One published number tells you whether your fixed rate resets badly
It is the comparison rate, and almost nobody reads it this way. Measured across every Australian lender publishing home loan rates under the Consumer Data Right.
When a fixed term ends, the loan reverts to a variable rate set by the contract. No lender advertises what that rate is. Most of the time it is unremarkable — 77% of the 522 fixed rate lines measured here revert to less than the fixed rate they advertise, so rolling off would be a small cut at today's prices, not a shock.
But 45 lines revert more than a full percentage point above their own advertised rate, and there is a published number that finds them. Compare the comparison rate with the advertised rate. If the comparison rate is more than about half a point higher, the loan resets high.
That test caught 45 of the 45 high-resetting lines in this data, with 16 false alarms across the other 477.
Be clear about what this is. It is arithmetic, not a prediction. The regulated comparison rate assumes the loan runs 25 years, so a 3-year fixed rate's comparison rate is already three years of the advertised rate blended with twenty-two years of the reversion. The reversion was never hidden — it was folded into a number lenders are required to publish. This page is a decoder, not an exposé.
What it looks like when it bites
Bankwest publishes two owner-occupied fixed loans for borrowers at 90.01–95% of value. At three years they advertise exactly the same rate.
| Product | Advertised | Comparison | Reverts to |
|---|---|---|---|
| Complete Fixed Home Loan Package | 7.49% | 7.45% | 6.91% |
| Fixed Rate Home Loan | 7.49% | 8.62% | 8.89% |
Identical headline rate. 1.99 percentage points apart once the fixed term ends — about $9,930 a year in interest on a $500,000 balance. The only published number that separates them before you sign is the comparison rate.
Where the reversion sits, against the rate the same line advertises
The lines that reset hardest
| Lender | Product | Deposit band | Advertised | Comparison | Reverts to |
|---|---|---|---|---|---|
| People First and Heritage Bank | Standard Fixed Rate Home Loans - Owner Occupied | up to 95% of value | 6.29% | 8.08% | 8.73% |
| Suncorp Bank | 1,2,3 or 5 Year Fixed Rate | up to 95% of value | 6.89% | 8.68% | 9.23% |
| Bankwest | Fixed Rate Home Loan | up to 80% of value | 6.69% | 8.37% | 8.90% |
| Maitland Mutual Limited | Standard Home Loan | up to 95% of value | 6.54% | 8.02% | 8.65% |
| Newcastle Permanent Building Society | Fixed Rate Home Loan Principal and Interest | up to 80% of value | 6.49% | 7.94% | 8.52% |
| Newcastle Permanent Building Society | Fixed Rate Home Loan Principal and Interest | up to 80% of value | 6.49% | 7.94% | 8.52% |
| Newcastle Permanent Building Society | Fixed Rate Home Loan Principal and Interest | up to 80% of value | 6.49% | 7.66% | 8.52% |
| Bank of China | Fixed Rate Home Loan (With Principal and Interest Repayment) | any deposit | 6.39% | 7.98% | 8.40% |
| Westpac | Fixed Rate Home Loan | up to 70% of value | 6.74% | 8.26% | 8.74% |
| Westpac | Fixed Rate Home Loan | 80–95% of value | 7.14% | 8.65% | 9.14% |
How the reversion was recovered
Schedule 6 of the National Credit Code fixes how a comparison rate is calculated: a $150,000 loan amortised over 25 years. A fixed rate cannot fill 25 years, so the calculation has to assume what follows. Given the advertised rate, the term and the published comparison rate, that assumption is the only unknown, and it can be solved for.
Two unknowns actually sit in the blend — the reversion and the product's annual fee — and they leave opposite signatures across a lender's ladder of fixed terms. A fee's effect on the comparison rate shrinks as the term lengthens; the reversion's grows. So the fee that makes a lender's one to five year terms agree with each other on a single reversion is fitted alongside it, and the agreement between those terms is the error bar. Across this sample the terms agree to a median of 0.009 percentage points.
The check that matters is out of sample: the fit never sees variable rates, yet the numbers it produces land on rates the same lenders publish. Bank of China's investment ladder recovers 8.71% against their published 8.70% Standard Investment Variable Rate. Westpac's six owner-occupied lines recover 6.38, 6.49, 6.82, 8.74, 8.83 and 9.14 — against published variable rates of 6.39, 6.49, 6.84, 8.73, 8.83 and 9.13.
What this does not tell you
It is not a forecast. These are today's prices. A borrower fixing for three years today meets their lender's variable rate in 2029, not this one. What the figure tells you is which rate your contract points at — the standard variable or a discounted one — not what that rate will be when you get there.
It is not independent of the comparison rate. The reversion is derived from the comparison rate, so "the comparison rate flags high reverters" is close to a restatement rather than a discovery. The value is not the correlation; it is that the answer was already published and is not read this way.
Package fees are not a general rule. At Westpac and Bankwest the dearer packaged line reverts far lower, and it is tempting to conclude the fee buys the discount. Across every lender offering lines at materially different fees in the same deposit band, the dearer line reverts lower only about a third of the time. Do not assume the expensive one is protecting you.
Lenders do not all revert to a standard variable rate. Some contracts point at a discounted variable rate applicable at the time. Where a fitted figure sits well above anything the lender currently sells, read it as a contractual ceiling rather than the rate you would meet — and lenders reprice on request, which no published data can capture.
Lines that cannot be separated are excluded, not guessed. Fitting needs three or more terms on one rate line. Where a lender publishes several lines in a deposit band that cannot be told apart from published fields, they are left out.
General information about published rates, not financial or credit advice, and not a recommendation of any product or lender. Figures are inferred from lenders' published data and may be wrong; confirm anything that matters with the lender or your own adviser before acting.