On this page
New Australian small-business loans averaged 7.23% a year in August 2026, according to the RBA. ABS data separately show new small-business fixed-term commitments excluding refinancing fell 6.9% over the year to the June quarter. Rates and lending volumes measure different things; neither is an individual approval forecast.
This reference brings together official statistics and linked SMBLoan studies. The observation period appears beside every number. A source checked in October can still describe an earlier month, quarter or financial year.
Key lending and business finance statistics
| Measure | Value | Observation period |
|---|---|---|
| New small-business loan average | 7.23% p.a. | August 2026 |
| Outstanding small-business loan average | 7.47% p.a. | August 2026 |
| New medium-business loan average | 6.18% p.a. | August 2026 |
| New large-business loan average | 5.54% p.a. | August 2026 |
| Small–large new-loan difference | 1.69 percentage points | August 2026 |
| Small-business new commitments excluding refinancing | $13,996.1 million | June quarter 2026 |
| Annual change in those commitments | −6.9% | June 2026 vs June 2025 |
| Working-capital commitments | $631.3 million | June quarter 2026 |
| Annual change in working-capital commitments | −1.2% | June 2026 vs June 2025 |
| External refinancing commitments | $1,830.0 million | June quarter 2026 |
| Internal refinancing commitments | $1,907.9 million | June quarter 2026 |
| Bank branches | 3,205 | June 2025 |
| Branch reduction since June 2017 | 2,489 | June 2017–June 2025 |
| Motor vehicles’ share of new PPSR registrations | 69% | 2024–25 |
| SMEs’ Commonwealth contract value | $14.703 billion | 2024–25 |
| Three-year survival of business entry cohort | 49.2% | 2022–23 entries to June 2026 |
Sources behind the numbers
- RBA loan rates and repayment model
- ABS borrowing by purpose, state and refinancing
- APRA bank branch access
- AFSA collateral and PPSR activity
- Commonwealth procurement statistics
- ABS business cohort survival
Primary sources: RBA lending rates; ABS Lending Indicators, June 2026; ABS business counts. The study pages link their downloadable source summaries and describe the calculations.
Loan rates: an average is not an offer
The new-loan and outstanding-loan averages cover different pools of lending. Security, amount, term and borrower mix can change the result. A smaller business cannot infer its price by adding the average small–large gap to a quoted large-company rate. Compare written offers using the amount received, all repayments, mandatory fees and security.
For the practical comparison, read the business loan interest-rate guide. It explains why an annual rate, flat fee, factor and invoice charge cannot be ranked by their headline percentages alone.
Borrowing volumes: commitments are not balances
The ABS series used here records new fixed-term commitments during a quarter. It does not measure how much credit is currently outstanding, all applications lodged, or approval percentages. Refinancing is shown separately because replacing debt does not necessarily provide additional money for new activity.
The June quarter result also differs from the earlier March-quarter headline of 36.9% annual growth in working capital commitments. That older growth rate belongs to a different quarter and base. The current report preserves its earlier CSV while displaying the new observation prominently.
Payment delays and the cash cycle
The Slow Payer Index is a dated July 2026 extract of Payment Times reports. It compares reporting entities’ payment measures, not the experience of every small supplier. Separate medians across reporting entities do not sum to a national invoice weighted distribution. The raw filing extract is needed to fully reproduce the study, so its current page explains that limitation.
The working capital calculator is different again: it models expenses over a selected period using ATO industry ranges. Its output is an illustration to test against a business’s own receipts and payment calendar, not a survey of cash reserves.
Collateral and physical banking access
A PPSR registration is not a loan or a unique borrower. Several registrations can relate to one arrangement, and a registration can remain active for years. The vehicle share therefore describes the register’s composition, not the percentage of business loan dollars spent on cars.
Branch counts show physical service locations. They do not measure online credit availability, approval speed or travel time to a specific service. A regional business can have digital borrowing options and still face additional costs depositing cash or resolving an account problem in person.
What these statistics cannot establish
There is no verified whole-market approval rate in this roundup. Platform conversion figures, a lender’s marketing claim and a representative market measure are different things. A source that covers only successful applications also cannot show the rejection rate without the unsuccessful base.
Do not use a population statistic to infer that a particular loan is affordable. The business needs a forecast of cash available after ordinary expenses, tax and existing debt. Guarantees and security can create consequences beyond the scheduled instalment.
Archive:2017 online-lender application data
The earlier article used Become/LendingExpress platform data for 15 July–15 October 2017. That was a platform-specific historical sample, not a measure of today’s Australian lending market. Its images remain preserved in the publishing package, but obsolete graphs and the unrelated 2022–2023 RBA table no longer sit beside current statistics without context. No present-day eligibility or lender limit is inferred from them. Its loan sizes, business ages and state mix describe that channel, not every Australian borrower. For a separate dated market benchmark, see the loan rate study.
Citation and maintenance
For a quotation, use the exact measure, period and source in the table. “Small-business lending fell” is too broad when the source specifically measures quarterly fixed-term commitments excluding refinancing. The linked research pages explain each denominator and any calculation.
For source questions or corrections, contact hello@smbloan.com.au. This page was checked on 10 October 2026. It is maintained manually; a future release from an agency does not automatically change these figures.