Australian economic view – March 2026

Market review

A hiking Reserve Bank of Australia (RBA), balanced against a global risk off environment, resulted in the Australian bond market, as measured by the Bloomberg AusBond Composite 0+ Yr Index, rising 0.88%.

The RBA kick started the hiking cycle in February, increasing the cash rate 25 basis points (bps) to 3.85%. Three‑month Bank Bill Swap Rates (BBSW) rose 15bps to 3.99% and six‑month BBSW 24bps to 4.33%. The three‑year government bond yield eased 5bps to 4.22% and ten‑year yields declined 16bps to 4.65%, resulting in a flatter domestic yield curve.

The global environment outweighed the RBA’s hawkishness as a myriad of intersecting concerns materialised. Broad questions are being asked of the implications of artificial intelligence (AI) for a range of economic sectors and broad labour markets. This has also led to questions relating to credit markets, private credit and software profitability. Time will tell with these themes, but it weighed on sentiment and generated uncertainty. Global economic activity was generally stable, with market focus elsewhere.

Australia isn’t immune from these themes, but the local dataflow has been more instrumental. Headline CPI rose 3.8%, unchanged from the prior month, while trimmed mean CPI was 3.4%, up from 3.3%. This was higher than expected, and increases were broader based than in prior months. The RBA are concerned about the persistence of inflation, so this was an unwelcome outcome. Labour market conditions remained firm, with the unemployment rate at 4.1% in January, with supportive inputs. The NAB Business Survey showed business conditions moderating slightly. Consumer sentiment weakened following the RBA move, with the Westpac–Melbourne Institute Consumer Sentiment Index falling to 90.5 in February from 92.9 in January, household spending was also weaker in the month.

Under these conditions, the RBA remain highly data dependent, and the advent of Middle East tensions at month end are likely to increase policy path uncertainty. The domestic economy shows signs of general resilience, but the global backdrop has the potential for disruption.

Market outlook

We see the RBA continuing to raise interest rates, to above what the market currently anticipates over the life of the current cycle. Our high case is one where inflation remains elevated and the RBA are forced to raise interest rates more than expected into 2027. Our low case reflects a weaker economic outcome, if global uncertainties are renewed and the labour market deteriorates. We hold no tilt at present. We hold a small long duration position, targeted on the curve, and remain vigilant to take advantage of market mispricing.

Monthly focus – AI productivity in the Australian context

Australia has always embraced general purpose technology (GPT) and adapted it to suit our unique economy. The AI transformation is expected to be no different. Using the Organisation for Economic Co-operation and Development (OECD) study of productivity gains from AI as a guide, we assume that AI will raise productivity levels in the Australian economy over the coming decade, as technology augments employees across the sectors Australia has a comparative advantage in. Rising productivity, given it is a significant input, leads to a higher neutral interest rate in the medium term.

The OECD examined the productivity benefits of AI¹ across the G7 economies. Their process applied a number of increasingly common methodologies studying this topic and pulled them together. First, they note the impact on specific tasks, then the exposure of tasks in industry sectors, finally the rate of adoption for the economy. If we put this in an Australian context, the industries in which AI is likely to augment work feature highly in the economy². Jobs and Skills Australia note that only 4% of the workforce have full automation exposure, while 96% have medium to high AI augmentation, where AI can be used to support and enhance worker output.

AI augmentation is expected to be high in significant industry sectors of the Australian economy. Knowledge based services are heavily positively impacted, such as the IT sectors, finance, professionals and telecoms. Less impacted are agriculture, mining and construction. Although, there are examples of increasing use in these sectors as well. This also raises the point that there are use cases and new roles that we haven’t even envisioned as yet. Australia’s GDP is weighted toward the knowledge-based activities, except for mining. If we cross reference the OECD’s top 25% of industries benefiting from AI, Australia’s production in these sectors accounts for around 34% of GDP. Given this, we can assume that the benefits from AI enhancing productivity comes at the upper end of estimates.

The final part of the productivity puzzle is when can we expect the benefits to flow through. Australia has an interesting history when it comes to adapting to new technology. We often hear that the AI benefits for the Australia economy are “years” away, thus assuming we are a slow adopter of new technology and benefit less from the creation of new technology. However, Fleming et al³ note in a pre-AI study, that Australia tends to adapt new technology to suit the local specialisations and adopt new technology fully in time. Australia imports new patents in broad industries and then specialises in what Australia does well. This seems like an appropriate template for AI adoption. We might not create the underlying technology, but are likely to embrace it, and in specialised, comparative advantage sectors, embrace and enhance its use. As the 2025 Nobel prize for economics theory would suggest, it isn’t just the innovation itself but the willingness to use these in practical applications which drives sustained economic growth⁴.

Given this, we would assume, in the OECD phraseology and methodology, that Australia is most likely to follow a more medium pace of adoption. This assumes that firms follow an s-shaped profile, raising their adoption of AI technology to around 50% of firms by 2034. Their 10-year horizon started in 2022, on the official start of the technology and uses the latest digital technology adoption as a guideline. The build up tends to be a slow rise, with a more rapid profile from 2030 onwards. This is, of course, highly uncertain and many assumptions underpin the academic research.

In turn, given the exposure to AI tasks and assumed adoption rate, the OECD estimates applied to the Australian environment would suggest an additional 0.5 – 0.8 percentage point (ppt) increase in annual labour productivity. This is above the estimates from the Australian Productivity Commission (0.4ppts), and below the OECD estimates for the United States. We would assume large uncertainty bands around the assumption.

^{Views as at 2 March 2026.}