Variable, fixed, or split — and at what ratio?
Choosing between variable, fixed, or a mix of both feels like guesswork. It isn't. The right answer is shaped by how much you keep in your offset account: the variable portion benefits from offset every day, the fixed portion locks in the rate on the rest. This calculator picks the split that minimises your interest — the part bank tools skip.
Want full repayment math instead? Use the repayment + offset calculator →
Your headline result
—
How we got this
Working out the math…
The split decision, made boring
Most discussion of variable vs fixed treats the split decision as a question of risk tolerance and rate forecasting. It can be — but for most borrowers with an offset facility, the maths picks the answer before you ever weigh in on rate-direction views. Three observations that flatten the question:
- Offset only works against variable balance. Standard package fixed loans either disallow offset entirely or cap it at a token amount ($10k typical). So every dollar of offset you hold needs a corresponding dollar of variable balance to "absorb" it. Beyond your offset balance, variable is just an undiscounted version of the fixed rate.
- The optimum is offset-equal-to-variable. Up to your offset balance, variable is effectively 0% interest. Beyond your offset, every dollar of variable that could have been fixed costs you (variable_rate − fixed_rate) per year. The cleanest split is therefore: variable portion ≈ your sustained offset, fixed portion = the rest at the lower rate.
- The interesting input is offset, not split. Once the rule above is accepted, the only soft input is "what's my realistic sustained offset balance going to be?" — and that's where the honesty matters. Most borrowers overestimate offset by 30–50% because they look at peaks rather than averages.
The calculator runs your specific numbers, but the rule worth remembering: bigger offset → more variable. Smaller offset → more fixed. The "what split should I have?" question is mostly a "how big will my offset realistically be?" question.
What the rate sensitivity grid is for
The fixed portion is locked in at your quoted rate; the variable portion drifts with the cash rate. The sensitivity grid above shows what 5-year interest looks like at the optimal split if variable rates move ±1% / ±2% over the fixed period. Two things it surfaces that calculators usually hide:
- The asymmetry of rate moves at this split. Because the variable portion is small (≈ offset), rate moves on variable have a smaller absolute dollar impact than they would at all-variable. The split is itself a hedge.
- The cost of wrong-way rate views. If you fix because you think rates will rise and they fall, the grid quantifies the cost. If you stay variable thinking rates will fall and they rise, the grid quantifies that too. Both costs are usually smaller than the marketing-led narrative suggests.
What this calculator doesn't model (yet)
- Lender split fees. Setup fees of $0–$395 typical per split, sometimes per portion thereafter. Eats some of the headline savings; not a deal-killer for most loan sizes but worth checking against the savings figure above.
- Multi-portion fixed structures. Some borrowers split fixed into several tranches with different expiry dates — fixed-rate-laddering. v1 models a single fixed portion only.
- Refixing decisions at year T. What happens at the end of the fixed period (refix at then-current rates, refinance, leave on revert) is its own decision. v1 assumes the fixed portion reverts to variable rate at year T.
- Offset growth schedule. Real-world offset balances grow over time (you're saving). v1 uses a single sustained-average input. v2 will accept a growth schedule.
- Cap-on-extras for fixed loans. Most fixed loans cap extras at $10–30k/year. v1 applies extras to variable first, then fixed — which approximately reflects the constraint without modelling per-lender caps.
- Tax effects for investors. Investment loan interest is deductible; the after-tax math can shift the variable/fixed preference for high-bracket investors. We surface this on file.
Bottom line
The right split is more a question of arithmetic than judgment, once you fix on a realistic offset balance. The calculator does the arithmetic; you provide the offset estimate; the rest follows. The interesting conversation is whether the savings figure justifies the lender's split fee, and whether your offset estimate is honest — both of which are conversations worth having before settlement, not after.