Estimate Hong Kong Re-export Markups and Reconcile Trade Statistics from China, Hong Kong and Their Major Trading Partners -- A Mathematical Programming Approach

1
Estimate Hong Kong Re-export Markups and
Reconcile Trade Statistics from China, Hong Kong
and Their Major Trading Partners -- A
Mathematical Programming Approach

• Zhi Wang
• United States International Trade Commission
• Mark Gehlhar
• United States Department of Agriculture
• Shunli Yao
• China Center for Economic Research, Peking
  University
• The views expressed in this presentation are
  solely those of the presenter. It is not meant to
  represent in anyway the views of the U.S.
  International Trade Commission or any of its
  individual Commissioners.
• .

2
Presentation Outline

• Issues and Objectives
• Literature Review
• Model Specification
• Steps to Implement the Model
• Preliminary Results
• Conclusions and Future Work

3
China and U.S. Merchandise Trade Balance,
1985-2005
4
Literature Review International Trade Statistics

• China-U.S. Joint Commission on Commerce and Trade
  (JCCT), 1995 Statistics Canada, 2002, 2005
• Fung and Lau, 1996, 1998, 2001, 2003, 2004
• Feenstra et al., 1999, 2004
• Schindler and Beckett, 2005
• The literature to date consistently shows that
  the re-export activities in Hong Kong are one of
  the major contributing factors to the statistical
  discrepancy.

5
Major Issues

• There are significant discrepancies between trade
  statistics reported by China and its partners,
  especially at commodity level, but adjustment in
  the literature to date is only made at the
  aggregate level
• There is no well developed consistent
  relationships between trade statistics reported
  by trading partners in a world in which
  transshipment and re-export activities have
  become increasingly important
• Re-export markup estimates are ad hoc and varies
  considerably, as a result, substantial
  inconsistencies still exist after adjusting trade
  via Hong Kong even at the aggregate level in
  recent years

6
Literature ReviewComputational Economics

• Constrained matrix-balancing as a procedure to
  solve under or over determined estimation
  problems. It is a core mathematical structure in
  diverse applications
• Estimating input-output tables and inter-regional
  trade flows in regional science
• Balancing of social/national accounts in
  economics
• Estimating interregional migration in demography
  
• Analysis of voting patterns in political science
• Treatment of census data and estimation of
  contingency tables in statistics
• Estimation of transition probabilities in
  stochastic modeling
• Projection of traffic within telecommunication
  and transportation networks
• Reconcile international trade statistics reported
  by partner countries

7
Objectives

• Develop and implement a formal model to estimate
  Hong Kong re-export markup and reconcile trade
  statistics from China, Hong Kong and their
  partners simultaneously in a consistent
  optimization framework.
• To apply the model to 2004 bilateral world trade
  data in GTAP sector classification to produce
  Hong Kong re-exports adjusted trade flows
  contributing to version 7 GTAP database.

8
Basic Ideas of the Adjustment Framework

• To apply this procedure as a data reconciliation
  tool (solve over determined estimation problem),
  three pieces of information are needed
• Initial estimates of the same economic variables
  from different sources (in national economic
  accounts, estimates of the same variables can
  often be obtained from income, expenditure or
  production data in international trade, export
  and import reported between partner countries).
• An accounting framework and other constraints
  (demands have to equal supplies, components have
  to sum to totals, global balance of trade equal
  to zero, ).
• Reliability information on the initial estimates
  (standard error, ranking index, )

9
Problems of Proportional Adjustment in
International Trade Statistics
Country Name China direct exports to Partners Hong Kong domestic export to partner China re-exports to partner via Hong Kong Partners total imports from China Partner imports with Hong Kong origin Statistical discrepancies
Mozambique 75 1 4 18 1 -280.5
Benin 577 1 4 48 3 -1019.8
Togo 399 4 29 56 14 -394.6
Kyrgyz Republic 492 0 0 80 0 -548.4
Malta 273 5 20 78 15 -196.7
Luxembourg 918 13 35 76 58 -447.1
Venezuela 596 15 73 185 34 -173.7
Sudan 730 2 2 263 6 -186.4
Singapore 12,684 2,095 3,456 16,012 3,599 9.1
United Kingdom 14,952 3,587 6,849 26,206 3,372 15.7
Korea Rep 27,810 2,117 2,832 29,585 3,268 -1.5
Germany 23,753 2,752 6,905 40,679 2,491 21.4
Japan 73,222 4,302 11,977 93,589 1,322 4.2
United States 125,118 17,707 35,587 208,153 9,141 17.2
10
Key Assumptions

