5232.0 - Australian National Accounts: Finance and Wealth, Sep 2019 Quality Declaration 
Latest ISSUE Released at 11:30 AM (CANBERRA TIME) 19/12/2019   
   Page tools: Print Print Page Print all pages in this productPrint All RSS Feed RSS Bookmark and Share Search this Product

Technical Note

This issue of Australian National Accounts: Finance and Wealth (cat. no. 5232.0) includes historical revisions. These revisions, spanning the entire time series back to June quarter 1988, are needed to reflect changes in data sources, concepts, classifications and methods and to maintain consistent time series.

This publication includes revisions for:

  • the incorporation of revised estimates sourced from phase 1 of the Economic and Financial Statistics (EFS) modernisation program;
  • improvements to the time series that incorporated the ABS Superannuation reporting standards for the Australian Prudential Regulation Authority (APRA) regulated funds implemented in September quarter 2017;
  • refining the assumptions used in the Australian Taxation Office (ATO) and ABS models that generate the estimates for Self-Managed Superannuation Funds (SMSF);
  • improvements to transaction estimates;
  • review of the sectoral classification of holding companies in EFS data; and
  • other quality assurance work undertaken during the historical revision cycle.

The incorporation of revised estimates sourced from phase 1 of the Economic and Financial Statistics (EFS) modernisation program

This publication incorporates estimates from the first phase of the EFS modernisation program. The modernisation program and its implications for ABS economic outputs are discussed in the previously released information papers:
In introducing the revised EFS data collections, a number of quality, classification and implementation issues arose. APRA, the Reserve Bank of Australia (RBA) and ABS have undertaken extensive and ongoing discussions with data providers on the EFS collection to clarify and address implementation issues. To help with quality assurance, the new and old data collections were conducted in parallel for the March and June quarters 2019. The ABS considers the quality and comparability of the statistics to have improved significantly with the introduction of the new EFS collection. Data quality is expected to continue to improve over time as lenders become accustomed to the new reporting basis and further refine the data they report.

Where the EFS data resulted in a shift in the level from the previous APRA domestic books collection the level shifts were, in most cases, backcast to the June quarter 2002. In a small number of cases the size of the revision required the series to be revised through the entire time series back to June quarter 1988.

This release removes the distinction between Bank and Non-bank ADIs to reflect the 2018 changes to the Banking Act. In line with this, changes have been made to the subsectors of the Depository Corporations Sector. The ‘Banks’ subsector has been renamed ‘Authorised Deposit taking Institutions’ and includes ‘Banks’ and ‘non-bank ADIs’.

The ‘Other Depository Corporations’ sub-sector has been renamed ‘Other Broad Money Institutions’ and excludes ‘non-bank ADIs’.

In March 2018, amendments to the Financial Sector (Collection of Data) Act 2001 broadened the scope of entities that must report information to APRA, resulting in an increased number of Registered Financial Corporations (RFCs). RFCs are classified to the ‘Other Broad Money Institutions’ subsector. This change in the scope of the RFC collection was backcast within the time series to the entry of each new institution into the Australian financial markets. The APRA RFC collection includes Securitisation units. These units are excluded from this sub sector and included in the Securitiser subsector within the publication.

The largest impact of the EFS phase 1 reporting has been the reclassification of financial assets and liabilities between the two subsectors within the ‘Depository Corporations’ sector, and in particular the deposit and loan market for all counterparties. Where the ‘Depository Corporations’ sector loans and deposits have been revised this has been driven by improved measurement from EFS rather than the impact of the reclassification.

Deposit liabilities of the Depository Corporations sector

Graph 1. Revisions to Depository Corporations deposit liabilities by sector - June quarter 2019

Graph 1 shows Revisions to Depository Corporations deposit liabilities by sector - June quarter 2019

In the June quarter 2019 the total deposit liabilities of ‘Authorised Deposit-taking Institutions’ and ‘Other Broad Money institutions were revised down $11.1b, of which other deposits were revised down $168.8b while transferable deposits were revised up $157.7b. The new EFS reporting showed downward revision to household deposits for over a 10 year period, with a revision of $26.8b in the June Quarter 2019.

