Revisions
The Canadian International Merchandise Trade (CIMT) Program of Statistics Canada produces monthly international merchandise trade values, price indices and volume indices on both a customs- and balance of payments-basis. These statistics are prepared under tight deadlines and depend primarily on large volumes of administrative records received from the Canadian Border Services Agency and the United States Customs and Border Protection Agency. In accordance with the agreement on the exchange of import data, Canadian and United States international merchandise trade data are released simultaneously by Statistics Canada and the United States Census Bureau approximately 35 days after the end of the reference month.
Factors influencing revisions include late receipt of Customs documentation, incorrect information on Customs forms, replacement of estimates with actual data, changes in classification of merchandise based on more current information, and changes to seasonal adjustment factors.
In general, merchandise trade data are revised on an ongoing basis for each month of the current year. Current year revisions are reflected in both the customs- and balance of payments-basis data. The previous year's customs-basis data are revised with the release of the January and February reference months as well as on a quarterly basis. The previous two years of customs-basis data are revised annually and are released in February with the December reference month. The previous year's balance of payments-basis data are revised with the release of the January, February, March and April reference months. Revisions to balance of payments-basis data for previous years are released annually in December with the October reference month.
Seasonal Adjustment
Seasonal adjustment of customs- and balance of payments-basis values and price and volume indices is performed at an aggregated commodity grouping level of the North American Product Classification System (NAPCS). Customs- and balance of payments-basis values are also seasonally adjusted at the principal trading partner level of geographical detail. Monthly fluctuations can occur as a result of weather patterns, number of trading days, roving holidays (such as Easter) and institutional factors (such as scheduled factory shutdowns). In order to isolate turning points or trends in the basic data, it is necessary to eliminate this effect of seasonal movement. To remove seasonal fluctuations from time series, Statistics Canada uses the SAS® X12 procedure (SAS Institute Inc., 2010) as well as an adaptation of the US Census Bureau X-12-ARIMA Seasonal Adjustment program (US Census Bureau, 2010). The seasonal adjustment process is applied following the Statistics Canada Quality Guidelines.
Revised data are available in the appropriate data tables and statistical products.
Reference
SAS Institute Inc. (2010), "The X12 Procedure", SAS 9.2 Documentation: SAS/ETS 9.22 User's Guide, Cary, NC: SAS Institute Inc.
US Census Bureau (2010), X-12-ARIMA Seasonal Adjustment Program, Version 0.3, Washington, DC.