0.04 0.20 0.17 0.14 0.11 0.07
892 Printed matter 0.15 0.01 0.18 0.27 0.29 0.17 0.08 0.09 0.02 0.06 0.15 0.17 0.12
893 Articles of artificial plas-
tic materials 0.07 0.20 0.34 0.31 0.19 0.01 0.18 0.18 0.12 0.05 0.09 0.13 0.01
90 The Chinese Economy

894 Perambulators, toys,

games, and sporting
goods 0.04 0.18 0.32 0.23 0.07 0.02 0.02 0.07 0.03 0.00 0.09 0.12 0.09
895 Office and stationery
supplies, n.e.s. 0.05 0.18 0.24 0.26 0.16 0.11 0.09 0.04 0.13 0.06 0.02 0.10 0.06
897 Jewelry and gold-/silver-
smiths watches 0.16 0.03 0.16 0.34 0.34 0.20 0.05 0.02 0.05 0.01 0.07 0.13 0.04
899 Manufactured articles,
n.e.s. 0.04 0.23 0.35 0.36 0.19 0.02 0.27 0.25 0.06 0.06 0.00 0.11 0.07

Source: Authors calculation.

Note: n.e.s. = not elsewhere specified.
 marchapril 2014 91

Results

The data used to generate the S-curve pattern were drawn annually from 1978 to
2010. We generate the S-curve using aggregate bilateral trade data between the
United Kingdom and China. We then disaggregate the aggregate bilateral trade data
and generate the S-curve for 47 industries that trade between the two countries.
As mentioned before, these industries for which continuous time series data were
available engage in 72 percent of the trade. We summarize our results in Table 2,
but report the S-curve only for industries that conform to the curve in Figure 2.
While the S-curve was not supported when aggregate bilateral trade data were
used (see Figure 3), it was supported in at least 12 of the 47 industries, as identified
Downloaded by [Tulane University] at 21:13 01 April 2016

in Table 2. Included among these 12 industries is the largest industry, coded 841
(clothing except fur clothing), which has more than 12 percent of market share.
While graphic presentation of S-curves helps us to identify the industries, it is not
easy to read the magnitudes of the correlation coefficients from the graphs. To that
end, we report the size of the coefficients in Table 3. Note that in six of the 12 in-
dustries that conform to the S-curve hypothesis (identified by numbers in bold type)
there is a reversal of the signs after the fifth lead. This could be due to the fact that
we have calculated cross-correlation coefficients over longer lags and leads since
the data became annual. If we stop at four leads, we see the sign reversal only in
one industry. Furthermore, the sizes of the coefficients are much smaller at longer
leads, implying that the effects of depreciation have been weakened.

Conclusion

The S-curve phenomenon, a concept developed by Backus et al. (1994), asserts

that while the past values of a trade balance and the current terms of trade or the
real exchange rate are negatively correlated, the future values of the trade balance
and the current real exchange rate are positively related. This implies that currency
depreciation or devaluation will improve the trade balance in the future. Research
reported in the literature has shown that the emergence of the S-curve depends on
the level of data aggregation. While aggregate trade flows between one country and
the rest of the world fail to support the S-curve, bilateral trade data between two
countries do support it. And where aggregate bilateral trade flows fail, disaggregating
the same data and using industry-level data between two countries does not.
This article considers the trade flows between China and one of its major trading
partners, the United Kingdom. When we use bilateral aggregate trade data between the
two countries for the period from 1978 to 2010, we are unable to support the S-curve
pattern. However, when we disaggregate the trade data by industry, we find support
for the S-curve in 12 of the 47 industries. While most of the 12 industries are small,
one large industry that engages in 12 percent of the trade between the two countries
conforms to the S-curve, implying that it will enjoy the positive association between
its trade balance and currency depreciation in the future.
 92 The Chinese Economy

