Advertised Versus Paid
What Australian lenders advertise on a variable owner-occupier home loan, against what people who already hold one are actually paying. The advertised side is our own daily reading of every lender’s published rate; the paid side is the Reserve Bank’s monthly housing lending series. The two measure the market differently, and this page says how before it says by how much.
What is advertised, and what is paid
The mean across 389 loans from 101 lenders, taken at each loan’s headline advertised rate, as at 19 Oct 2026. Every lender counts once per loan, however large or small its book.
The midpoint of the same 389 loans on the same day. Half sit either side of it, so the lenders advertising as much as 13.75% cannot move it the way they move the average.
What is actually being paid on outstanding variable-rate owner-occupier housing credit in August 2026, weighted by loan balances. Source: Reserve Bank of Australia, RBA Statistical Table F6, Housing Lending Rates.
These two numbers are not measured the same way
The Reserve Bank’s figure is weighted by loan balances and covers only lenders holding more than $6 billion of housing credit, which it says captures over 95 per cent of total housing credit. That is a small number of large lenders. Our average covers 101 lenders, counts only the loans we list, and weights every loan equally, because market share is not published in the Consumer Data Right feed and we will not estimate it.
So the difference between the two is partly borrowers paying less than the list price, and partly the two series describing different markets. The figures below separate as much of that as the data honestly allows, and label plainly the part that cannot be separated. The Reserve Bank also notes that these rates are indicative only and are likely to be regularly revised, so the paid side is a revisable estimate rather than a settled number.
Where the gap comes from
| Measure | Gap | What it means |
|---|---|---|
| Advertised average against paid6.85% against 6.24% | 0.61pp | The headline number, and the one this page will not publish on its own. It mixes discounting with the difference in how the two series weight the market. |
| Of which: the shape of the list pricesOur own average against our own median, same loans, same day | 0.41pp | Measured, not estimated. This is purely the long tail of small lenders advertising high rates (the range runs 4.74% to 13.75%) pulling the average above the midpoint. A balance-weighted series would barely feel it. |
| What is leftOur median 6.44% against paid 6.24% | 0.20pp | The closest honest reading of how much less than the advertised rate is being paid. Still an upper bound: a median across every loan is not the same thing as a balance-weighted average across a small number of large lenders. |
| Control: the majors against everyoneInside the Reserve Bank’s own data, both sides balance weighted | 0.00pp | ANZ, CBA, NAB and Westpac against every lender reporting to the collection. This is how much “the official figure is dominated by the big four” actually moves it, and the answer is almost nothing. |
| Control: new loans against the bookLoans funded in August 2026 against loans already held | -0.01pp | Whether an existing borrower is paying more than someone taking out the same loan today. A negative figure is the loyalty tax; around zero means there is not one to find this month. |
Gaps are in percentage points, not per cent, so a difference between two rates cannot be misread as a rate. Rounding is applied once, at display.
The typical advertised rate is 0.20 percentage points above what is actually being paid. The average advertised rate is 0.61 percentage points above it, and 0.41pp of that difference is the gap between our own average and our own median rather than anything a borrower has negotiated. Source: Reserve Bank of Australia
How the two have moved
- Advertised: Variable, owner-occupier, P&I6.85%
- Paid: outstanding owner-occupier variable6.24%
The two lines start at different times and move at different speeds by construction. The advertised series is daily and begins when our archive does. The Reserve Bank’s is monthly and runs back to July 2019, when the collection behind it started; the chart draws the last 3 years of it so both lines are legible, and we hold all 86 months.
