Bulletin – September 2020 Finance Insights from the New Economic and Financial Statistics Collection

The New EFS Collection

Over the past few years, the ABS, APRA and the Reserve Bank have worked together in close consultation with financial institutions to modernise and expand the data collected from Australia's financial sector. These data are collectively known as the EFS collection.^[1] The new data have enhanced the quality of information on financial institutions and their lending and borrowing activities available to policymakers and the wider community. The EFS collection has been a large-scale and complex project, involving considerable collaboration between the three agencies and the industry.

The EFS collection has been implemented in three phases:

• Phase one improved existing data on financial institutions' balance sheets used to compile Australia's financial aggregates. The financial aggregates are statistics on the stocks of money and credit outstanding in the Australian economy (Bank, Durrani and Hatzvi 2019). Since August 2019, the financial aggregates have been published using data from the EFS collection. The balance sheet data from the EFS collection are also used as an input into the national accounts finance and wealth estimates for Australia.^[2]
• Phase two improved existing data on housing and business finance. This phase of the collection has also provided policymakers with much more granular information on reporting institutions' lending, liabilities, and interest rates.
• Phase three has provided information on other aspects of reporting institutions' activity and performance, including profits and activity in specific financial markets. It will also provide information on the fees that reporting institutions charge and further information on specific financial products in the future. Some of these data are, or will be, used in the measurement of Australia's Gross Domestic Product (GDP).

One of the most important changes in the EFS collection is the more detailed definitions of the data reported.^[3] This helps institutions report consistent data which, in turn, increases the quality of the statistics compiled and used by the ABS, APRA and the Reserve Bank. A number of definitions were clarified or updated to align with international standards for compiling economic statistics. These include the definitions of different types of deposits, industry sectors, and the residency status of households and businesses (which are consistent with the compilation of Australia's national accounts). The definitions used in the EFS collection are accompanied by comprehensive guidance.

The EFS collection also captures types of financial institutions that were not included in previous data collected from the financial sector. This has improved the accuracy of the statistics compiled using the new collection. The new data are collected from authorised deposit-taking institutions (ADIs), including banks, building societies and credit unions, and non-ADIs (also known as registered financial corporations).^[4] Previously, only a small number of non-ADIs reported data to APRA. Legislative changes mean that more of these entities are now registered with APRA and reporting data as part of the EFS collection, which has increased the coverage of the activities of non-ADIs.

This article focusses on the data and findings from phase two of the EFS collection. In particular, it describes the more comprehensive set of statistics now available on reporting institutions' lending and interest rates. Much of these data are now published – in an aggregate view – in some of the statistical tables on the Reserve Bank's website. APRA publishes selected information on individual banks within the domestic market using the EFS collection. The ABS also uses the EFS collection to compile aggregate statistics on new lending to households and businesses.

The following two sections outline some of the key insights on the provision of finance to businesses and households from phase two of the new EFS collection.

Better Data on Business Finance

The EFS collection provides a more complete view of the way that Australian businesses access finance from the banking sector. The new collection includes data on business credit outstanding by business size. The measured stock of outstanding business credit is higher than previous estimates due to improved coverage and more comprehensive measurement (Graph 1). The Reserve Bank's previous measure was based on data reported by a smaller number of lenders and did not include business lending for some business purposes. Currently, the ADI and non-ADI lenders reporting these data account for a little over 95 per cent of total business credit outstanding. Business lending by size, industry and interest rate type are now published on the Reserve Bank's website in Statistical Tables D14 and D14.1.^[5]

These data have assisted the Reserve Bank in assessing developments in business finance since the onset of the COVID-19 pandemic. Large businesses drew on lines of credit in March and April and held these funds as deposits, though about three quarters of this has since been repaid. In contrast, the overall volume of lending to small and medium-sized businesses has been little changed since the onset of the COVID-19 outbreak. This could reflect weak demand for new loans due to the heightened uncertainty for businesses in the current environment (Lewis and Liu 2020). In addition, there are various short-term initiatives helping small and medium-sized businesses cover their operating costs, which lessen the need for finance from the banking sector.

