How to read major-bank cash rate forecasts without taking them as truth

The four major banks each employ a chief economist whose cash-rate forecasts make headlines roughly once a month. Those forecasts are real macro research, with proper modelling behind them, and they are also a marketing artefact for institutions that price mortgages and write business loans against the same rate path they are publishing views on. Both things are true at the same time. A reader who treats the forecasts as oracle is guessing; a reader who ignores them entirely is missing information the swap curve already weighs. The skill is somewhere in between: read the consensus, read the spread, and compare both to where the market is priced.

CommBank, NAB Group Economics, Westpac Economics, and ANZ Research each publish quarterly forecast packs and ad-hoc updates after every RBA meeting and every major data print. The four publications together are the bank-economist consensus, and they are the most frequently cited reference in property and finance coverage in this country. Knowing how to read them without being captured by them is worth the half-hour it takes to learn the shape.

Why the four banks publish forecasts at all

At the surface, the four publish forecasts because their clients ask for them. Mortgage holders, businesses, fund managers, and treasurers all want a number to plan against. Below that, there are two less-discussed motives. The first is positioning: a bank whose economist is quoted in the financial press every week becomes the default authoritative voice when a household is choosing a lender. The second is loan-book pricing: when the bank's own published view is that rates are heading lower, retail customers are more receptive to variable products, and the bank's own funding desk has a reference point its sales teams can lean on.

None of that makes the forecasts dishonest. The modelling is real; the economists are typically the most macro-literate people in the building. It does mean the published number lives downstream of an institutional process, and the public version is the version after editorial and stakeholder review. Read accordingly.

The four houses and their tendencies

The reputations shift with personnel, but over the last decade or so the four banks have developed visible tendencies in their published forecasts.

• CommBank tends to sit on the dovish side of the consensus, forecasting cuts earlier and more aggressively than the others. The retail-mortgage franchise is the largest in the country, so the house view is read closely by owner-occupier households.
• Westpac Economicshas historically been the most hawkish of the four. The Bill Evans era set that tone for decades, and the house view often diverges from the swap curve's pricing, sometimes by more than 50 basis points at the twelve-month horizon.
• NAB Group Economics sits closer to the middle of the range, with a stronger business-banking and SME lens. The forecasts tend to weigh corporate credit conditions more than owner-occupier mortgage stress.
• ANZ Research also sits centrally but brings a broader Asia-Pacific frame, with more attention to global cycles, China, and trans-Tasman conditions than the others.

These are tendencies, not constants. When a chief economist changes, the house view can move with them. Treat the labels as a starting point for reading any given publication, not as a permanent guide.

The forecast horizons and what each is worth

Bank publications generally include four horizons, in descending order of usefulness.

• Next-meeting probability: cut, hold, or raise. This is the highest-confidence call the economists make, because the data window is short and the signal from the previous communiqué is fresh. Direction calls here are right around 70-80% of the time within three months on the published research record.
• End-quarter forecast: where the cash rate sits at the end of the current or next quarter. Useful for fixed-vs-variable timing, less so for long-horizon planning.
• End-year forecast: where the cash rate sits at the end of the calendar year. The error band here widens fast. By twelve months out, the absolute level forecast is typically wrong by 50 basis points or more across most published cycles.
• Terminal-rate estimate: where the cycle bottoms out or peaks. This is the most interesting number for thesis-building and the least accurate for planning. Treat it as a directional anchor, not a level.

Market-implied path versus bank consensus

The Overnight Indexed Swap market prices a forward cash-rate path every trading day. The OIS curve is the market's consensus, money on the table, and it updates in seconds where bank economists update in weeks. Comparing the two is one of the most useful habits a reader of these forecasts can build.

When the OIS curve is more dovish than the bank consensus, the market is pricing a weaker economy than bank economists are. When the curve is more hawkish, the market is pricing inflation persistence the economists have not yet priced in. Neither side is right by default. The historical record is that both miss turning points regularly. Divergence is information about where the next surprise might land, not a signal of which side to trade.

How to read the consensus in practice

Four steps, in order.

