Serviceability is the bank's answer to a simple question: if interest rates rose, could you still pay the loan while living a life that looks realistic on paper? It is separate from the headline repayment on a free calculator, and it is often the constraint that binds before a regulatory DTI cap does.
Assessment rates sit above carded rates
Banks apply a buffer — often roughly two to two-and-a-half percentage points above a reference rate — when they test repayments. A loan that looks comfortable at 4.9% may be assessed as if you were paying something nearer the mid-to-high sixes. That is why brokers talk about “servicing” as its own hurdle.
Living costs are assumed, not wished away
Banks use living-cost models that rise with dependants and sometimes with income. Declaring that you “hardly spend anything” rarely overrides their floor. Accurate expenses help; fantasy budgets do not.
Other debts reduce surplus
Card limits, personal loans and student loan repayments all reduce the surplus available for a mortgage under the test. This is the same family of issues as DTI, but expressed through monthly cashflow rather than a stock of debt alone.
The student loan case is a good illustration of why the two tests differ. Repayments are deducted at 12% of earnings above $24,128, so the loan reduces your monthly surplus every pay cycle regardless of how large the remaining balance is. A borrower with a small balance and a high salary still loses that slice of cashflow until the loan clears. Serviceability sees the deduction; DTI sees the balance. They are measuring different things and can disagree about which applicant is the safer bet.
How income is graded, not just counted
Not all income is treated as equally reliable. Base salary from permanent employment is generally taken at face value. Overtime, commission and bonuses are often shaded — a lender may count only a portion, and usually wants a track record before counting them at all. Self-employed income is typically assessed on two years of financial statements, with the lower or an average of the two years used rather than the better one. Contract income sits somewhere in between and depends heavily on how long the contract has left. Two applicants with identical gross figures on paper can therefore be assessed very differently.
Why your surplus is not the same as your savings rate
Borrowers often argue that they have demonstrably saved a deposit while renting, which proves they can afford the repayments. It is a fair argument and lenders do give weight to a genuine savings record. But the test is forward-looking and deliberately conservative: it assumes a rate well above the one on offer, applies a living-cost floor whether or not you live that way, and does not assume your rent disappears cleanly into a mortgage payment, because rates, insurance and maintenance arrive with the house. A strong savings history helps at the margin. It does not override the model.
DTI versus servicing
DTI caps (around 6× for many owner-occupier cases, with limited lending above the threshold) constrain total debt against income. Servicing asks whether the repayments fit. You must clear both the bank's internal tests and the regulatory environment they lend under. Whichever is tighter wins.
What to do with FinLab's range
Treat the borrowing range as a planning band, then stress your mortgage repayment at a higher rate in the mortgage calculator. If the higher-rate repayment breaks your budget, the bank's serviceability test is trying to tell you the same thing before you apply.
It also pays to know which constraint is binding on you specifically, because the fix is different in each case. If servicing is the limit, the levers are cashflow ones: clear a personal loan, reduce a card limit, or wait for a pay rise to land and season. If the DTI cap is the limit, extra income helps but a larger deposit does not, because the cap compares total debt to income rather than to the property value. If LVR is the limit, the deposit is the only thing that moves it. Applicants often spend months on the wrong lever because nobody told them which test they actually failed — so ask, and ask specifically.