Statement on Monetary Policy – November 2025Technical Note: The Transition to a Complete Monthly CPI

Key takeaways

• The Australian Bureau of Statistics (ABS) will start publishing a complete monthly Consumer Price Index (CPI) in late November 2025. The complete monthly CPI (referred to hereafter as the ‘monthly CPI’) will be the benchmark variable for Australia’s inflation target.
• The RBA welcomes the introduction of the monthly CPI. It will provide more timely information on consumer price inflation than the current quarterly CPI, assisting with monitoring inflation in the economy. The complete monthly CPI is also more comprehensive than the ‘monthly CPI Indicator’ (which has now been discontinued), as it measures monthly price changes for a wider range of household goods and services.
• The ABS will continue to publish quarterly seasonally adjusted CPI data based on the pre-October 2025 collection frequency (referred to hereafter as the ‘quarterly CPI’) for at least 18 months, while the complete monthly seasonal patterns are established. The ABS will also publish select measures of underlying inflation in the monthly CPI. Initially, for price categories with a relatively short data series available, the ABS will not be able to apply standard seasonal adjustment techniques and will use alternative methods to adjust data for seasonality.
• During this transition, the RBA will continue to focus on measures of underlying inflation from the quarterly CPI, such as the quarterly trimmed mean series. These data, which have well-understood properties and established quarterly seasonal patterns, will provide an important source of continuity while the new monthly time series mature. The RBA will also consider underlying inflation measures constructed using the monthly CPI. As the seasonal properties of the monthly data become clearer over time, the RBA will transition away from using measures of underlying inflation from the quarterly CPI and focus on the monthly data when assessing inflationary pressures.
• The RBA will continue to produce and publish quarterly (rather than monthly) forecasts for inflation. This is because some of the key variables that the RBA uses to forecast the economic outlook, such as GDP and labour costs, are only published on a quarterly basis. For headline CPI inflation, we will publish forecasts for the year-ended change of the quarter-average of the monthly CPI (non-seasonally adjusted). Many other central banks with monthly inflation data also forecast a quarter-average inflation rate. For underlying inflation, initially we will continue to publish forecasts for the year-ended change in quarterly trimmed mean inflation from the quarterly CPI, noting that the near-term forecast will be informed by the monthly CPI data. Over time, the RBA will transition to forecasting a measure of underlying inflation that is constructed from the monthly CPI; as we accumulate a longer history of data, we will better understand the properties of these measures.
• This technical note details how the staff will transition from using quarterly data to monthly data to monitor and forecast inflation. The changes are summarised in Table 1.

How are Australia’s inflation data changing?

The monthly CPI will provide much more timely information on inflation than existing measures.

The monthly CPI will commence with the release of the October 2025 data on 26 November, replacing the quarterly CPI as Australia’s primary measure of headline consumer price inflation. The aggregate monthly CPI series will be published back to April 2024. The ABS collects prices for thousands of items consumed by households, which are grouped into 87 Expenditure Classes. In April 2024, the ABS expanded its collection of monthly data for many more Expenditure Classes, including for most consumer durables (e.g. motor vehicles, furniture, appliances and clothing), market services (e.g. restaurant meals, takeaway and fast foods, hairdressing and insurance) and administered services (e.g. child care and pharmaceuticals) (Graph 1).

In addition to increasing collection frequency, the ABS has incorporated new data sources – including online price web scraping, transaction records and government-sourced administrative datasets. Note that monthly price collections had commenced before 2024 for around half of the Expenditure Classes, including housing, holiday travel and fuel, and were the basis for the monthly CPI Indicator, which has now been discontinued ahead of the introduction of the monthly CPI.¹ These data were previously reported in the partial monthly CPI Indicator on a non-seasonally adjusted basis, but will be published on a monthly seasonally adjusted basis for the first time in the monthly CPI.²

The introduction of the monthly CPI is a significant milestone that brings Australia in line with international best practice. It will provide more timely information on consumer price inflation than the existing quarterly CPI data. This will help us to analyse inflation in the economy better – e.g. higher frequency data may make it easier to identify any changes in inflation trends.

What CPI data will the RBA use to assess headline inflation?

The monthly CPI will be the benchmark variable for Australia’s inflation target.

