Patterns of trade

1
Patterns of trade

• How to work with trade data
• Frank van Tongeren
• Based on presentations by
• Henk Kelholt and Thom Achterbosch

2
What is a trade database?

• Products
• Reporting countries
• Partner countries
• Periods
• Streams
• Value/quantity/

• Examples of trade databases
• ITC/WTO PCTAS www.intracen.org/pctas/pctas00.htm
  
• EUROSTAT COMEXT www.dsidata.com/products/comext.
  htm
• UN COMTRADE

3
Products

• HS
• Harmonized System 6 digits 8000
  products
• 01 - Live animals
• 0101 - Live horses, donkeys, mules
• 010111 - Pure bred live horses
• SITC
• Standard International Trade Classification 5
  digits 5150 products
• 0 - Food and live animals
• 00 - Live animals
• 001 - Live animals
• 0011 - Bovine animals, live
• 00111 - Bovine animals, for breeding

4
Countries

• Reporters
• countries who register their imports and exports,
  the commodity and the country of origin or
  country of destination (depending on the database
  between 10 and 150)
• Partners
• countries of origin or destination (usually
  200)
• Number of reporting countries
• The higher the number of reporters, the larger
  the share in world trade that is covered by the
  database

5
Periods Streams Value Quantity

• Periods - usually years or months
• Streams - imports or exports
• Value - in thousands or millions of US or
• Quantity - tons and/or special units, e.g.
  number, pairs, cubic meter

6
Example

• The value of the export of wheat from Canada to
  Iran in the year 2000

• Version 1
• Product Wheat
• Reporting country 124 Canada
• Partner country 364 Iran
• Period 2000
• Stream export
• Value dollar x 1000

• Version 2
• Product Wheat
• Reporting country 364 Iran
• Partner country 124 Canada
• Period 2000
• Stream import
• Value dollar x 1000

7
Assignments

• Using the trade data provided in IRAN_trade.XLS
• Which are the main (broad) import commodities,
  which are the main export commodities?
• What is the development over time of the trade
  balance (X-M) for Food products?
• Which patterns can we see?
• Which are the main trade partners for food
  exports and imports?

8
Why TSA-express?

• Time Series Analysis Express
• Different data sources
• Speed, performance, ease of use
• Research gt exploration of data
• Meta data/information
• TSA by
• Dr. Wietse Dol - econometrics senior software
  consultant/developer
• Henk Kelholt - trade data specialist
• full TSA manual www.lei.dlo.nl/tsa
• wietse.dol_at_wur.nl
• henk.kelholt_at_wur.nl

9
Building trees

• Products/countries/periods have relationships,
  e.g.

10
Relations, relevance and functions

• All items in a tree have
• a period of relevance
• a function that specifies how the item is
  calculated

• 058 - Belgium/Luxembourg
• Period of relevance 1-1-1993 until 31-12-1998
• Function available in database
• 058a - Belgium Luxembourg
• Period of relevance 1-1-1993 until 31-12-9999
• Function add children when they are valid
  (period of relevance)

11
The program

• With the tree elements you select, TSA-express
  knows the minimal dataset that has to be
  collected to present the query (i.e. walk through
  the tree until you only have items with the
  function available in database). Load the data
  of these elements to memory and perform the
  function calculations in the memory (using sparse
  matrices and pointer techniques)
• All functions are calculated at the moment the
  viewer needs them
• Speed is more important than disk size or memory
  usage, i.e. trees use a lot of memory, but show
  what you are doing
• Using special software for speed and performance
  (tree, grid, graphics). Using special memory
  techniques for sparse matrices. Programmed
  modular in Delphi using objects.
• Oracle or Access XP database (Access is faster)

12
User friendly

• Using trees gt making a query is very
  visual/easy
• Using a username/password all settings are saved
  for next time usage
• You can define your own groups of
  products/countries etc.
• You can save selections/queries, i.e. makes
  periodical publications easy
• Multiple languages in database (trees and
  reports) as well as in the Graphical User
  Interface
• The multidimensional data viewer makes the report
  you want
• Save report to many formats (Word, Excel, CSV,
  HTML) or just print it, make a graphic,
  portfolio plot, trade balance
• Speed is important
• We have developed data loading routines (makes it
  easy to add a new data for COMEXT and PCTAS)

13
TSA table
14
TSA graphics
15
TSA portfolio analysis
Indonesian exports to World - 5.5 USA imports
from Indonesia 0.24
16
TSA trade balance
17
Things to keep in mind 1.

• Do not focus on a single product code. Look also
  at other products from the same group (or level
  of aggregation) and look at one level of
  aggregation higher.
• If possible study the exports of country A to
  country B and the imports of country B from
  country A.
• Quantities at a high level of aggregation may be
  the sum of very different products. A slight
  change in the composition of the group may cause
  a great change in the total quantity.

18
Things to keep in mind 2.

• Values have been converted from national
  currencies into US. Fluctuations in the exchange
  rate influence statistical values
• Values are always f.o.b. (exports) or c.i.f.
  (imports) and are exclusive of subsidies or taxes

19
National currency and US
20
Value and quantity
21
Value/quantity
22
Tariff data

• Main sources
• UN TRAINS
• WTO CTS
• Accessible through WITS (World Integrated Trade
  Solution)
• World Bank initiative

23
Tariff data snowploughs in Bangladesh
24
Tariff data snowploughs in Bangladesh
25
Assignment

• Compare avg tariffs agriculture (SITC 0)
  non-agriculture Iran. Which is higher? 2003
• Within agric. What is the pattern (crops,
  lvstck)
• Changes over time 2000 2004?