Although weakness in lending growth to small and medium-sized businesses appears to be mostly driven by subdued demand, the availability of credit to businesses has also tightened since earlier in the year (Lewis and Liu 2020). Much of this reflects the application of existing lending standards in an environment of weak economic conditions and great uncertainty. In addition, banks are requiring a greater degree of verification of borrowers' information, and some banks are cautious about lending to new customers and to sectors significantly affected by the pandemic (such as smaller retailers, tourism and commercial property).

The EFS data on business lending by size have also been important for the implementation of the Reserve Bank's Term Funding Facility (TFF), which was announced in March 2020 as part of a comprehensive policy package (Reserve Bank of Australia 2020). The TFF provides a guaranteed source of low-cost funding to banks, and includes an incentive for the banking sector to increase lending to businesses, especially small and medium-sized businesses.^[6] This is because banks can borrow extra funding under the TFF if they increase their lending to businesses; for every dollar of extra lending to small or medium-sized businesses, banks can access five dollars of extra funding under the TFF (for large businesses, the amount is one dollar of extra funding). Where available, the Reserve Bank is calculating the extra funding from the TFF using the data on business lending from the EFS collection.

The EFS collection also provides more frequent and disaggregated data on the interest rates paid by businesses (Graph 2). These data are now monthly, instead of quarterly, and broken down into rates for small, medium and large businesses, instead of rates for small and large business loans. The definitions of business size are also more accurate. Previously, small business loans were defined as those loans that were less than $2 million, while all larger loans were considered to be large business loans. In other words, lending was classified by the size of the loan and not the size of the business.^[7]

Business interest rates are published in the new Statistical Table F7 and the Lenders' Interest Rates page on the Reserve Bank's website. These data have been important for monitoring how the Reserve Bank's recent package of policy measures is flowing through to business lending rates. Indeed, business lending rates have recently declined to historically low levels.

The new business interest rate data from the EFS collection also confirm our understanding that smaller businesses pay higher interest rates for finance than larger businesses. A key reason for the difference is that smaller businesses typically have a higher risk profile due to their more volatile revenue streams (Lewis and Liu 2020). However, small-business borrowers pay significantly lower interest rates on loans secured by residential property compared with their other loans (Graph 3). This is consistent with the security provided by their property reducing the risk for lenders relative to unsecured finance.

Better Data on Household Finance

The EFS collection also provides a better, more detailed view of the way that Australian households access finance from the banking sector. Households generally seek finance for two main purposes – to purchase a house or for consumption purposes (via a personal loan to buy a car, for example).

Housing finance

The stages of housing finance

The EFS collection includes data on four stages in the provision of housing finance. This is a considerable improvement on previously available data, as these new data provide more complete information on changes in the characteristics of and trends in housing finance. These data are reported by the largest providers of housing credit, accounting for around 95 per cent of total housing credit outstanding. The four stages of housing finance are:

• A housing loan application, which is typically received shortly after a contract of sale is agreed between the buyer and seller of a property. A housing loan application is the first stage reported in the establishment of a housing loan, and the earliest indicator of housing finance activity.
• A housing loan commitment, which exists for a loan once a borrower has accepted an offer of finance from a lender. These data have been published by the ABS since November 2019.^[8]
• A funded housing loan is one where the lender makes funds available to be drawn on by the borrower. These new data include details on the characteristics of housing lending, such as loan size, whether a loan has a variable or fixed interest rate, and whether a borrower makes repayments for the principal and interest of a loan or the interest component only.
• A new drawdown occurs when a borrower draws on the funds made available by the lender during settlement of a property. This is the final stage of the housing finance process, where the creation of credit has taken place and where the loan enters the Reserve Bank's credit aggregates.

The new EFS data on housing loan applications have been very valuable in recent months. Recently, applications for housing loans have been a little above the levels of earlier this year, despite the weak housing market activity observed since the end of March (Graph 4). The broadly stable level of housing loan applications in recent months is consistent with the very large volumes of refinancing activity that has occurred of late, as borrowers have sought to take advantage of lower interest rates and offers of cash back by refinancing an existing loan previously held with another lender.