Notes

1. For additional issues involved in trade development in China, see He (2010), Li

(2009), Montinari and Prodi (2011), and Xiaoding (2007).
2. This definition of the real trade balance is espoused by Backus et al. (1994) and
others.
3. For additional details on the movement of the real exchange rate in China, see Zhang
(2009).
4. For example, when k = 3, we are constructing the cross-correlation coefficient
between the current exchange rate and the trade balance of three years ago, since data are
annual.
5. The S-curve concept is similar to the J-curve concept. For more on the J-curve, see
Bahmani-Oskooee (1985).
Downloaded by [Tulane University] at 21:13 01 April 2016

References

Backus, D.K.; P.J. Kehoe; and F.E. Kydland. 1994, Dynamics of the Trade Balance and
the Terms of Trade: The J-Curve? American Economic Review 84: 84103.
Bahmani-Oskooee, M. 1985. Devaluation and the J-Curve: Some Evidence from LDCs.
Review of Economics and Statistics 67: 500504.
Bahmani-Oskooee, M., and S.W. Hegerty. 2010. The J- and S-Curves: A Survey of the
Recent Literature. Journal of Economic Studies 37: 58096.
Bahmani-Oskooee, M., and A. Ratha. 2007. The S-Curve Dynamics of US Bilateral
Trade. Review of International Economics 15: 43039.
. 2008. S-Curve at Industry Level: Evidence from US-UK Bilateral Trade. Empiri-
cal Economics 35: 14152.
. 2010. S-Curve Dynamics of Trade Between U.S. and China. China Economic
Review 21: 21223.
Bahmani-Oskooee, M.; A. Gelan; and A. Ratha. 2008. S-Curve Dynamics of Trade in
Africa. African Development Review 20: 33542.
Bahmani-Oskooee, M.; A. Kutan; and A. Ratha. 2008. The S-Curve in Emerging Markets.
Comparative Economic Studies 50, no. 2: 34151.
He, Sh.-Q. 2010. Dynamics of Chinese Agricultural Trade Patterns: Pre- and Post WTO
Periods. Chinese Economy 43, no. 2: 525.
Li, K.-W. 2009. Finance and Trade Development in Contemporary China. The Chinese
Economy 42, no. 2: 36.
Montinari, L., and G. Prodi. 2011. Chinas Impact on Intra-African Trade. The Chinese
Economy 44, no. 4: 7591.
Parikh, A., and M. Shibata. 2004. Dynamics of the Relationship Between the Terms
of Trade and the Trade Balance in Developing Countries of Asia, Africa, and Latin
America. Journal of Quantitative Economics 2: 10421.
Senhadji, A.S. 1998. Dynamics of the Trade Balance and the Terms of Trade in LDCs:
the S-Curve. Journal of International Economics 46: 10531.
Xiaoding, R. 2007. The Characteristics and Trend of Chinas Processing Trade Develop-
ment in Recent Years. The Chinese Economy 40, no. 4: 7078.
Zhang, Y. 2009. Foreign Currency Inflow and Real Exchange Rate Movement. Chinese
Economy 42, no. 2: 6390.
 marchapril 2014 93

Appendix

Data Definition and Sources

All data are annual from 1978 to 2010 and were collected from the following
sources:
a. International financial statistics from International Monetary Fund (CD-
ROM).
b. World Bank, World Integrated Trade Solution System.
c. World Bank, World Development Indicators.
Downloaded by [Tulane University] at 21:13 01 April 2016

Variables in Equation (1)

TB = Trade balance of an industry. Defined as U.K. exports of industry i to Chi-

na minus U.K. imports of same industry from China divided by U.K.s GDP.
Since industry export and import data come in terms of the U.S. dollar, U.K.
GDP had to be in dollars also. Industry-level data come from source b; U.K.
GDP in dollars comes from source c.
REX = Bilateral real exchange rate between the British pound and Chinese
renminbi. Defined as PUK, where PC is price level in China, PUK is price level
in the United Kingdom, and NEX is nominal rate defined as number of
pounds per renminbi. An increase in REX reflects a real depreciation of the
pound. Price levels measured by the GDP deflator come from source c. The
nominal exchange rate was not directly available and was calculated using
cross-exchange rates against the dollar. These rates come from source a.