Implementing the new economic and financial statistics collection in ABS economic outputs

The purpose of this information paper is to inform users of upcoming changes to some ABS economic outputs from the implementation of a new, improved data collection called the Economic and Financial Statistics collection.

The Australian Prudential Regulation Authority (APRA) collects data from the financial sector (Authorised Deposit-taking Institutions (ADIs) and Registered Financial Corporations (RFCs)) for the Australian Bureau of Statistics (ABS) and the Reserve Bank of Australia (RBA). Some of these data collection forms have not been updated since they were first introduced in 2002. Over the past few years, APRA, the ABS, the RBA and the financial sector have worked together to modernise the data collection. This has been a large and complex project. The new set of forms and related guidance is called the Economic and Financial Statistics (EFS) collection. The EFS collection will be implemented in four phases.

1. The first phase involves new forms for balance sheets, including loans and deposits and other supplementary balance sheet forms.
2. The second phase will update current forms on housing and business loan approvals.
3. The third phase will provide information on other aspects of reporting institutions’ activity and performance, including profits, fees charged and activity in specific financial products and markets. The first phase data will be used in addition to this performance data in the compilation of Australia’s Gross Domestic Product (GDP).
4. The fourth and final phase will collect quarterly data on derivatives activity.
    

EFS will improve data quality. For example, it will provide:

• much better guidance on what to report, so institutions can provide consistent data;
• updated definitions to align with international standards for compiling economic statistics; and
• more detailed data on home loans to investors and first home buyers.
   

The improved data will be used in a number of ABS outputs (see Table 1), which will in turn improve the quality of these estimates.

The first EFS Phase 1 release of Australian National Accounts: Finance and Wealth (cat. no. 5232.0) will be on 19 December 2019 (September 2019 issue). Expected impacts of the new EFS collection include movement within the deposit market from ‘other deposits’ to ‘transferable deposits’ (due to more clarity in the reporting guidance to enable alignment of the concept of transferability required within the Australian System of National Accounts (ASNA) and a shift from housing to personal loans (due to reporting loans by purpose).

The new collection will significantly improve alignment with current ASNA requirements. The previous collection’s concepts, statistical units and disaggregation did not always align with detailed ASNA requirements for Australian National Accounts: Finance and Wealth (cat. no. 5232.0). The degree of alignment varied depending on the concept being measured; the consolidation basis on which it was reported; and the level of disaggregation within a reported concept. As a result, data sourced from the old APRA collection was subject to data modelling, adjustment and the substitution of data from other sources. One example of this is when the ABS implemented an updated statistical framework, the 2008 System of National Accounts (SNA) (an internationally recognised guide for compiling measures of economic activity). The amount of modelling, adjustments and data substitution materially increased to maintain consistency with this framework. The new EFS collection provided an opportunity to review and reduce the number of models and adjustments applied to align with the 2008 SNA framework and other ASNA requirements.

The new EFS collection removes the distinction between Banks and Non-bank Authorised Deposit-taking Institutions (ADIs). On 5 May 2018, changes were made to the Banking Act 1959 and from that date, ADIs were allowed to use the word ‘bank’ without restriction under the Banking Act. In line with this, the following changes will be made to the subsectors of the Depository Corporations sector in the Australian National Accounts: Finance and Wealth (cat. no. 5232.0) publication:

• the ‘Banks’ subsector will be renamed ‘Authorised Deposit-taking Institutions’ and will now include Banks and non-bank ADIs; and
• the ‘Other Depository Corporations’ subsector will be renamed ‘Other Broad Money Institutions’ and will exclude non-bank ADIs.

The reclassification will redistribute financial assets and liabilities between the two subsectors within the Depository Corporations sector. Where the Depository Corporations sector financial assets and liabilities are revised this will be driven by improved measurement from EFS rather than the impact of the reclassification.

For more information, including more detail on what items will change and information on the new table formats and series identifiers, please see Information Paper: Product changes to the Australian National Accounts: Finance and Wealth, 2019 (cat. no. 5232.0.55.005).

The first EFS release of this monthly lending product, using data from Phase 2 of the new collection, will be on 17 December 2019 (for the October 2019 issue). To reflect the new content, the publication will be renamed Lending Indicators, Australia (cat. no. 5601.0).

Overall, there will be more housing data available in the new publication, Lending Indicators, Australia (cat. no. 5601.0), compared with Lending to households and businesses, Australia (cat. no. 5601.0) but fewer data on business loan commitments. Some existing series will no longer be collected and some existing data series will be suspended from publication pending further analysis of the newly collected data. The ABS plans to increase the content of Lending Indicators, Australia (cat. no. 5601.0) over time.

The changes to the concepts and classifications are significant. There is a high likelihood of revisions in future reporting periods as APRA, the ABS and the RBA continue to work with the financial sector to ensure consistent reporting. It is expected data quality will continue to improve over time.

Where new concepts are similar to the current concepts, the ABS will produce an historic series back to July 2002 (i.e. a ‘backcast’ series) to avoid breaks in series.

For household lending indicators, the new data items available from the EFS collection will include:

• the number of new loan commitments to investors;
• investor lending by purpose (such as ‘construction of dwellings’ and ‘established dwellings’); and
• lending to households to buy residential land (in the current collection, this is reported as ‘personal finance’ whereas in EFS it will be a component of lending for households, to better meet user needs).
   

