Probability first
Where practical, model outputs will show the relative likelihood of different outcomes rather than a single “most likely” number without context.
RATE CHALLENGE FORECASTS & MODELS
Forecasts & Models
Evidence-led probability models and scenario analysis for Australian interest rates, home-loan pricing and the Victorian property market — built to show uncertainty, assumptions and alternative outcomes rather than present a single prediction as fact.
Probability-based
Scenario-led
Methodology disclosed
Performance tracked
This is the permanent Forecasts & Models gateway. Individual Rate Challenge models will be published only when their methodology, testing and review are complete.
Forecasts are uncertain by nature. Published model outputs will describe probabilities, ranges and scenarios — not guarantees of future outcomes.
How we approach forecasting
Useful forecasts should explain uncertainty, not hide it
A single number can create false precision. The Rate Challenge approach is designed around probability distributions, scenario ranges, changing inputs and transparent performance review so readers can see what may happen, what would need to change, and how confident the model actually is.
Scenario ranges
Base, stronger and weaker scenarios can help show how different economic, policy or market conditions may alter the path ahead.
Visible assumptions
Key inputs, assumptions, data periods and limitations should be visible enough for readers to understand what is driving an output.
Performance review
Where a model produces dated forecasts, its historical accuracy and calibration should be reviewed rather than quietly replacing old forecasts with new ones.
Planned model areas
The forecast areas this hub is designed to cover
These are the permanent model areas in the Rate Challenge research structure. A model page becomes active only when the underlying model and publication controls are ready.
Cash Rate Probability Model
Probability analysis around future Reserve Bank cash-rate decisions and the distribution of possible cash-rate outcomes over time.
Home-Loan Rate Forecasts
Analysis of possible movements in Australian variable and fixed mortgage pricing, including ranges and directional scenarios.
Property Price Forecasts
Quality-gated Victorian property-market scenarios, beginning at broader geographic levels before any more granular forecast coverage.
Rental Market Forecasts
Forward-looking scenarios for rental conditions, including rents, vacancy and yield-related market measures where suitable data supports them.
Housing Affordability Forecasts
Scenarios connecting property prices, deposits, mortgage rates and repayments to changes in housing affordability.
Lending Activity Forecasts
Possible future paths for lending activity such as approvals, refinancing and credit growth where sufficiently robust inputs are available.
Model accountability
A forecast should be judged against what happened next
Rate Challenge has a dedicated Model Performance structure so published forecasts can be assessed over time. The intention is to retain dated outputs, compare outcomes with forecasts and explain where models performed well or poorly.
- Historical forecast archive
- Forecast versus actual comparison
- Probability calibration review
- Methodology changes recorded
Core principle
Keep the old forecast visible.
Performance can only be evaluated if dated forecasts are preserved rather than overwritten.
Forecast output
Probability + range
Review
Forecast vs outcome
Publication standard
01
02
03
04
What a Rate Challenge model should show before you rely on it
Inputs
The economic, lending, rates or property-market information used to inform the model should be identified.
Method
The broad modelling approach, important assumptions and interpretation of outputs should be explained.
Uncertainty
Probabilities, ranges, scenarios or confidence information should make uncertainty visible rather than imply certainty.
Performance
Dated forecasts should be retained and evaluated against actual outcomes wherever a meaningful comparison is possible.
Research that informs the models
Explore the underlying Rate Challenge research
Forecasts & models FAQs
Understanding probabilistic forecasts
Does a high-probability outcome mean it will definitely happen?
No. A probability describes uncertainty; it is not a guarantee. Even the highest-probability outcome can fail to occur, particularly when new information changes market or economic conditions.
Why use ranges and scenarios instead of one forecast number?
Ranges and scenarios make uncertainty more visible. They can also show how an outlook changes when assumptions about rates, inflation, employment, lending, housing supply or other drivers change.
Will Rate Challenge publish old forecasts after the outcome is known?
The intended performance framework is to preserve dated forecasts so they can be compared with later outcomes rather than silently replacing them with updated forecasts.
Are these forecasts personal financial advice?
No. Forecasts and model outputs are general research. They do not take into account your objectives, financial situation, property, loan or personal circumstances.
Why are some model areas not yet linked?
The permanent research structure is being created before individual models are released. A model page will become active only after the relevant methodology, testing and publication checks are complete.
Research before prediction
Start with the current data before looking ahead.
Explore current home-loan rates, Victorian property research and government buyer support while the forecast library develops.