Fixed or variable, when nobody knows what rates will do

A fixed rate holds your interest rate for an agreed period, commonly one to five years. A variable rate moves when the lender moves it. That is the whole of the technical difference, and it is not where the decision lives.

The decision lives in what fixing does to the rest of the loan. Fixing buys certainty of repayment and pays for it in flexibility: restricted extra repayments, limited or absent offset, and a cost to exit early that is calculated rather than capped.

It is tempting to treat this as a prediction problem. It is not one. Lenders price fixed rates off wholesale funding markets that already contain the market's collective expectation of where rates are going. When a fixed rate is offered below the variable rate, the market is not doing you a favour. It is pricing an expectation of cuts. Fixing does not let you beat that expectation. It lets you opt out of the consequences of being wrong about it.

Fixing is not a forecast. It is insurance. You buy it because of what you cannot afford, not because of what you expect.

• Extra repayments are usually capped. Many fixed loans allow a set amount per year above the required repayment, and break costs can apply beyond it.
• Offset is often unavailable, or available only partially. If your strategy depends on holding cash against the loan, fixing can undo it.
• Redraw on a fixed loan is commonly restricted or absent altogether.
• Refinancing during the fixed term triggers a break cost calculation.
• Selling the property during the term does the same, unless the loan is portable and the lender agrees to port it to the new security.

None of that is a reason not to fix. It is a reason to fix deliberately, with a clear view of what you intend to do with the loan over the next few years. A borrower who expects to sell within two years and fixes for five has bought certainty they will pay to hand back.

This is the most misunderstood number in Australian lending. A break cost is not a penalty and it is not a flat charge. It is the lender recovering an economic loss.

When you fix, the lender funds your loan by taking a matching position in wholesale markets at the rates prevailing on that day. If you exit early, that funding position still exists and has to be unwound at whatever rates prevail at the time. Where wholesale rates have fallen since you fixed, the unwind produces a loss, and the break cost passes that loss on to you.

Three things drive the size of it.

1. 01How far wholesale rates have moved since you fixed, and in which direction. Falling rates produce break costs. Rising rates generally do not.
2. 02How much time is left on the fixed term. Two years remaining costs far more than two months.
3. 03How large the balance is. The calculation applies across the amount being repaid early.

The consequence is counter-intuitive and worth sitting with. Break costs are largest in exactly the environment where you most want to refinance, because the same rate movement that makes a new deal attractive is the movement that makes your exit expensive.

A split loan is one loan divided into two accounts, one fixed and one variable, in whatever proportion you choose. It is the most under-used structure in Australian lending, and for a lot of borrowers it is simply the correct answer.

The fixed portion holds repayment certainty over the part of the debt that needs to be predictable. The variable portion keeps the offset, the unrestricted extra repayments, and the ability to make lump-sum reductions without triggering a break cost.

Two administrative notes. Splits usually mean two loan accounts and, at some lenders, two sets of fees. And the fixed portion reverts at the end of its term exactly as a whole fixed loan would, so the diary note still applies.

Between the day you apply and the day you settle, the fixed rate on offer can move. Most lenders offer a rate lock: for a fee, usually a percentage of the loan amount or a flat charge, the fixed rate available at application is held for a defined window.

Whether it is worth paying depends on how long settlement is likely to take and how much a movement would hurt. Without a lock, the rate you receive is the rate on the day of settlement or drawdown, not the day of application. Where a construction loan or a long settlement is involved, that gap can run to months. Ask what the lock costs, what window it covers, and whether the fee is refundable if the loan does not proceed.

Break cost

The lender's economic loss when a fixed loan is repaid early. It is calculated on the day, not set as a fee, and it can be nil where rates have risen.

Revert rate

The variable rate a fixed loan rolls onto when the term ends. It is set by the lender and is often not the sharpest rate that lender offers.

Rate lock

A fee paid at application to hold the fixed rate on offer for a defined window through to settlement or drawdown.

Portability

Moving an existing loan onto a new security instead of discharging it. It is conditional, and the lender has to agree to it.

Split loan

One debt held in two accounts, one fixed and one variable, in whatever proportion you choose.

You cannot forecast rates and neither can we. What you can do is answer four questions about yourself, which is a far more reliable basis for the decision than anyone's view of the next announcement.

1. 01

   What happens to your household if the repayment rises materially

   Not whether it would be annoying. Whether it would be survivable. If a meaningful rise would force a change you cannot make, certainty is worth paying for regardless of where you think rates are heading.

2. 02

   What do you intend to do with this property in the next three years

   Sell, renovate, refinance, rent it out. Any of those interacts badly with a long fixed term, and all of them are knowable now.

3. 03

   Do you have cash to deploy against the loan

   A regular surplus or an expected lump sum both argue for keeping at least a portion variable, so the offset and the extra repayments stay available to you.

4. 04

   How much of the loan actually needs to be certain

   This is the question that leads to a split, and it is usually the right question. The decision is rarely all or nothing, and treating it that way is what produces regret.

With those four answers in hand the product decision becomes close to mechanical. That is the point of framing it this way. The decision is about your circumstances, not about the next rate announcement.