The EFS housing loan commitments data provide a timely indicator of housing finance activity. They also give a better estimate of the average loan size for new housing lending than was previously available (Graph 5). Average loan sizes are calculated as the total value of new lending divided by the number of new lending facilities. Previous estimates of loan size were based on the number of loans extended by a lender. However, a borrower can have a single mortgage facility for a property that may include multiple loans, such as a fixed-rate loan for part of a mortgage and a variable-rate loan for the rest. The EFS collection includes data on the number of loan facilities instead of the number of loans. Since November 2019, the average new loan size (excluding refinancing) has been around $530,000, higher than the measure of about $480,000 based on the data available prior to the introduction of the EFS collection.

The new disaggregated EFS data available on housing loan characteristics can provide timely insights for policymakers. Monthly data are now available on the value and number of loans funded at different loan-to-valuation ratios (LVRs) and at different loan sizes (Graph 6). The LVR is an important indicator of the riskiness of a loan. Higher LVR loans have lower equity buffers to absorb any potential declines in the value of the property.^[9]

The characteristics of funded housing loans reported in the EFS collection have also been useful for explaining recent trends in housing finance. For instance, these data show a very sharp increase in the share of housing loans funded with fixed interest rates in recent months (Graph 7). The rise in housing loans funded with fixed interest rates is due to borrowers refinancing from variable- to fixed-rate housing loans (discussed below). As a result, fixed-rate loans now account for around 25 per cent of total housing credit outstanding.

Improving the transparency of mortgage interest rates

The EFS collection has greatly enhanced the transparency of interest rates paid by borrowers in Australia, which assists consumers in making better-informed choices about their finances. The absence of readily available data on interest rates paid by borrowers in relation to their mortgages had been highlighted as one factor impinging on competition in the Australian financial system (Productivity Commission 2018). Very few borrowers actually pay interest rates as high as the standard variable rates (SVRs) published by lenders. While SVRs are the reference rates against which variable-rate loans are priced, lenders also advertise a range of interest rates that are materially lower than their SVRs. In addition, most individual borrowers are offered, or may be able to negotiate, further discounts on the interest rate applied to their loan (Reserve Bank of Australia 2019).

To enhance the transparency of mortgage interest rates, the Lenders' Interest Rates page on the Reserve Bank's website and Statistical Table F6 now provide interest rates paid by borrowers on new and outstanding housing loans. This includes interest rates by loan repayment type (interest-only and principal-and-interest repayments), for variable- and fixed-rate loans, by different loan sizes, and by different LVRs. The Council of Financial Regulators has also worked with the Australian Competition and Consumer Commission to enhance the Mortgage Calculator on the Australian Securities and Investments Commission's Moneysmart website, based on the EFS collection (Council of Financial Regulators 2019).

The EFS data show significant differences in interest rates across new interest-only and new principal-and-interest loans (Graph 8). For example, new owner-occupier borrowers have paid around 60 basis points more, on average, for interest-only loans than for principal-and-interest loans over the past year or so. Higher interest rates for interest-only loans have been evident since 2015, following measures taken by APRA to place limits on investor and interest-only lending.^[10] By contrast, there is much less differentiation in the average interest rates paid by borrowers with different loan sizes or LVRs.

The information on new loans available from the EFS data is especially valuable given that competitive pressures are greater for these loans. These data confirm that new borrowers, on average, pay lower interest rates than existing borrowers (Graph 9). This reflects the tendency for competition to be strongest for borrowers who are in the process of shopping around for a loan. By offering lower interest rates to new or refinancing borrowers, lenders are able to compete for these borrowers without lowering the interest rates charged to existing borrowers.