Every series we publish from F6
| Series | RBA id | Month | Rate |
|---|---|---|---|
| Owner-occupier, all loansThe average rate being paid across every owner-occupier housing loan on the books, weighted by the value of credit outstanding. | FLRHOOTA | August 2026 | 6.21% |
| Owner-occupier, variable rateThe average rate paid on outstanding variable-rate owner-occupier loans, weighted by the value of credit outstanding. | FLRHOOVA | August 2026 | 6.24% |
| Owner-occupier, all loans, major banks onlyThe same measure restricted to ANZ, CBA, NAB and Westpac. Its distance from the all-institutions figure is how much F6 being dominated by the majors moves the number. | FLRHOOTL | August 2026 | 6.21% |
| Owner-occupier, principal and interestOutstanding owner-occupier loans on principal-and-interest repayments, which is the great majority of the book. | FLRHOOP | August 2026 | 6.19% |
| Owner-occupier, interest onlyOutstanding owner-occupier loans on interest-only repayments. | FLRHOOI | August 2026 | 6.85% |
| Owner-occupier, new loansThe average rate on owner-occupier loans actually funded during the month, weighted by the value of new lending. | FLRHOFTA | August 2026 | 6.24% |
| Owner-occupier, new variable-rate loansThe average rate on variable-rate owner-occupier loans funded during the month. This is the series that lines up in time with an advertised variable rate. | FLRHOFVA | August 2026 | 6.23% |
| Investor, all loansThe average rate being paid across every investment housing loan on the books, weighted by the value of credit outstanding. | FLRHIOTA | August 2026 | 6.44% |
| Investor, variable rateThe average rate paid on outstanding variable-rate investment loans. | FLRHIOVA | August 2026 | 6.47% |
| Investor, new variable-rate loansThe average rate on variable-rate investment loans funded during the month. | FLRHIFVA | August 2026 | 6.40% |
10 of the 42 series the Reserve Bank publishes in RBA Statistical Table F6, Housing Lending Rates, chosen for this comparison. The full table is published free by the RBA. Source: Reserve Bank of Australia. The Reserve Bank does not endorse this site or anything on it.
CompareThat Rate Tracker, advertised versus paid home loan rates, 19 Oct 2026; Reserve Bank of Australia, RBA Statistical Table F6, Housing Lending Rates, August 2026. https://comparethat.com.au/rates/advertised-vs-paid
Home Loan Market AveragesCompare Home Loans
How these figures are calculated
- Where the numbers come from
- Every provider’s own published product data, read directly from their Consumer Data Right product reference endpoints and re-checked daily. We do not survey lenders and we do not ask them what to print. See where our data comes from.
- What is averaged
- Every product we list in the category, taken at the headline advertised rate shown on its card, so an average on this page is taken on the same rates as the tables on the rest of the site. We do not compare every product in the market, so these figures describe the products we list. A segment narrows that set to one combination of rate type, borrower purpose, repayment type or term, and the figure carries the count of products and providers behind it. The home loan headline is one such loan: each lender’s variable owner-occupier principal and interest rate, because a fixed special and a variable rate averaged together describe no loan anyone can take out.
- Loans and their comparison rates
- Each home, personal and car loan average is printed with an average comparison rate, taken over the loans in the average rate that publish a comparison rate. A loan whose lender publishes no comparison rate is not listed in our tables unless we choose to list it, so it is usually left out of every figure here, including the counts of products and providers. Where a loan is listed more than once, such as once per fixed term, each listing is judged on its own: a listing with no comparison rate is left out unless we list the loan anyway, and the loan counts at its lowest listed rate. Where we do list a loan without a comparison rate, our tables show “Comparison rate not published” beside its rate, and it counts in the average rate, the median and the counts, but never in the average comparison rate or as the lowest rate. Personal and car lenders often price by risk and publish a range of rates for one loan: the averages take the lowest rate in that range, the rate our tables sort on, and most applicants are offered more than that. A loan listed with both fixed and variable rates counts once in the headline average, at its lowest listed rate, and once in each of the fixed and variable segments it is listed in.
- Each product counts once
- The averages are unweighted. A lender with four hundred loans on its book and a lender with four count the same per product, because market share is not published in the feed. These figures measure what the market advertises, not what Australians hold.
- Advertised rates, not offered rates
- Providers publish the rate they advertise. Package discounts and anything negotiated at application are not in the data, and where a lender publishes a risk-based range only its lowest rate is averaged, so a borrower with a strong application may be offered less than the figures here, and one with a thin file may be offered more.
- One reading a day, kept
- The aggregate is computed once per calendar day and stored, so a figure published on a given date does not move afterwards, unless we change how a figure is calculated. Then the whole series is recalculated on the new basis, so a change between two readings is never a change in our method. The home, personal and car loan series were last recalculated on 7 Oct 2026, to count only the loans we list. The archive currently holds 90 daily readings, the first taken on 22 Jul 2026.
- How a move is measured
- A change is the current average against the oldest reading in the last 40 days, not against a fixed calendar month: the series only holds a row for a day the aggregation ran. Every movement figure names the date it is measured from.