Economist Rodney Dickens wonders if bank lending criteria are the reason the housing market remains sluggish despite interest rate cuts

By Rodney Dickens

If dwelling sales reported by the REINZ had responded normally to the significant fall in mortgage interest rates since 2017 the monthly seasonally adjusted number of sales would be around 7,500 now versus just over 6,300 in November (i.e. 15-20% higher).

The lack of upside in dwelling sales since 2017 isn't because interest rates have become irrelevant, it is because numerous special factors have offset the boost from the fall in interest rates.

Examples of the special factors are the October 2018 ban on most foreigners and banks tightening lending criteria in response to the findings of the Australian Financial Services  Royal Commission.

The impact of some of these factors will have ended or largely ended, however, two of the special factors will have an ongoing impact, with one of these - the interest rates banks use to stress test borrowers - possibly changing the way interest rates impact on the housing market permanently.

This article reviews the special factors.

In light of the potential importance of what banks do with the interest rates used to stress test borrowers, Gareth Vaughan of Interest.Co is surveying banks and mortgage brokers to find out what is happening with the results to be released later this week.

The response to the latest fall in interest rates has been nothing like a normal one

The chart below shows that the previous sizeable falls in the average mortgage rates on offer by the major banks since 1993 were all followed roughly four months later by a large increase in the number of existing dwelling sales; but not the latest one (the boxed area). Reflecting dwelling sales reacting to interest rates with around a four month lag, the red mortgage rate line has been advanced or shifted to the right by four months.

The number of sales should have already benefited from most of the fall in the average mortgage rate from 5.31% in June 2017 to 4.17%.

To put the fall in proper context, the 1.14 percentage point fall in the average rate represents a 21% fall in interest costs. It is a large fall that would normally drive a large increase in existing dwelling sales vs. the sideways drift that has occurred since 2017.

The list of culprits for the lack of upside in sales is quite long and includes one factor that warrants further investigation

The list of culprits I can come up with is:

• The ban on most foreign investors.
• Banks tightening lending criteria following the Australian Royal Commission.
• The dumped plan for a Capital Gains Tax and changes in rental laws favouring tenants.
• KiwiBuild stealing some first home buyers and "second chancers" from the market.
• Banks possibly acting in advance of the increase in capital levels the Reserve Bank confirmed earlier this month.
• Something having a negative impact in the Lower North Island, especially Wellington and to a lesser extent in the South Island.
• Immigration changes and especially the cut in residency visas by the government.
• Banks not cutting the interest rates used to stress test borrowers less than they have cut mortgage rates.

What follows is a brief review of the special factors. Some are no longer relevant, while what banks do with the interest rates used to stress test borrowers could be the most important one in the future.

That is why it is the focus of Gareth's investigation.

The October 2018 ban on most foreign investors: This had a moderate negative impact on dwelling sales in late-2018, but the impact ended in early-2019 based on the number of sales to foreigners having stabilised this year (next chart). Australians and Singaporeans are exempt from the ban so there will always be some sales to foreigners.

An Interest.Co article investigated how much the Australian-driven tightening in lending criteria made it harder for would-be borrowers to get mortgages, partly because of large increases in the interest rates used to stress test borrower (link below).

This was driven by the findings of the Australian Banking Royal Commission that were released in February 2018.

The next chart gives an indication of the impact tighter bank lending criteria in Australia in response to the findings of the Royal Commission had on the existing housing market.

Despite mortgage interest rates not changing, the value of new bank loans to owner-occupiers and investors who were borrowing to buy existing dwellings tumbled in 2018.

However lending has spiked since mid-year in response to falling mortgage rates.

In Australia bank lending on existing housing seems to be responding to interest rates normally again (next chart).

Tighter lending criteria undermining the boost to sales from the fall in interest rates significantly: The Australian experience suggests it stopped being a negative factor some months ago but in NZ banks have been scrutinising expenses more this year.

The government's proposal of a Capital Gains Tax: In 2018 this had a minor negative impact on borrowing by investors over the second half of the year and the first half of 2019 based on new mortgage lending by banks to investors (see the blue line in the next chart). Reports at the time also suggested the plan for a CGT encouraged more investors than normal to sell, as did plans for changes to rent legislation that favoured tenants over landlords.

However, in the last few months lending to investors has been the strongest performing of the three categories the Reserve Bank releases info on (next chart). The limited upside in existing dwelling sales this year is down to owner-occupiers including first home buyers not so much investors.