The EFS collection also includes more comprehensive data on interest rates for new and existing fixed-rate loans. These data show that rates for new fixed-rate loans are on average around 20–50 basis points lower than lenders' advertised interest rates (Graph 10). These discounts are much smaller than those for variable-rate loans, which are on average roughly 150 basis points below lenders' SVRs. There is less incentive for lenders to compete for new borrowers by offering large unadvertised discounts on new fixed-rate loans than there is on new variable-rate loans. This is because for fixed-rate borrowers, lowering the advertised fixed interest rate does not result in a lowering of the interest rates charged to existing fixed-rate borrowers. In contrast, for variable-rate borrowers, lowering the advertised SVR would result in a reduction of the interest rates paid by all existing variable-rate borrowers. As a result, much of the competition for new variable-rate borrowers has generally occurred via unadvertised discounting instead.

These new interest rate data can also help to explain the sharp rise in the share of housing loans funded with fixed interest rates in recent months. Interest rates on new fixed-rate loans have declined by around 65 basis points since February this year, consistent with the fixed interest rates derived from interest rate swaps (the benchmark for pricing fixed-rate loans). The interest rates on new fixed-rate loans are now around 60–70 basis points below new variable interest rates (see Graph 7 above). A large proportion of the increase in new fixed-rate housing loans is due to borrowers refinancing their mortgages to take advantage of the lower level of fixed interest rates.

Personal finance

The data from the EFS collection also allow us to analyse trends in personal finance in more detail than was previously the case. Personal finance is extended to households for purposes other than housing, and includes products like credit cards and personal loans (such as for a holiday, some furniture, or whitegoods). This type of finance also includes margin loans – which are loans that enable households to borrow to invest directly in shares or managed funds – and finance leases, where the borrower essentially pays to lease an asset such as a car.

Following a steady decline for some years, personal credit has contracted sharply over the past six months or so, alongside the introduction of COVID-19 containment measures (Graph 11). Around half of the decline in the stock of personal credit outstanding since February has reflected a decline in the balances on credit cards. This is consistent with the decline in personal credit card transactions that occurred in March and April, as retail sales and household consumption declined. That is, households were spending less and so accumulating less credit card debt. In addition, repayments on credit card balances also declined during that period, but to a lesser extent than transactions. Taken together, these effects reduced the value of credit card debt outstanding.

The acceleration in the decline of personal credit since 2018 has occurred alongside the interim and final reports of the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry and the associated policy changes that came into effect at around the same time.^[11] For instance, amendments to credit card contracts in 2018 and 2019 prohibited lenders offering unsolicited credit cards and credit limit increases, tightened serviceability requirements, banned retrospective interest charges on balances benefiting from an interest-free period, and provided consumers with the ability to reduce their credit limit or terminate their contract.

More broadly, there has been a decline in the use of personal credit over the past decade or so (Graph 12). As a share of household disposable income, personal credit has been steadily declining since 2008 (in contrast to household debt to income; Lowe 2017). The decline in the stock of personal credit has coincided with households' increasing use of mortgage offset and redraw accounts. These accounts offer borrowers the ability to finance spending through home equity, acting as a line of credit that can be used for any purpose. Offset accounts are often linked to a debit card for convenience, and so are a substitute for personal credit products such as credit cards. For instance, a household could redraw from their mortgage, for which they are currently charged an interest rate of around 2–4 per cent, and use the funds to repay balances outstanding on a credit card, for which they are currently charged an interest rate of around 16–17 per cent following the expiry of the interest-free period.

Similarly, mortgage debt has also become a more attractively priced source of funding relative to personal finance products over the past decade, as mortgage interest rates have drifted lower while unsecured personal lending rates remained stable. The new EFS data on interest rates for personal finance show that households pay much higher interest rates on personal credit card balances, which are a form of unsecured finance, compared with products that are secured by residential property (Graph 13). This reflects the lower risk of secured lending, as the lender has recourse to the security pledged by a borrower in the event of default. These new EFS interest rate data for personal lending are available for public use in Statistical Table F8.

Conclusion

A broader range of information on the lending and borrowing activities of the financial sector is now available to policymakers and the wider community as a result of the introduction of the new EFS collection, and the quality of existing data has been greatly improved. The new data has aided policymakers' understanding of recent developments relating to business and household finance. Moreover, the interest rate data from the EFS collection have improved the transparency of interest rates paid by borrowers in Australia. This large-scale project has involved considerable cooperation and effort by the ABS, APRA, the Reserve Bank, and many financial institutions.

