RDP 2021-07: Macroprudential Limits on Mortgage Products: The Australian Experience Non-technical summary for ‘Macroprudential Limits on Mortgage Products: The Australian Experience’

The Australian Prudential Regulation Authority (APRA) implemented macroprudential policies targeting the housing market twice between 2014 and 2018. These policies were aimed at risks that were building in speculative components of the mortgage market, which, at the time, were seen to have the potential to amplify housing market dynamics and economic fluctuations. The first policy, announced in late 2014, required banks to limit their annual growth in investor mortgages to no more than 10 per cent. The second policy, announced in early 2017, required banks to limit their lending in interest-only mortgages to no more than 30 per cent of their total new housing lending. APRA removed the limits when they felt there had been sufficient improvement in lending standards and the identified risks had subsided.

Our paper focuses on 2 questions: how successful were these policies in curbing the targeted types of lending? And what other effects did these policies have? To answer these questions, we analyse lending and interest rate data from the 28 largest Australian banks, which account for the vast majority of outstanding housing credit.

As seen in Figures 1 and 2 below, both policies quickly reduced aggregate growth in the types of lending that APRA was targeting. To meet APRA's limits, banks cut the growth of new lending of the targeted loan types by around 20 to 40 percentage points within a year, having raised interest rate spreads on such mortgages by around 10 to 30 basis points. In our paper, we delve into a precise and robust measurement of the policy impact and, particularly, a deeper understanding of how the effects varied across banks.

Large and mid-sized lenders reacted differently to APRA's policies. Large banks (that is, the 4 major banks) transitioned their lending into mortgage types not targeted by the policies. For example, when financial institutions cut interest-only lending, large banks increased their principal and interest lending, while midsized banks did not. In addition, the policies had some effect on competition among the 28 banks we analyse, though the impact was short lived. While the policies did not slow aggregate mortgage growth, they slowed growth in the targeted mortgage types. Moreover, we notice signs of a positive effect on business credit.

The paper also explores implementation issues associated with these macroprudential policies. Because of their novelty in Australia, such policies brought unique challenges for banks and regulators. As a result, we observe a lag of 2 quarters between when the investor policy was announced in 2014 and when it began to have a significant effect. The delay was due to different interpretations of the credit growth limits and because banks' systems were not initially capable of targeting particular components of their mortgage portfolios. These challenges had largely been overcome by the time of the interest-only policy, when the policy had an immediate effect.

The broad aim of macroprudential policy is to manage systemic risk in the financial sector. Our results show that, despite some initial difficulties and unexpected effects, banks reacted by reducing growth in risky types of lending targeted by the regulator. As such, our results suggest that these macroprudential policies achieved their stated aims and contributed to a reduction in risk in the financial system.

Back to abstract