• Hong Kong is the only entrepot between China and
  its partner countries
• All reporting countries, including China, can
  correctly identify the country of origin of their
  imports, whatever the imports are directly from
  the partners or indirectly from Hong Kong, but
  can not determine the final destination when
  exports leave their ports.

11
Mirror Relation among U.S. , China and Hong Kong
Reported Trade StatisticsEastbound Flows

China reported total exports to U.S. (fob)
12
Mirror Relation among U.S. , China and Hong Kong
Reported Trade StatisticsWestbound Flows

U.S. reported total exports to China (fas)
U.S. Reported Total exports to Hong Kong (fas)
Hong Kong reported re-exports of U.S originated
goods
Hong Kongs reported re-export U.S. originated
goods to other countries
U.S. indirect exports to China via Hong Kong
-
Hong Kong reported imports of U.S originated
products (cif)
China reported imports of U.S. originated
prodicts (cif)
fob/cif adj
-
-

fob/cif adj
13
Notation and Conventions

• DX Direct exports between custom regions
• RX Indirect exports via Hong Kong between custom
  regions
• TX Total exports between custom regions
• TM Total imports between custom regions
• DM Direct imports between custom regions
• RXM Re-export markups
• cif CIF/FOB margins rate
• wex a countrys total exports
• wmxa countrys total imports

• Superscripts denote custom regions, subscripts
  denote commodities and years
• Capital letter denote variables, lower latter
  denote parameters
• Variables with a 0 in the end denote initial
  estimates for that variable
• An additional w before the variable indicates
  the reliability measure for that variable
• Exports are measured on a fob basis, imports are
  measured on cif basis.

14
Mirror Relation among U.S. , China and Hong Kong
Reported Trade StatisticsEastbound Flows
China reported direct exports to U.S. (fob)
fob/cif adj


15
Consistency Constraints in the Model

• There are 16 consistency constraints in total
• Five for Eastbound trade, five for Westbound
  trade, four for China-Hong Kong bilateral trade,
  and two for global exports and imports
• The first constraint in both East and Westbound
  trade is intended to isolate the statistical
  discrepancies with re-export markups, and defines
  two side of the mirror-trade statistics in each
  directions in the presence of entrepot trade
• The rest of the constraints define trade
  variables in the first constraint and give the
  mathematical relations between observed and
  unobserved trade statistics

16
Basic Consistency ConstraintsEastbound flows
China and Hong Kong export to partners

• Partner reported imports from China and Hong Kong
  after fob/cif adjustment should equal the sum of
  China reported exports to the partner, Kong Kong
  reported domestic exports, Hong Kong re-exports
  for China originated goods to the partner, plus a
  statistical discrepancy
• Chinas total exports to partner equals Chinas
  direct exports plus Hong Kongs re-exports for
  China minus Hong Kongs re-export makeup
• Hong Kongs domestic exports to a partner equal
  its total exports to that partner minus its
  re-exports for all other countries to the partner
  plus its markup earnings from re-exports

17
Basic Consistency Constraints Eastbound flows
China and Hong Kong export to partners
18
Mirror Relation among U.S. , China and Hong Kong
Reported Trade StatisticsWestbound Flows
U.S. reported direct exports to China (fas)
U.S. Reported Total exports to Hong Kong (fas)
Hong Kong Reported Total re-exports of U.S
originated goods

-
Hong Kongs reported re-export U.S. originated
goods to other countries
U.S. indirect exports to China via Hong Kong
fob/cif adj
-


Hong Kong retained imports of U.S originated
products (cif)
China reported imports of U.S. originated
products (cif)
fob/cif adj
-
-
fob/cif adj
19
Basic Consistency ConstraintsWestbound flows
partners export to China and Hong Kong