The key changes to existing housing data will include:

• Changes to the number and value of loan commitments to owner occupier first home buyers, partly because lenders have improved their identification of first home buyers and partly because of an improvement in the way the loans are counted.
• A shift from owner occupier to investment loans due to the underlying principle of EFS that loans will be classified by purpose, not product type. Owner occupier loans are only for the customer’s principal place of residency in EFS so a customer can only have one owner occupier loan. Other loans will be reclassified to ‘investor’ loans. Loans for non-housing purposes (e.g. to buy a car) but secured by a dwelling are often classified as ‘housing finance’ currently. In the new statistics, these will be classified according to the purpose of the loan (as personal finance if the loan is to buy a car, for example).
• A smaller number of loans. Many lenders currently report the number of loan accounts, even if two or more loans are for the same dwelling. Multiple loans or loan components (such as a fixed and variable component) for the same dwelling will be considered one lending facility in EFS. This should reduce the total number of loan commitments compared with current reporting.
   

There will be significant changes to new loan commitments to businesses, including:

• temporarily suspending some series such as new loan commitments to businesses by industry;
• ceasing some series in the new EFS collection, such as total commitments by purpose and lending commitments by state; and
• making new data available for new loan commitments to small and medium businesses by detailed purpose category, and by state for the purchase of property and construction.
   

The changes listed above mean there will be fewer data items than previously available on business loan commitments. Some series published in Lending to households and businesses, Australia (cat. no. 5601.0) will be suspended pending further assessment of data quality and the effect of EFS changes such as the inclusion of internal refinancing.

In addition, some items such as the total lending to businesses for purchase of property and construction will no longer be available at the state level.

There will be some new items available including fixed term lending to small and medium sized businesses by detailed purpose in original terms.

For more information, including more detail on what items will change and information on the new table formats and series identifiers, please see Information Paper: Upcoming changes to Lending to Households and Businesses, Australia, 2019 (cat. no. 5601.0.55.002).

EFS data is not directly used in compiling Balance of Payments and International Investment Position (cat. no. 5302.0) and International Investment Position, Australia: Supplementary Statistics (cat. no. 5352.0). The ABS Survey of International Investment (SII) is used. However, there is a small amount of overlap between data collected from EFS and SII. During a parallel run process in 2019 the ABS reconciled the overlapping data, for authorised deposit-taking institutions and registered financial corporations’ interaction with non-residents, from the two collections. The ABS then worked with data providers to resolve any significant reporting issues and ensure coherence between the two sources (please see the final section of this information paper, ‘Quality assurance of the new EFS data’ for more information).

EFS won’t flow through to published estimates of GDP, and therefore Gross State Product (GSP), until late 2021, when the ABS has data from all EFS phases and will do historical revisions in the Australian System of National Accounts (beyond the regular annual revisions which are limited to the previous four years of data).

This timing of the implementation in 2021 gives the ABS time to thoroughly assess the impacts ahead of implementing the EFS data in Australian National Accounts: National Income, Expenditure and Product (cat. no. 5206.0), Australian System of National Accounts (cat. no. 5204.0) and Australian National Accounts: State Accounts (cat. no. 5220.0). At this stage, the ABS expects impact on total GDP to be minor. We also expect minimal impact to GSP. The national impact will be distributed using employment indicators from the Labour Force Survey. Within the sectoral income accounts, we expect some redistribution between household interest paid on dwellings and consumer debt. The size of this redistribution is not yet known. The lag in implementing EFS data will mean there’s a period of incoherence between estimates published in Australian National Accounts: Finance and Wealth (cat. no. 5232.0) and the other Australian national accounts and international accounts products referenced in this paper, between late 2019 and 2021.

If needed for data quality reasons (e.g. to ensure significant economic events are captured in the estimates), the ABS may make some adjustments to the national accounts products listed above from late 2020 to reflect some of the new data from the first phase of EFS.

The ABS will publish more information on the expected impacts on the ABS’s other national accounts releases closer to the time of release.

The ABS will complete, and in some cases has already completed, multiple parallel runs to assess the data for each phase of the new EFS collection before publishing the new data. During parallel runs the financial sector provides data to APRA on both the current basis and the new basis and the ABS analyses the data from the new EFS collection while continuing to collect and publish data on the current basis.

As part of the parallel run process for Australian National Accounts: Finance and Wealth (cat. no. 5232.0), the ABS undertook checks to ensure coherence between the small amount of overlapping data from the new EFS collection and the SII. The result of this quality assurance process has no implications for the SII now or going forward. The SII remains the key statistical input for estimates of Australia’s financial interaction with non-residents in the Australian National Accounts: Finance and Wealth (cat. no. 5232.0) publication.

One of the most important changes in the new EFS collection is the increased detail and precision in definitions of the data to be reported. These definitions are accompanied by comprehensive guidance to assist ADIs and other reporting institutions to provide consistent data. APRA, RBA and the ABS have worked, and continue to work, with the financial sector to resolve queries about what to report through EFS and have produced answers to several frequently asked questions to ensure high quality reporting.

This release previously used catalogue number 5232.0.55.006