Loan assets of the Depository Corporations sector

In the June quarter 2019 the loan assets of depository corporations were revised up $51.6b. Loans to Other Financial Corporations and Households were revised up $41.7b and $39.7b respectively. These were partly offset by a downward revision to loans to Other Private Non-Financial Corporations (-$87.7b). The upward revision to household loans was mainly due to the misclassification of unincorporated loans being classified as loans to private non-financial corporations.

Graph 2. Revisions to Depository Corporations loan assets by sector - June quarter 2019

Graph 2 shows Revisions to Depository Corporations loan assets by sector - June quarter 2019

ABS Superannuation reporting standards for APRA regulated funds

The September quarter 2017 issue implemented for the first time the ABS Superannuation reporting standards for APRA regulated funds. Reported data was received from the September quarter 2016, and a time series was backcast for all financial instruments and counterparty sectors back to September quarter 2005 to ensure there were no series breaks. Recent quality assurance work has revealed that some series from this collection was not backcasted appropriately and some incorrect counterparties were allocated for the accounts receivable and payable market. The improvements made to the APRA regulated superannuation time series estimates for this release of Australian National Accounts: Finance and Wealth had the following impacts:

  • significant upward revision to pension fund holding of unlisted retail trusts;
  • upward revision to pension fund holdings of listed Other Private Non-Financial Corporation (OPNFC) shares, these are more than offset by downward revisions to holdings by Other Financial Corporations and Non-Money Market investment Funds (NMMF);
  • downward revision to pension fund holding of listed ADIs, offset by upward revision to rest of the world and household holdings; and
  • significant downward revisions of pension fund accounts payable and receivable with households; and accounts receivable from the rest of the world.

Self-Managed Superannuation Funds

This release includes a new table which presents separately a balance sheet of SMSFs (Table 19), which is a part of the overall pension fund sector (Table 18) in this publication. In light of this, the ABS reviewed and improved the SMSF modelled estimates provided by the ATO and the subsequent detailed (financial instrument by counterparty) estimates generated by the ABS included within Table 18 and 19. The ABS implemented a new method to estimate the value of quarterly total assets, and types of financial assets and liabilities for the quarters when ATO compliance data is not available. For the detailed ABS model, research was undertaken including contacts with industry specialists to update the assumptions in the model and formulate the detailed financial instruments by counterparty included within the SMSF estimates.

The upward revision to the net equity in reserves of pension funds (Tables 18 and 34) is mainly due to including SMSF investments in non-financial assets which were not previously captured in the estimates.

Graph 3 and 4 below shows the revisions to the time series from the previous quarter for pension fund and household total assets and components due to the implementation of the EFS data and quality assurance work undertaken for the historical revision cycle.

Graph 3. Revisions from the previous quarter to Pension Fund Financial Assets
Graph 3 shows Revisions from the previous quarter to Pension Fund Financial Assets

For the 30 year time series, Graph 3 shows total assets of pension funds were revised up in the early 2000s and the revisions declined after the global financial crisis (GFC) in 2008. Since the GFC total assets have been downwardly revised and have picked up in recent quarters to display small revisions. Within the financial instruments, equity assets have been revised up by nearly $100b in recent quarters (due mainly to holdings of unlisted retail trusts) while accounts payable was revised down by a similar magnitude, for households and rest of the world sectors.