• Sum of U.S. reported exports to China and Hong
  Kong equals the sum of China and Hong Kong
  reported imports originated from U.S. after
  fob/cif adjustment, minus Kong Kong re-exports
  for the U.S. to China
• Hong Kongs domestic use of imports plus its
  re-exports for a partner minus re-exports markup
  equals Hong Kongs total imports from that
  partner
• Chinas direct imports from a partner equal
  Chinas total imports from that partner minus
  Hong Kongs re-exports to China for that partner
  adjusted by Hong Kongs re-export markups

20
Basic Consistency Constraints Westbound flows
partners export to China and Hong Kong
21
Basic Consistency ConstraintsGlobal balance

• Actual exports from China and Hong Kong to all
  their partners after adjustment should still
  equal to the sum of their reported total exports
  to the world
• China and Hong Kongs imports and Hong Kongs
  re-exports minus re-exports markup after
  adjustment should still equal the sum of China
  and Hong Kongs total imports from the world

22
The Adjustment Problems
Adjust a given set of initial trade statistics
according to the following objective function to
satisfy the 16 consistency constraints
23
Theoretical Properties

• Statistical interpretations underlying the model
  differ when different reliability weights are
  used
• Estimation of Hong Kongs re-export markups,
  rearrange sources and destinations of Chinas and
  Hong Kongs exports and imports, adjust bilateral
  trade balance for China and all its partners are
  made in a consistent simultaneous manner, thus
  impose global consistence to the adjusted trade
  data
• In all but the trivial case, posterior estimates
  derived from entropy or quadratic loss minimand
  will always be closer to the unknown, true values
  than the associated initial statistics
• The choice of weights in the objective function
  has a large impact on the estimation results

24
Why Adjusted Estimates Better?

• D0 Initial estimates
• W variance matrix of initial estimates ,
• A coefficient matrix of all linear constraints
  AD 0
• The BLUE
• D will never be worse than D0 with equal or
  smaller variance

25
Empirical Advantages

• Convenience and details
• Hong Kong's re-export markup rate, each country's
  re-exports via Hong Kong as percent of the
  country's total exports, and adjusted bilateral
  BOT between China, Hong Kong and their partners
  all are part of the model solution
• Completeness
• Complete use of all information from official
  trade statistics
• Flexibility
• The model permits a wider variety and volume of
  information to be brought into the adjustment
  process than is possible with scaling methods
  such as RAS
• Incorporation of data reliabilities in a systemic
  way
• The weights in the objective function reflect the
  relative reliability of a given trade flow.
  Trade statistics with higher reliability should
  undergo less adjustment than trade statistics
  with lower reliability

26
Five Key Steps to Implement the Model

• Obtain initial estimates
• Calculate initial Hong Kong re-export markups
• Estimates of fob/cif margins
• Country and commodity aggregation
• The choice and estimation of reliability weights

27
Observed and Derived Trade Statistics East bound
trade

• Observed
• Chinas direct exports to partner countries
• Hong Kongs total exports to partner countries
• Partners total imports from China
• Partners imports of product originated from Hong
  Kong
• Derived
• Chinas total exports to partner countries
• Hong Kongs domestic exports to partner countries
• Partner countries direct imports from China
• Partner countries total imports from Hong Kong

28
Observed and Derived Trade Statistics West bound
trade

• Observed
• Partner countries total exports to Hong Kong
• Partner countries direct exports to China
• Chinas total imports from partner countries
• Hong Kongs total imports from partner countries
• Derived
• Hong Kongs imports from partner countries for
  domestic use
• Chinas direct imports from partner countries
• Partner countries total exports to China
• Partner countries exports for Hong Kongs
  domestic market

29
Data Source

• All data are from 1995-2004
• China Customs General Administration, 8-digit
  Harmonized System (HS)
• Hong Kong Census and Statistical Department, 8-
  digit HS
• United States Census Bureau, 10-digit HS
• Other Partner Countries World Integrated Trade
  Solution (WITS) managed by the World Bank, 6-
  digit HS, more than 150 reporting countries.