Australia’s flexible inflation target is for consumer price inflation to be between 2 and 3 per cent.³ The Monetary Policy Board sets monetary policy so that inflation is expected to return to the midpoint of the target range within an appropriate period of time and remain there on an ongoing basis.

Historically, the target variable has been specified as year-ended headline inflation from the quarterly CPI, which is the broadest measure of prices in the CPI. In assessing inflation, the staff have closely monitored headline and underlying inflation measures from the quarterly CPI (see below). Throughout each quarter, RBA staff have updated their assessment of current inflation using data on Expenditure Classes from the monthly CPI Indicator.

From November, the monthly CPI will be the benchmark variable for the inflation target, benefiting from improved data coverage and timeliness. Of course, monthly observations may still be volatile, even in year-ended terms, so in judging underlying inflation the RBA will continue to have recourse to a range of other measures, as discussed in the next section.

What CPI data will the RBA use to assess underlying inflation?

The RBA currently focuses on trimmed mean inflation to assess the degree of underlying price pressures in the economy.

While Australia’s inflation target is expressed in terms of headline CPI, the staff also closely monitor measures of underlying inflation. These measures help us look through volatility in prices and through the effects of one-off or temporary measures, to get a better sense of the degree of underlying price pressures in the economy.⁴

The RBA currently considers a number of underlying inflation measures constructed from the quarterly CPI. These include ‘trim-based’ measures that exclude items with large price changes in a given period, regardless of whether the item is typically volatile or not. Trimmed mean inflation is the average rate of inflation after ‘trimming’ away the items with the largest price changes (positive and negative) in seasonally adjusted terms.⁵ We also consider ‘exclusion-based’ underlying inflation measures, which exclude items that often exhibit large price changes, such as fruit, vegetables and fuel.

A good measure of underlying inflation should satisfy several criteria, including that it is unbiased with respect to headline inflation, helps to predict near-term headline inflation, and tracks a trend measure of headline inflation over time.⁶ Over the past decade or so, the RBA has placed most emphasis on quarterly trimmed mean inflation when assessing underlying inflation, reflecting that it has performed well against these desired characteristics.⁷ Between quarterly CPI releases, we have drawn on component-level data from the monthly CPI Indicator for our assessment of trimmed mean inflation for the current quarter, but have generally not placed emphasis on aggregate measures of underlying inflation derived from the monthly Indicator data. For example, the monthly CPI Indicator trimmed mean, based on the year-ended distribution, is typically more volatile and backward looking than the quarterly trimmed mean series.

The short history of monthly data for some items means that the ABS will not be able to apply standard seasonal adjustment techniques initially.

For some items, initially the ABS will use alternative techniques to account for seasonal patterns in monthly price changes, so there may be some residual seasonality for a time. Standard seasonal adjustment techniques for monthly data require at least three years of historical data. As a result, for Expenditure Classes with monthly time series beginning in 2024 (46 per cent of the basket by weight) the ABS will instead be producing seasonally adjusted data using non-standard approaches. Specifically:

• For 15 per cent of the basket, the ABS will use historical data to first ‘backcast’ a longer monthly time series that can then be seasonally adjusted.
• For 10 per cent of the basket, when price changes are identified as being largely driven by seasonality, the ABS will smooth the non-seasonally adjusted data.⁸
• For 21 per cent of the basket, the ABS expects that the data will have either low or no identifiable monthly seasonality, based on its knowledge of the series and the absence of a seasonal pattern in the quarterly data.

This implies that there may be some residual seasonality in the seasonally adjusted monthly CPI for a time. This will limit our ability to rely solely on the monthly CPI data to gauge persistent inflationary pressures. The use of ‘smoothing’ techniques for a small share of items may mask some signal in non-seasonal price movements. There may also be relatively large revisions to historical seasonally adjusted data from month to month, as well as further revisions when standard seasonal adjustment techniques are able to be applied to the remaining Expenditure Classes from April 2027.

Residual seasonality may affect some measures of underlying inflation calculated using the monthly CPI data. For example, trimmed mean and weighted mean inflation are constructed using seasonally adjusted data. So even year-ended measures of trimmed mean inflation rely on seasonally adjusted data and so may be impacted by residual seasonality. This is in contrast to year-ended measures of headline inflation that rely solely on non-seasonally adjusted data.