Graph 4. Revisions from the previous quarter to Household Financial Assets
Graph 4 shows Revisions from the previous quarter to Household Financial Assets

Similarly to pension fund revisions, Graph 4 shows total assets of households for the 30 year time series were revised up in the early 2000s and the revisions declined after the GFC in 2008. Since the GFC, total assets have been downwardly revised and have picked up in last few years, with a significant spike in June quarter 2019 related to an actuarial adjustment of commonwealth general government unfunded superannuation liabilities. The similar revision pattern for total assets between households and pension funds reflects the household’s largest financial asset being insurance technical reserves (ITRs) generated by pension funds. For this revision cycle, these ITRs are driven by SMSFs.

The most significant liability of households are their loans, and the graph shows that the new EFS data has revised up both short and long term loans since the mid-2000s from $20b to nearly $40b in June quarter 2109. The significant downward revision to accounts payable of nearly $60b in the June quarter 2019 was due to misclassification of those payable from pension funds.

Transaction estimates

The ABS implemented the following improvements to the derivation of transactions:
  • To compile estimates for the listed equity market, the ABS needs market capitalisation data and shares issued (transactions) during the quarter. This information is sourced from the Australian Securities Exchange (ASX). To produce accurate quarterly transaction estimates, information on delisting activity from the ASX needs to be implemented. Prior to this historical revision cycle, delisting activity was included in the transaction estimates from 2014 onwards. For this release, delisting activity has now been included in the transactions from March quarter 1997 to June quarter 2014. Its implementation impacted most transaction estimates within listed equity markets.
  • The ABS implemented a new method to calculate unlisted equity liability transaction in wholesale trusts (classified within the NMMF sector). The method was based on a model that took into account the asset classes, their overall proportions within the NMMF sector (which was used as a proxy for wholesale trusts) and their respective price movements. The new method had impacts on sectors that hold significant amounts of wholesale trusts, such as pension funds, life and non-life insurance corporations and other NMMF.
  • For the rest of the world (RoW) asset and liabilities, the data source for stocks and transactions for some sectoral instruments and counterparties is the ABS Survey of International Investment (SII). For the sectors for which SII does not have the relevant data for interactions with the RoW, for example pension funds, APRA data is used. The APRA data does not provide transaction estimates. A basic method based on price movements (e.g. exchange rate between the $A/$US) of financial instruments was previously applied to derive these transactions. A new, more sophisticated, method using weighted averages of exchange rates and MSCI indexes (where applicable) is now used for the price changes to derive the new transaction estimates. The new method was developed for the relevant sectors and instruments for both debt and equity assets and liabilities. The method had significant impacts on the pension fund sector, and specifically the derivation of the ITR transaction estimates, which in turn had impacts on household net lending/borrowing estimates derived from the financial account.

Review of the sectoral classification of holding companies in EFS data

Recent guidance provided to reporting entities such as ADIs states that the holding company (that is the unit that holds the equity assets of the ADI’s direct investment in their subsidiary corporations such as Life Insurance) be classified as financial auxiliaries. Further, the EFS domestic books consolidation principles specifically states to exclude all subsidiaries. This has resulted in ADIs reporting their unlisted equity assets of their investment in subsidiaries as investment in Financial Auxiliaries (classified as other financial corporation sector within the publication). Ideally, for national accounts purposes, the investment in the holding company (financial axillary) is “looked through” and the ADIs direct investment in the life insurance corporation sector (in this example) is recorded. For the historical revision cycle, we applied a model to look through the significant investment by ADI in the OFC sector to the primary activity of their subsidiaries. This has resulted in upward revisions of ADIs unlisted equity holdings of Life Insurance, Non-Life Insurance corporations and RFCs.

Other quality assurance work

The following additional improvements were made during the historical revision cycle:
  • targeted quality assurance work to produce better estimates within the sectoral financial balance sheets for Money Market Funds, Non-money Market Financial Investment Funds, and Other Financial Corporations; and
  • the quality assurance processes associated with the historical revision have resulted in the identification and removal of a number of breaks in time-series. These series breaks were the result of revisions outside historical revisions windows; new collection forms being implemented without back casting methods and errors in the compilation process.