Housing Credit MoM

Housing Credit MoM is the Housing row of the RBA monthly Financial Aggregates. Full detail is in tables D1, D2 and D3, starting monthly in September 1976. No acronym exists, and the RBA constructs the aggregates for Australia.

It covers debt funding in loans and securities from domestic intermediaries in Australia to the private non-financial sector. It means households and businesses, and housing credit funds housing. Owner-occupier credit funds the main home, generally one per borrower, while investor credit covers all other housing loans.

The panel is a census of reporting intermediaries, not a sample survey. Banks, building societies, credit unions and registered financial corporations reported balance sheets to APRA for the RBA and ABS, plus RBA additions from non-traditional lenders. Since July and August 2019, inputs use the EFS collection with purpose-based loan definitions.

It is not new mortgage lending, approvals or commitments, but growth of outstanding stock including securitised loans. It excludes cross-border and non-intermediated lending such as corporate bonds measured separately.

Stocks are balance-sheet snapshots at the end of the reference month. The release appears monthly about 4 weeks after month end on a pre-announced date. For example, the March 2019 reference month was published on April 30, 2019.

Each release shows monthly and year-ended growth for total credit, housing, personal and business plus broad money. Revisions and the August 2019 EFS and March 2023 warehouse-trusts changes were announced in advance.

Levels and growth revise after resubmissions, fresh seasonal factors and securitisation updates. The RBA warned revisions could run larger than normal after the August 2019 switch.

There is no target or threshold, the series reads as pace, faster or slower. Rapid growth can signal strength but also stability risks when debt is high. Splits track access and willingness to lend, while housing credit reflects and shapes housing conditions.

Housing growth reflects its two parts, with finer detail in Table D2. December 2018 shares give context, with housing at 62% of total credit or 95% of GDP. Households were 67% of total credit, and shares move slowly so writers must date any share.

• Owner-occupier, 42%: Loans for the main home, generally one per borrower, about two-thirds of housing.
• Investor, 21%: All other housing loans, more volatile, about one-third of housing.
• Personal, 5%: 5% of total credit outside housing.
• Business, 33%: Lending to non-financial businesses, separate from housing.

In March 2019, published on April 30, 2019, housing was +0.2% after +0.3% in February 2019. Year-ended housing fell to 4.0% from 6.1% a year earlier in March 2018. In the same release business was +0.5% after +0.3% and 4.9% year-ended after 4.0%. Personal was -0.3% monthly and -2.8% year-ended, total was +0.3% and 3.9% with broad money at +1.1%.

A soft housing print did not mean soft credit overall, so read housing on its own.

1. Collect end-of-month loan balances from every reporting intermediary in the APRA and EFS collection. The RBA queries unusual moves via APRA, counting only the final loan to avoid double counting.
2. Aggregate to a credit stock that includes securitised loans in the headline measure.
3. Seasonally adjust the stock with concurrent X-12-ARIMA factors re-estimated every month. The Bank has used this method since 2006 after annual ABS adjustment.
4. Break-adjust growth only by removing jumps from mergers, new banks, transfers and reclassifications. Published levels stay as reported, with breaks listed in technical notes.
5. Take same-basis adjusted stocks S for month t and t minus 1, subtract, divide by prior stock and multiply by 100. In symbols, g(t) = 100 x [S*(t) minus S*(t-1)] / S*(t-1). Year-ended growth uses the same method over 12 months on the same basis.

For illustration, adjusted stocks are $2,200.0bn for January and $2,213.2bn for February with no new break. Then February growth is 100 x 13.2 / 2,200.0, or +0.6%, shown as Housing plus +0.6%. A $5bn merger jump is removed before math while the level keeps the jump.

Do not calculate growth from published levels since levels lack break adjustment. Do not read owner-occupier jumps as new lending when they reflect reclassification.