• Take the median of the four banks at a given horizon. That is the central tendency. Mean is biased by an outlier; median is sturdier.
• Read the spread. A 60 basis-point gap between the most dovish and most hawkish house is a market with high uncertainty. A 15 basis-point gap is a market where the four houses are converging on the same path.
• Compare to the previous publication. A sudden 50 basis-point shift in any house view means something has happened: a data print, a communiqué revision, a global event. The change is the signal more often than the level.
• Cross-check the OIS curve. If the bank median lines up with the curve, the consensus is well-priced. If it diverges materially, the divergence is where the next move probably lives.

Common reading errors

Four traps catch most casual readers.

• Treating a forecast as a guarantee. The economists themselves publish confidence intervals; the press cycle strips them out.
• Anchoring on whichever forecast was published most recently, regardless of whether the underlying data has moved since.
• Ignoring the spread. A consensus of 3.85% with the four houses ranging from 3.50% to 4.25% is barely a consensus at all.
• Acting on the twelve-month or terminal-rate forecast as if it were a planning number. At those horizons, the error bars typically dominate the central estimate.

What this means for property buyers

The use case depends on the decision in front of you.

• Buying and choosing fixed vs variable: the three-year fixed rate already prices in the consensus path. If the consensus and the OIS curve both suggest cuts, the fixed rate is usually below the current variable. If the published fixed rate is above the current variable, the market is pricing rate stability or rises.
• Refinancing: the consensus path is informative for timing. If the median view is cuts within six months, refinancing into a variable product preserves optionality. If the view is hold or rise, locking in a competitive fixed rate has more to recommend it.
• Long-term investing: the forecasts are mostly noise. Cash-rate cycles run five to seven years; the next twelve months are a small fraction of a typical holding period. Structural trajectory, rental demand, and supply pipeline carry more weight than any single year of rate calls.

A worked illustrative example

Suppose the four-bank consensus puts the cash rate near 3.85% by Q4 2026, with the spread running from a dovish house around 3.60% to a hawkish house near 4.10%. The OIS market is pricing roughly 3.75%. The median is close to the curve, but a 50 basis-point spread between houses says the economists are not converged. Wider spread, wider uncertainty.

For a $600,000 owner-occupier variable loan at 6.20%, 30 years principal and interest, monthly repayments land near $3,679. A 25 basis-point cut, broadly in line with what most of the four houses are projecting over the coming year, takes the contracted rate to 5.95% and the repayment to roughly $3,584, saving about $95 a month or $1,140 over a year. A 50 basis-point cut roughly doubles that. Run the math against your own balance on the mortgage calculator; the scenario worth modelling is not the central forecast but the dovish and hawkish edges of the spread, because that is the range your repayment can actually land in.

Borrowing capacity moves on a similar arithmetic. The same 25 basis-point cut typically lifts the marginal household's ceiling by $15,000 to $20,000 once the APRA serviceability buffer is applied on top. The borrowing power calculator will translate the spread into a range rather than a point estimate.

What the banks don't forecast publicly

The published forecasts are about the RBA cash rate. They are not about the rates a bank actually offers its own customers. That second layer is set internally and is not disclosed.

• Internal mortgage book risk appetite, which influences how aggressively a bank prices for new business at any given cash rate.
• Lending-policy changes coming through the credit committee: tightening or loosening serviceability assumptions, debt-to-income caps, or specific collateral rules.
• Lending-volume targets by segment, which can shift discount margins on new originations independent of the cash-rate path.

The customer-facing rate is the cash rate plus a margin the bank controls and adjusts on its own timetable. A correct cash-rate forecast does not guarantee the customer rate will move in lockstep, and a bank's public dovish view does not guarantee its own variable product will reprice to match.

Reading the forecasts alongside the rest of the calendar

Bank cash-rate forecasts do not stand alone. They sit between the RBA's own communications and the data the RBA reads. The companion piece on reading the RBA Statement on Monetary Policy is the right starting point for understanding the documents the bank economists are themselves reading. For the demand pulse the economists weigh, ABS Lending Indicators is the cleanest monthly read. For the regulatory layer that sits between the cash rate and borrowing capacity, the APRA serviceability buffer is the missing piece.

On Burbfinder, suburb and region pages surface the rate-sensitive metrics alongside local prices, rents, and vacancy. The bank consensus is a national number, but the decision it informs is always a local one. The forecasts are a useful input when read as a spread; they become misleading when read as a single number on a headline.