Footnotes

The author is from Domestic Markets Department, and would like to thank Shane Aves, Susan Black, Stephanie Crews, Kassim Durrani, Lea Jurkovic, Qiang Liu, Kateryna Occhiutto, Dmitry Titkov, and Charlie Wenk for their assistance with this work. [*]

The EFS collection is administered on behalf of the ABS and the Reserve Bank by APRA in APRA's role as the national statistical agency for the financial sector. The Financial Sector (Collection of Data) Act 2001 enables APRA to collect information from the financial sector for the purposes of performing its functions or to assist another financial sector agency to do so. [1]

See Australian Bureau of Statistics (2019), ‘Implementing the new Economic and Financial Statistics collection in ABS economic outputs’, Information Paper 5232.0.55.006. Available at: <https://www.abs.gov.au/ausstats/abs@.nsf/mf/5232.0.55.006>. [2]

Supporting documentation for the EFS collection is available at: <https://www.apra.gov.au/modernised-economic-and-financial-statistics>. [3]

APRA's list of Registered ADIs is available at: <https://www.apra.gov.au/register-of-authorised-deposit-taking-institutions>. [4]

Historical data on bank lending to businesses is available from the Reserve Bank at: <https://www.rba.gov.au/statistics/discontinued-data.html>. [5]

More information on the TFF is available at: <https://www.rba.gov.au/mkt-operations/term-funding-facility/overview.html>. [6]

Businesses with turnover greater than or equal to $50 million are classified as large businesses. For businesses with turnover less than $50 million, when the lender has an exposure of more than $1 million, the business is generally classified as medium. When the exposure is less than $1 million, the business is usually classified as small. [7]

For more information see Australian Bureau of Statistics (2019), ‘Upcoming changes to Lending to Households and Businesses’, Information Paper 5601.0.55.002, available at: <https://www.abs.gov.au/ausstats/abs@.nsf/mf/5601.0.55.002>. [8]

Borrowers with an LVR above 80 per cent are also typically required to pay for lenders' mortgage insurance (LMI). Although LMI should reduce the risk to the lender, borrowers who are required to pay for LMI may also be more cash-constrained and have lower capacity to pay back a mortgage. [9]

These measures have since been removed. For more information see APRA (2018), ‘APRA to remove interest-only benchmark for residential mortgage lending’, Media Release, 19 December. [10]

The final report of the Royal Commission into Misconduct in the Banking, Superannuation and Financial Services Industry is available at: <https://financialservices.royalcommission.gov.au/Pages/reports.aspx>. [11]

References

Australian Bureau of Statistics (2019), ‘Implementing the new Economic and Financial Statistics collection in ABS economic outputs’, Information Paper 5232.0.55.006.

Australian Bureau of Statistics (2019), ‘Upcoming changes to Lending to Households and Businesses’, Information Paper 5601.0.55.002.

Bank J, K Durrani and E Hatzvi (2019), ‘Updates to Australia's Financial Aggregates’, RBA Bulletin, 21 March.

Council of Financial Regulators (2019), ‘Quarterly Statement by the Council of Financial Regulators’, Media Release No 2019-02, 10 July. Available at <https://www.cfr.gov.au/news/2019/mr-19-02.html>.

Lewis M and Q Liu (2020), ‘The COVID-19 Outbreak and Access to Small Business Finance’, RBA Bulletin, 17 September.

Lowe P (2017), ‘Household Debt, Housing Prices and Resilience’, Speech at the Economic Society of Australia (QLD) Business Lunch, Brisbane, 4 May.

Productivity Commission (2018), ‘Competition in the Australian Financial System’, Inquiry report, 3 August. Available at <https://www.pc.gov.au/inquiries/completed/financial-system#report>.

Reserve Bank of Australia (2019), ‘Box D: The Distribution of Variable Housing Interest Rates’, Statement on Monetary Policy, November, pp 59–60.

Reserve Bank of Australia (2020), ‘Box E: The Term Funding Facility’, Statement on Monetary Policy, August, pp 1–3.