Some component-level data may also be impacted by residual seasonality. This is particularly the case for some consumer durables prices, and to a lesser extent market services prices,⁹ where ‘smoothing’ techniques will be used in some months until more data become available.For items that already have a longer monthly time series, such as housing and groceries, or which display little seasonal variation in prices, the monthly CPI data will provide a timely read on inflation from the outset.

Once more data become available, standard techniques will be applied. The likelihood of there being residual seasonality will be substantially reduced once all Expenditure Classes can be seasonally adjusted from April 2027.

The continued publication of the seasonally adjusted quarterly CPI data for at least 18 months will ensure we retain access to data with well-understood properties, while also making use of the new and more timely monthly data.

The ABS will continue to publish quarterly seasonally adjusted data for many series for at least the next 18 months, at which point standard seasonal adjustment approaches will be able to be used for all Expenditure Classes in the monthly data. The collection frequency of the prices data underpinning the seasonally adjusted quarterly CPI data will remain unchanged (i.e. it will continue to use the same timing of price collection as it did before October 2025).

The seasonally adjusted monthly and quarterly CPI data will each provide important information on inflation developments. Corresponding data series from the monthly CPI and quarterly CPI may differ, including because of differences in seasonal factors and collection frequencies between the data sets. The monthly CPI will be timelier and have greater coverage than the quarterly CPI, but seasonally adjusted data from the monthly CPI are likely to have more residual seasonality until mid-2027 and some non-seasonal movements may be obscured by smoothing techniques. While the quarterly CPI is less timely, it will provide continuity as it has a long history with established seasonal patterns and well-understood properties.

The RBA’s use of the monthly CPI for monitoring and forecasting inflation will evolve gradually as we learn about its properties, and as a longer history allows seasonal patterns to be better understood and accounted for.

The RBA will continue to focus on trimmed mean inflation from the quarterly CPI to discern underlying inflation, until there is enough data to adequately assess the properties of underlying inflation measures constructed from the monthly CPI.

Following the commencement of the monthly CPI, for a time the RBA will continue to focus on trimmed mean inflation from the quarterly CPI when assessing underlying inflation. This is because quarterly trimmed mean inflation has well-understood properties, is based on seasonal factors calculated using standard techniques and has a long time series. Throughout the quarter, the RBA will use both aggregate and component-level data from the monthly CPI to help predict the quarterly trimmed mean rate in the near-term.

The monthly CPI will contain valuable additional information, and we will consider measures of underlying inflation based on the monthly CPI, complementing the quarterly trimmed mean series. For example, the ABS will publish monthly trimmed mean and monthly weighted median inflation – based on the monthly distribution of seasonally adjusted price changes. The RBA will also calculate other underlying inflation measures that smooth the monthly CPI data and reduce volatility. Examples include smoothed measures of monthly trimmed mean inflation, or trimmed mean inflation calculated using price changes over longer time periods.¹⁰ As more monthly data become available, we will assess the relative strengths and weaknesses of the various measures of underlying inflation, including against the desirable criteria of underlying inflation outlined above.

Underlying inflation measures from the monthly and quarterly CPI will differ, just as alternative measures of underlying inflation from the existing quarterly CPI vary. This is expected and does not reflect problems with either measure – simply that they are all constructed differently and therefore measure different things. For instance, the monthly trimmed mean uses more frequent, but also more volatile, price changes than the quarterly trimmed mean.¹¹

Over time, the RBA will transition to focus on underlying inflation measures based on the monthly CPI.

Given that the monthly CPI will initially have a short time series, the properties of different underlying inflation measures constructed from it will not become fully apparent for some time. We will continue to monitor and analyse these measures over time to understand how they can best inform our view of underlying inflation. This includes assessing how these measures perform against the desirable criteria of underlying inflation outlined above, and evaluating their performance in our forecasting frameworks. Based on this analysis, the RBA may continue to focus on a trimmed mean measure of inflation, focus on a different measure of underlying inflation or emphasise a suite of complementary measures.

What inflation series will the RBA forecast?

Historically, the RBA has published quarterly forecasts for headline inflation and trimmed mean inflation.

The RBA uses the quarterly CPI data to produce forecasts for headline inflation, which is the benchmark inflation target variable, as well as for underlying inflation, which is less volatile and helps assess the underlying trend in inflation. Most of the RBA’s economic models are specified in terms of underlying inflation.