KiwiBuild: Kiwibuild will have stolen some buyers away from the existing housing market last year, but with only 303 KiwiBuild homes sold since June 2018 it hasn't had a significant impact. This is in part because many of the 303 KiwiBuild home buyers probably won't have bought an existing house if they hadn't bought a KiwiBuild home. Progress with KiwiBuild remains slow based on the changes between the September and October "dashboards" (next chart). KiwiBuild may become a bit more of a threat to the existing housing market next year.

The next chart suggests banks were planning on significantly tightening lending criteria for rural and commercial property borrowers and to a lesser extent SMEs based on the Reserve Bank's latest survey of banks conducted in September 2019, but mortgage borrowers didn't appear to face much of a threat (black line).

We assume the tighter lending criteria implied by the survey was because banks were responding in advance of the Reserve Bank's plan to significantly increase minimum bank capital levels.

Banks acting in advance of the increase in capital levels didn't appear to be a significant threat to mortgage borrowers, but may have played a minor part in muting the boost to dwelling sales from the fall in interest rates this year.

A breakdown of NZ dwelling sales into four super-regions provides some insights but also raises some issues (next chart).

Following the negative impact of the ban on most foreign investors in late-2018 to early-2019, Auckland sales have increased roughly as much in the last seven months as would be expected given the size of the fall in mortgage interest rates (blue line, next chart).

However, this will partly reflect increased investor buying in recent months that is likely to be focused quite a bit on the Auckland market, with a bit of this also relevant to the rest of the upper North Island that has also had a reasonable increase in sales (black line).

Sales in the lower North Island have been a bit of a drag on the national figures (gold line, chart above), with this most the case for Wellington but also the case in other parts of the lower NI. Sales in the South Island have not increased much (green line).

Southland has been a bit of a drag on South Island sales but there has been limited upside in general.

Sales don't behave the same all over the country all the time but it is a bit unusual for sales in the lower North Island not to behave much like sales in the upper North Island.

It isn't strange for South Island sales to behave a bit differently from sales in the North Island but the lack of upside in sales in the second half of the year is a bit puzzling.

It is unlikely banks have adopted different lending criteria for different regions, so the divergence in the behaviour of sales in the four super-regions most likely reflects regional factors at work (e.g. maybe higher insurance premiums in Wellington and more buying by investors possibly focused most on the upper North Island).

The regional differences go some way to explaining the lack of upside in NZ sales in response to the fall in mortgage interest rates, even if it isn't obvious why there hasn't been some recovery in lower NI sales or why upside in the South Island has been so muted compared to the upper NI.

Traditionally immigration has been the second most powerful driver of existing dwelling sales; interest rates having been the most powerful. The official immigration estimates derived from a model Stats NZ built in response to the November 2018 dumping of departure cards show a sharp spike in in late-2018 that was more than fully reversed in early-2019.

Recently there has been a bit of upside (blue line in the next chart).

By contrast, based on info derived from arrival cards, that still exist, immigration tumbled in late-2018 to be followed by a partial rebound this year (black line).

Statistics NZ favours the official estimates (the blue line), but we have no faith in the estimates or the card-based numbers that were somehow corrupted by the dumping of departure cards in a manner Stats NZ has failed to explain adequately.

The lack of certainty over what immigration did this year makes it hard to say if it helped mute the boost to dwelling sales from the fall in mortgage rates,but the government cuts in residency visas (black line, next chart) in conjunction with the October 2018 ban on most foreign investors from buying will have had a mild negative impact on dwelling sales.

The fall in the number of immigrants arriving with residency visas has been more than offset by upside in the other visa categories.

However, the ban on most foreigners from buying existing house applies to immigrants with work, student and worker visas.

The change in the mix of immigrants and the ban on most foreign buyers will mean immigration has had a least a mild negative impact on dwelling sales.

Based on anecdotes from the market like those from this link, banks haven't cut the interest rates used to stress test borrowers as much as they have cut mortgage rates.

This could be creating a bottleneck constraint that mutes the impact of the changes in mortgage rates.

If this is the case it is possible what banks do with the stress-test rates will be more important than changes in the actual mortgage interest rates banks offer.

To help shed light on this potentially important issue, Gareth Vaughan of Interest.Co is conducting a survey of mortgage brokers and banks to find out what banks are doing with stress-test rates and more.

If banks' stress-test rates move much less than mortgage rates in the future, it implies interest rates will be a much less powerful driver of housing and economic cycles than has been the case in the past.

This poses a range of questions including what impact OCR changes will have on the housing market and economy.

This article is here with permission.