Skip to content

Latest RBNZ figures show NZ's new mortgage borrowers stretching themselves much less than has been seen in recent years in regards to the amount of debt they're taking on relative to income

Banker_house.jpg

By David Hargreaves

New mortgage borrowers are this year continuing to take on very much less debt relative to their incomes than has been seen in recent years.

New figures from the Reserve Bank (RBNZ) show that overall the debt-to-income ratios of new borrowers in March 2023 were the lowest since the RBNZ started compiling this information - albeit that this only goes back to 2017.

The RBNZ said that in March, the monthly share of new mortgage commitments for all borrowers with a debt-to-income (DTI) ratio of over five times (that's debt more than five times annual income) was 34.4%, down from 38.8% in December 20221.

The RBNZ said the monthly share of new commitments with DTI of over five is the lowest recorded since the data collection began in June 2017. The share has fallen from a recent high of 60.2% in November 2021.

This reduction will be giving the RBNZ considerable comfort as it looks ahead to the possible introduction of DTI limits early next year.

The detailed DTI figures are compiled monthly, but released quarterly. What the data has shown in the time the RBNZ has been producing it is that DTIs were at quite high levels in 2017, dropped through 2018, started rising again in 2019 and became stratospheric through 2020-21, hitting peak levels in late 2021.

Since then the declines in DTI ratios of new mortgage borrowers have been pretty spectacular. Clearly a number of factors are prompting this trend. House prices have on average reduced by about 17.5% from their late 2021 peaks - so, less money needs to be borrowed. Incomes have increased. But mortgage rates have increased enormously too and this coupled with more conservative lending policies from the banks means it's no longer possible for some would-be homeowners to borrow at the sort of elevated DTI ratios previously seen.

This comes through very clearly in the latest DTI figures produced by the RBZ.

It will be interesting to see how the current declining trend in DTI ratios for new mortgages influences the RBNZ's thoughts on possible implementation of DTI limits.

The RBNZ has wanted to have a DTI tool in its 'macro-prudential toolkit' (a toolkit that already includes the loan to value ratio or LVR limits already in use) since at least 2016. But the RBNZ struggled to secure government support for DTI measures, firstly from the National-led government and then the current Labour government. This was due to concerns about the potential impact on first home buyers.

Having finally received government approval in 2021 the RBNZ then began preparatory work and earlier this year released a debt servicing framework, which will enable restrictions to be possibly brought in by March 2024 if needed - with the banks therefore getting 12 months to get their systems ready, should they be required.

Notably, however, the RBNZ says given the housing market is currently in a downturn, there's no immediate need to implement DTI restrictions.

The RBNZ keeps a close eye on borrowing that's done on DTIs of over five - in other words where the amount borrowed is over five times the annual income of those taking out the mortgage. It's not completely clear what sort of DTI levels the RBNZ would be 'happy' with. And the question of what sort of limits might be imposed if a debt servicing framework is introduced have not yet been explicitly addressed.

What would be clear though is that the RBNZ would now be very happy with how the DTI ratios of new borrowers are tracking.

In terms of some of the highlights pinpointed by the RBNZ in the latest data release, which, remember is put out quarterly, but with monthly figures, the RBNZ says the monthly share of new mortgage commitments with DTI of above 5 has now decreased for 13 months in a row.

As we've done since the start of this data series we are comparing the latest month's figures (March 2023) with the last month from the previous release (December 2022) and we are also comparing both these with March 2022 - and this time as a treat we've added an extra column to our tables so you may also compare the two-years-ago figures from March 2021.

As ever, we've got two tables for you with the first one (immediately below) showing the figures for first home buyers (FHBs) and other owner occupiers, while the second table looks at figures for investors and owner-occupiers who have investment property collateral.

So, as for the first table immediately below, DTIs of above five are regarded as getting up there, so we highlight the percentages of total mortgage money that is borrowed by both first home buyers and other owner occupiers at DTI ratios of above FIVE. Please note that our calculations here exclude the (small) amount where the DTI size is unknown.

The table below shows the percentage of new mortgage money for first home buyers and other owner-occupiers that is on debt-to-income ratios of over five times:

Table_latest_int.JPG

So, some pretty substantial falls evident there.

That's the FHBs and the owner-occupiers. Our second table looks at the investor and those owner-occupiers with investment collateral. For this table we choose a more bracing DTI level and look at the percentages of those with debt-to-income ratios of over SEVEN times. Again our calculations exclude the (small) amount of mortgage money where the DTI size is not known.

The next table shows the percentage of new mortgage money for both investors and owner occupiers that have investment collateral that is on debt-to-income ratios over seven times:

Table_latest_int2.JPG

So, there again we have it.

How much lower can they go? Or are we going to see upward movement again? It is interesting to see that some of the owner-occupier with investment collateral figures have blipped up a little - but from fairly low levels.

We'll be keeping an eye on the data as it is released.

This story was originally published on Interest.co.nz and has been republished here with permission.