Following the publication of the monthly CPI, the RBA will continue to produce and publish quarterly forecasts for headline and underlying inflation.

Our continued focus on quarterly outcomes aligns with our forecasts for other economic variables, including many series that are available only on a quarterly frequency, including GDP. This is the same approach used for other monthly data such as the unemployment rate. Many other central banks with monthly inflation data also forecast a quarter-average inflation rate.

For headline inflation, the RBA will now forecast a quarterly measure constructed from the monthly CPI. Specifically, the RBA will forecast the year-ended change in the quarter-average of the headline price index (non-seasonally adjusted) from the monthly CPI.¹²

For underlying inflation, we will continue to forecast trimmed mean inflation from the quarterly data until there is enough data to assess underlying measures from the monthly CPI. Initially we will continue to forecast the year-ended change in quarterly trimmed mean inflation. This series uses the quarterly CPI collection timing and will be published by the ABS in Appendix 1a of the monthly CPI. This is consistent with how we plan to monitor the underlying inflation data and allows our forecasting models to be estimated on the longer time series available for the quarterly CPI.

Over time, the RBA will transition to forecasting an underlying inflation measure that is derived from the monthly CPI.

This transition will be assisted by learning about the properties of the monthly data in the period ahead and the shift to standard seasonal adjustment for all Expenditure Classes as more monthly data become available. The choice of which underlying measure to forecast will be informed by how they perform against the desired characteristics of underlying inflation outlined above and in our modelling frameworks.

Endnotes

1 For more detail, see ABS (2025), ‘Monthly Consumer Price Index Indicator Methodology’, August. The RBA’s preliminary analysis of the Monthly CPI Indicator was laid out in RBA (2022), ‘Box B: The Introduction of a Monthly CPI Indicator for Australia’, Statement on Monetary Policy, November. However, since then, the ABS have made several changes to the series, including changing its calculation of trimmed mean inflation.

2 A small share of items will continue to be collected on an annual or quarterly basis, including education fees and private health insurance, typically because these items change price less frequently (Graph 1).

3 See the Treasurer and the Monetary Policy Board (2025), Statement on the Conduct of Monetary Policy, July.

4 See RBA (2024), ‘Box C: Headline and Underlying Inflation’, Statement on Monetary Policy, August.

5 That is, after ranking each item in the basket by the size of seasonally adjusted price changes, trimmed mean inflation is calculated by the weighted mean of the middle 70 per cent of the quarterly distribution of price changes. Relatedly, weighted median inflation is the inflation rate of the item at the middle of the price changes in the CPI basket (the 50th percentile of the quarterly distribution by weight). See RBA, n 4; RBA, ‘Inflation and its Measurement’, Explainer.

6 Other desirable properties of underlying inflation include that it is subject to minimal revisions, can easily be replicated by others, and is easy to understand. For a more comprehensive discussion of the desirable properties for underlying inflation, see Brischetto A and A Richards (2007), ‘The Performance of Trimmed Mean Measures of Underlying Inflation’, RBA Research Discussion Paper No 2006-10.

7 See Brischetto and Richards, n 6; Richards AJ (2024), ‘The Use of Trimmed-Mean Measures of Inflation in Monetary Policy Making: Some Differences between Old and New World Central Banks’, September.

8 For a description of the ‘smoothing’ approach, see ABS (2025), ‘Seasonal Adjustment for the New Complete Monthly Consumer Price Index (CPI)’.

9 The ABS has indicated that it may use smoothing techniques in some months for around 46 per cent of consumer durables and 8 per cent of market services. For more information, see ABS, n 8.

10 Examples of these different kinds of smoothed measures are three-month-on-three-month growth in the monthly trimmed mean index and trimmed mean calculated using the distribution of three-month-on-three-month inflation rates respectively. The latter measure could be constructed by calculating the three-month-on-three-month growth rate in the seasonally adjusted index for each Expenditure Class. Neither of these measures would require seasonal adjustment to be performed by users, meaning they can be replicated more easily.

11 Analysis of other series of trimmed mean inflation suggests there can be persistent differences between measures when they are constructed using short-horizon and longer horizon price changes. For example, in Australia, we currently observe differences between the quarterly trimmed mean and the annual trimmed mean (i.e. trimmed mean inflation for the annual distribution of price changes).

12 This series will be published by the ABS in Table 17.