30
Reliability of Reported Trade Statistics

• Mirror trade statistics are the major data source
  to estimate the reliability weights
• Econometric analysis of discrepancies between the
  two "reported" trade data of the same trade flows
  provide estimates of data reliability

31
Auto regression with dummy variables

e is the mirror trade statistics discrepancies b
is the symmetric bias dummy variables represent
events have a significant impact on the reporting
practice in the two data reporting countries
(change of commodity classifications,
implementing better custom information systems,
enforcing effective anti-smuggling programs,
ect.) the variance
32
Country and Commodity Aggregation

• 42 GTAP merchandise sectors aggregated from
  original data at 6 and 8 digit HS level
• 215 countries identified in the GTAP global
  bilateral trade data base, while only 157
  countries reported at least one year of their
  exports to or import from China and Hong Kong
  during 2002 to 2004
• First aggregate all the non-reporting country
  into one block
• Then use two cut off criteria to separate the 157
  reporting country into two blocks. The first
  block has 96 countries, including all single
  countries in version 6 GTAP database and the sum
  of exports from China and Hong Kong to the world
  greater than 300 million dollars in 2004
  identified either by China and Hong Kong reported
  data or their partner reported data.
• The second block is consisted of 61 remaining
  reporting countries, which aggregated to one
  other reporting country block.
• Include China and Hong Kong, the model has 98
  countries.

33
Characters of the Initial Data

• Reported westbound trade is less problematic than
  reported eastbound trade. (20 of the 97 reported
  bilateral routines with more than 100
  discrepancies in eastbound trade, only two
  routines in the westbound trade see such large
  discrepancies)
• Trade with developing country partners shows
  greater discrepancies than developed countries in
  general
• Small trade flows often associate with large
  discrepancies than large flows. Extremely large
  discrepancies come from partners only have small
  trade values

34
Initial and Adjusted Re-export Markup Rate, from
Partner to China, in
35
Initial and Adjusted Re-export Markup Rate, from
Partner to China, in
36
Initial and Adjusted Re-export Markup Rate, from
China to Partner, in
37
Initial and Adjusted Re-export Markup Rate, from
China to Partner, in
38
Initial and Adjusted Re-export Earnings to
China, Million of U.S. Dollars
39
Initial and Adjusted Re-export Earnings from
China, Million of U.S. Dollars
40
Initial and Adjusted Retaining Imports in Hong
Kong, Millions of U.S. Dollars
41
Reported and Adjusted Bilateral TradeBalance
between China and Selected PartnersMillions of
U.S. Dollars
42
Reported and Adjusted Bilateral TradeBalance
between China and Selected PartnersMillions of
U.S. Dollars
43
Reported and Adjusted Chinas Net Exports to All
Partners Except Hong Kongby Major Commodities,
Millions of U.S. Dollars
44
Reported and Adjusted Chinas Net Exports to All
Partners Except Hong Kongby Major Commodities,
Millions of U.S. Dollars
45
Reported and Adjusted China-U.S. Bilateral
TradeBalance by Major Commodities Millions of
U.S. Dollars
46
Reported and Adjusted China-U.S. Bilateral
TradeBalance by Major Commodities Millions of
U.S. Dollars
47
Conclusions

• Preliminary result shows that the model developed
  in this study provides a flexible tool to
  estimate Hong Kong re-export markup and reconcile
  trade statistics from China, Hong Kong and their
  trading partners simultaneously to fully use all
  available official trade statistics.

48
Implications

• The model can be applied to reconcile direct and
  indirect trade for other regions of the world
  where transshipment creates major discrepancies.
• It not only provides a tool for the preparation
  of global trade data in future versions of GTAP
  database, but also contributes to the
  methodological development to estimate and
  reconcile discrepancies in international trade
  statistics when transshipment and re-export
  activities heavily diminish the ability of a
  country identifying its correct partner
  countries.

49
Future Work

• Missed and double counted transshipments through
  Hong Kong
• Reliability of Hong Kongs re-export statistics
• Countries total exports, imports and supply-
  demand balance at global level
• Econometrically estimate reliability weights and
  fob/cif margins

50
Thank You

• My E-mail Address
• Zhi.Wang_at_USITC.GOV