5342.0 - Balance of Payments Statistics, Information Paper on Quality , 1996  
ARCHIVED ISSUE Released at 11:30 AM (CANBERRA TIME) 20/02/1996   
   Page tools: Print Print Page Print all pages in this productPrint All  
Contents >> Chapter 2. Analysis of accuracy of the statistics >> Consistent behaviour of time series

2.45. An analyst of economic statistics is usually concerned with the behaviour of series over time. Analysis of individual time series can provide important information about changing patterns occurring in the economy. By relating two or more series, information can be obtained about economic behaviour and structural relationships over time. Within the balance of payments there are many such relationships that can be established, e.g. freight earnings on imports and merchandise imports, passenger fares earnings and travel, and income on foreign investment related to the level of investment.

2.46. Various relationships may also be established between balance of payments statistics and other series. An article on Balance of Payments Ratios, published in Balance of Payments and International Investment Position, Australia, 1993-94 (5363.0) gives many examples of such ratios.

2.47. A consistent relationship over time between various series may be an indication of data accuracy. An inconsistent relationship between the series being compared (where a consistent relationship is expected) may indicate errors in the statistics, changed economic circumstances, special factors, or significant statistical noise (i.e. the data are suspect but there is insufficient information available to resolve whether the data are accurate or not).

2.48. Some of the more direct relationships between series occur when the data sources for the series are related. For example, there is a relationship between the two balance of payments components shipment debits and imports f.o.b. The principal component of shipment debits is freight on imports earned by non-resident carriers and the main source for this estimate is international merchandise trade statistics compiled from customs documents. These statistics are also the main source for the merchandise imports estimates in the balance of payments. As might be expected, given the common source data, the ratio of shipment debits to merchandise imports has been largely consistent over many years, although exhibiting a gradual reduction in the last ten years. This change in ratio may, for example, have occurred as a result of competitive forces producing lower freight rates in the face of surplus carrier capacity, increased freight being carried by Australian operated craft, a compositional change from cargoes with lower to higher value-to-volume ratios or, perhaps, due to efficiency gains in cargo handling and carrier running costs.

2.49. Another example of a relationship between series comes from income yield analysis. Graph 2; shows selected investment income yield series for categories of Australia's foreign borrowing. Yields are calculated as interest payable abroad for a year expressed as a percentage of corresponding period average levels of debt. The accuracy of the underlying series may be tested by comparing the derived yields with representative market yields for different classes of debt and differing maturities, by examining the implied yields in balance of payments statistics at a finer level of disaggregation, and by looking at the contributions of individual enterprises. This type of data validation will usually identify any errors in reported data and improve understanding of the underlying data relationships. In the case of Graph 2; the broad correspondence between the series in the graph suggests that there are no fundamental errors in the underlying series. The shift, from 1986-87, in the yield on official sector borrowing relative to other yield curves is a reflection of the declining proportion of very long-term (over ten years) debt domiciled overseas in total official sector borrowing. While in 1984-85 and 1985-86 long-term debt domiciled overseas with ten years or more to maturity was about 30% of the total official sector borrowing, by 1987-88 it had fallen to about 10%. In contrast, short-term (repayable within one year) debt domiciled overseas has increased from less than 10% of official sector borrowing in the early 1980s to about 30% by 1993-94.



      GRAPH 2: ANNUAL INCOME YIELDS - FOREIGN INVESTMENT IN AUSTRALIA, BORROWING







Previous PageNext Page