Database Specific
Database Specific Abstract AbstractDespite the growing importance of international trade, driven in large part by the rise of globalisation and the accompanying international fragmentation of production, the availability of statistics on price change in international merchandise trade at more granular level is still limited. To fill this data gap, the OECD has developed this new Merchandise Trade Price Index (MTPI), which is constructed based on information on trade values and quantities as reported in the UN COMTRADE database. This release covers about 170 countries from 2011 to 2020. Indices by reporting country, estimated from data expressed in national currency, are available for both exports and imports, broken down by 30 products, aligned with the 2-digit level of the Classification of Products of Activity (CPA, version 2.1).
Source Source Contact person/organisation Contact person/organisationFor further information, contact the OECD Statistics Directorate at: STAT.Contact@oecd.org
Source Periodicity Source PeriodicityYearly
Data source(s) used Data source(s) usedUN COMTRADE
Data Characteristics Data Characteristics Periodicity PeriodicityYearly
Unit of measure used Unit of measure used Concepts & Classifications Concepts & Classifications Classification(s) used Classification(s) usedCPA version 2.1
Estimation EstimationConstructing trade price indices from detailed merchandise trade data using unit values requires an intricate process of data cleaning. A wide variation in unit values is typically observed for individual commodity codes, at least partially attributable to erroneous data in customs declarations, which requires a sophisticated method to detect outliers but one that still retains sufficient observations to reliably reflect price change of the product group. This approach builds on previous efforts, and refines them through the incorporation of three-stages of outlier removal process.
Index type Index typeLaspeyres Price Index (change compared to the previous period, expressed in national currency and benchmarked to the aggregate implicit trade prices using national statistics).
Key statistical concept Key statistical conceptUnit values are defined by trade value over quantity. These unit values are then divided by the unit values of the previous year to obtain unit value indices at HS 6-digit level. Unit value indices are then aggregated over countries and commodities using the Laspeyres formulae (weighted).
Validation ValidationWherever possible the unit value index estimates are validated and compared with existing data, including implicit price deflators from national accounts, official price indices for selected countries at the Standard International Trade Classification (SITC) 1-digit level, and global commodity prices.
Despite the growing importance of international trade, driven in large part by the rise of globalisation and the accompanying international fragmentation of production, the availability of statistics on price change in international merchandise trade at more granular level is still limited. To fill this data gap, the OECD has developed this new Merchandise Trade Price Index (MTPI), which is constructed based on information on trade values and quantities as reported in the UN COMTRADE database. This release covers about 170 countries from 2011 to 2020. Indices by reporting country, estimated from data expressed in national currency, are available for both exports and imports, broken down by 30 products, aligned with the 2-digit level of the Classification of Products of Activity (CPA, version 2.1).
For further information, contact the OECD Statistics Directorate at: STAT.Contact@oecd.org
UN COMTRADE
Yearly
Yearly
Unit values are defined by trade value over quantity. These unit values are then divided by the unit values of the previous year to obtain unit value indices at HS 6-digit level. Unit value indices are then aggregated over countries and commodities using the Laspeyres formulae (weighted).
CPA version 2.1
Constructing trade price indices from detailed merchandise trade data using unit values requires an intricate process of data cleaning. A wide variation in unit values is typically observed for individual commodity codes, at least partially attributable to erroneous data in customs declarations, which requires a sophisticated method to detect outliers but one that still retains sufficient observations to reliably reflect price change of the product group. This approach builds on previous efforts, and refines them through the incorporation of three-stages of outlier removal process.
Wherever possible the unit value index estimates are validated and compared with existing data, including implicit price deflators from national accounts, official price indices for selected countries at the Standard International Trade Classification (SITC) 1-digit level, and global commodity prices.
Laspeyres Price Index (change compared to the previous period, expressed in national currency and benchmarked to the aggregate implicit trade prices using national statistics).