18.2 Data sources and estimation methods for balance of payments and international investment position statistics are frequently changing as some sources are found to be less satisfactory and new ones come to light. For example, in recent times the ABS has improved the sources and methods for education-related travel credits, financial services and financial derivatives. Improvements to sources for estimating portfolio investment in Australia will be introduced during 1999, while work is also underway on evaluating new sources for measuring the migrants’ transfers.
Longitudinal Survey of Immigrants to Australia
18.3 The Longitudinal Survey of Immigrants to Australia, being conducted on behalf of the Department of Immigration and Multicultural Affairs, collects information from about 5,000 recently arrived migrants and accompanying family members reflecting their experience of settlement in Australia. Migrants in the survey are approached on three separate occasions over three and a half years. The ABS has contributed to the survey to enable questions to be asked on the value of goods and funds brought into Australia by migrants, with a view to using these data to replace some of the current data sources used to estimate the migrants’ transfers items in the balance of payments.
18.4 The current methodology used to measure immigrants’ transfers is based on the Business Skills Category information supplied by the Department of Immigration and Multicultural Affairs, and the Migrants’ Transfers Model for migrants arriving under non-business skills categories. Although the Department’s information is considered reliable, there are reservations about data from the SOFUT (see paragraph 5.38), an important input into the Migrants’ Transfers Model.
18.5 The work on the Longitudinal Survey of Immigrants is expected to be completed during 1999 and the results of the survey may be used in a revised Migrants’ Transfers Model.
Portfolio investment in securities issued in Australia
18.6 Chapter 13 explained in some detail the complex methods used to measure foreign portfolio investment in shares, bonds and notes, and money market instruments issued in Australia. As pointed out there, the resident issuer generally does not know who holds the share or debt security issued. Therefore the ABS must approach the custodians as the primary data source so that the levels, financial flows and stock of investment can be estimated.
18.7 Currently, some of the larger custodians supply annual data on a security by security basis which provides the paid up and face value of securities held by non-residents at year end; a separate line is reported for the value held by residents of each partner country. Other custodians report the lines security by security on a quarterly basis. These data are supplemented by annual and quarterly aggregate data (i.e. not security by security). The data become the basis of the data models described in Chapter 13. However, data models’ ability to approximate real world activity is limited by the information that is currently reported, and the ABS has been consulting custodians for some time on ways to extend quarterly reporting and improve these data. Broadly the objective is to have all custodians report quarterly data on a security by security basis, with a separate line being reported for each partner country. Data to be reported would include the market value of security held, and the financial transactions in securities. The system would be extended to include holdings of financial derivatives issued by residents and held by non-residents through custodians.
18.8 The new collection methodology, which will be progressively implemented from late 1998, will involve all major custodians reporting quarterly on a security by security basis. On an ongoing basis, the data provider will report gross transactions and levels held at the end of the reference period, of each security line.
18.9 Security by security reporting in the medium term saves custodians from aggregating data to the various classification levels required on collection forms. In the longer term, collection at this level of detail will minimise the impact on data providers of future changes in statistical requirements. From an estimation perspective the individual security information allows the ABS to apply classification variables (such as institutional sector and maturity) more accurately, and facilitates a more flexible approach to estimation of aggregates such as accrued income.