Review of the Reserve Bank's loan-to-value ratio policy

A loan-to-value ratio (LVR) is a measure of how much a bank lends to a borrower, relative to the value of the borrower’s property secured against the lending. The more the banking system as a whole lends to high-LVR borrowers the greater the risk that the banking system will suffer large losses in a severe downturn. If banks do suffer large losses they may become unwilling to continue lending to the wider economy. Experience in other countries suggests this can have a damaging and lasting effect on the economy. The Reserve Bank of New Zealand (RBNZ) introduced LVR restrictions in October 2013 in response to financial stability risks associated with a potential house price correction and high-LVR mortgage lending, and has adjusted policy settings in response to changing risks. The LVR policy is aimed at constraining excessive house price inflation and household credit growth to lean against the risks of a severe correction. In doing so, it can support both financial system stability and sustainable economic growth. The Reserve Bank had considered, but decided against, capital macroprudential tools instead of the LVRs, although baseline capital rules for high-LVR loans were tightened. The LVR restrictions tend to have a greater impact in directly reducing housing and household sector risks, and in mitigating the scale of an economic downturn, than capital-based macroprudential tools that are focused on building additional bank capital buffers for absorbing shocks. This review traces changes in the LVR policy over the past five years, analyses the effect they have had on banks and households, and asks what the Reserve Bank can learn from this experience. The review suggests that the LVR policy has been effective in improving financial stability. By mitigating the scale of house price falls during a potential downturn, and limiting the indebtedness of households, the policy has made the financial system more resilient to a housing-led downturn. Declining risk weights for housing loans have offset some of the resilience benefit of LVRs, although the Reserve Bank has adjusted baseline housing capital calibrations to stabilise risk weights and support bank resilience since 2013. The LVR policy has also mitigated the likely decline in household spending and economic activity during a stress scenario. All prudential tools have an efficiency cost, which must be weighed against the benefits. The LVR policy is likely to have had a larger impact in constraining mortgage borrowing capacity than alternative macroprudential tools. This suggests that the LVR policy should be deployed primarily when housing and household sector risks are high, to maximise the benefits of the policy relative to the efficiency costs, which are reduced credit access for credit-worthy borrowers and the potential for slower economic growth in the short-term. The speed limit of high-LVR loans is an important calibration tool to mitigate the efficiency cost of the policy. The review also assessed the potential tension between the LVR policy and other social objectives, and other unintended effects of the policy: The initial LVR calibration disproportionately restricted purchases by first home buyers (FHBs), creating a tension with the public policy goal of housing affordability. Over time, the policy has increasingly operated by restricting property investor purchases, and rebalanced the policy burden away from FHBs. The Auckland regional LVR policy (applied for about a year starting in late 2015) contributed to a spillover of housing demand from Auckland to other regions, despite being effective at addressing risks in Auckland itself. Although the provision of mortgages by non-banks has grown since the policy came into place, the scale of this disintermediation remains too small to significantly erode the effectiveness of the LVR restrictions. This suggests that the LVR policy will remain effective in improving bank resilience for a longer period than originally expected. The policy could impact on competition, the construction of new housing, and rents. If these effects have been present, they have not been of a sufficient scale to be identified. There are limits to what the LVR policy can achieve. The LVR tool is relatively ineffective at influencing aggregate demand and inflationary pressures, and therefore should not be used as a lever of monetary policy. More importantly, the pressures on the housing market and the rental market from a growing population, a limited supply of housing, and low mortgage rates require a range of responses that are outside the scope of prudential policy.