Cutting back on highly leveraged mortgages
Thu 17 Oct 2013 by Gareth Kiernan in Construction

The Reserve Bank’s new policy of limiting the amount of high loan-to-value mortgage lending aims to improve the robustness of the financial system, by trying to reduce the risk of default faced by retail banks in the face of a large drop in house prices.  The policy may also affect sales activity and prices in the housing market, but there is considerable uncertainty about exactly how big those effects will be.

What are the policy details?

From October 1, the retail banks are required to restrict mortgage lending with an LVR of over 80% to no more than 10% of the dollar value of all new mortgage lending.  The restriction does not apply to existing loans, lending on farms, Welcome Home Loans under the Housing NZ scheme, or bridging loans.  There will be some ability for existing high-LVR borrowers to transfer their loan if they move house or change bank without the loan being subject to the LVR restrictions, as long as the value of the loan does not increase.  However, loan top-ups will be subject to the new restrictions.

The calculation of the 10% restriction will initially be applied over a six-month period from October 2013 to March 2014 to give the banks time to adjust their lending practices, with the proportion calculated on a three-month rolling basis after that.

Confused aims

As we mentioned in the introduction to this article, the Reserve Bank’s primary aim with the policy is to reduce the threat to financial stability of a substantial fall in house prices.  By limiting the amount of high-LVR lending on banks’ balance sheets, the policy will reduce the number of loans that could move into negative equity territory if there was a large drop in house prices – thereby reducing the chance, and magnitude, of loan defaults following such a drop in prices.

When he announced the policy in August, Reserve Bank governor Graeme Wheeler also stated that the “restrictions are designed to help slow the rate of housing-related credit growth and house price inflation”, as well as “dampening housing demand.”  In much of its communication, the Bank has been careful to separate the financial stability agenda of the LVR restrictions from its monetary policy functions that are carried out through setting the official cash rate.  But Mr Wheeler’s speeches and recent editorial in the New Zealand Herald, along with a previous speech by Grant Spencer in June, seemed to present the macro-prudential restrictions as an alternative or compliment to the traditional monetary policy mechanism of interest rates.1

From a monetary policy perspective, the Reserve Bank should not be concerned about house prices for their own sake – even though the current Policy Targets Agreement mentions asset prices specifically as something that the Bank should monitor.  House prices do not enter directly into the CPI – the closest subcomponents are residential rents and residential building costs, both of which have some correlation with house prices.

The main monetary policy concern arising from rapidly rising house prices is the correlation with household consumption spending, aggregate demand, and, ultimately, broader inflationary pressures.  Research into last decade’s housing boom internationally is mixed about whether the wealth increases associated with the surge in house prices directly caused faster consumption growth through top-up loans and equity withdrawal2

An alternative explanation is that homeowners who are credit-constrained use their increased collateral to lift their current consumption.  However, either way, the Reserve Bank should only be concerned about a housing market boom if it is unsustainably boosting aggregate demand and generating broader inflationary pressures.  In this context, the solution should not be macro-prudential regulations, but a simple increase in the official cash rate.

On the prudential side, introducing loan-to-value limits in the case of a house price ”bubble” will decrease the number and value of loan defaults that will be experienced by a retail bank when the bubble pops.  The Reserve Bank has made a song and a dance about the fact that its policies may reduce the size of the bubble, but even if there is no effect on house prices, LVRs will ensure that there is a smaller default rate – improving the stability of the banking system.

Even in this framework, it is difficult for economists, analysts, and officials to determine whether an asset price bubble is occurring at any particular point in time.  In other words, trying to fine-tune policy settings on the basis that the housing market, for example, is out of equilibrium and is due for a correction, is highly problematic and assumes a level of certainty about current conditions that is unlikely to be obtained.

Furthermore, even if there is a bubble, loan-to-value limits are an unusual tool for dealing with these financial stability concerns.  As we’ve noted, the limits on LVRs work by reducing the chance of default on mortgages.  However, a retail bank has an entire balance sheet of loans – and as a result, if part of its balance sheet now has lower risk, it may be willing to increase risk in other parts of the balance sheet to increase returns.  This “waterbed effect” is a common unintended consequence of such regulations.

How the banks currently stand

The residential mortgage books of the five largest banks totalled $164.8bn at the end of June 2013, up from $155.5bn a year earlier.  Of the $9.3bn increase in mortgage lending over the year, $4.3bn, or 47%, was on loans with an LVR of over 80%.  Most of the banks have expanded their high-LVR lending in line with their market share, apart from ASB, which has been responsible 54% of the growth in high-LVR lending over the last year (compared with a 23% market share a year ago).  At the same time, Westpac has actually shrunk its high-LVR loan book over the last 12 months.

The Reserve Bank estimates that, once the exemptions to its new policy that we have outlined above are taken into account, about 15% of new mortgage lending will be able to have an LVR of over 80%.  All other things being equal, this estimate implies that, rather than the $4.3bn of high-LVR lending that occurred over the last year, the policy would have limited high-LVR lending to closer to $900m.  Or, put another way, annual growth in total mortgage lending could have been about 3.7% instead of 6.9%.

The mortgage market response

With restrictions on the quantity of high-LVR loans available, the segment of the market that is most likely to miss out on loans are first-home buyers.  Most existing homeowners are likely to already have more than 20% equity in their property, while those with less than 20% equity have the ability to transfer their loan (although the policy will make it more difficult for them to trade up to a more expensive house).  Existing bank lending policies mean that few property investors will have loans with LVRs of over 80%.

Banks have historically had policies in place that have made borrowing with an LVR of over 80% more expensive than borrowing with a deposit of 20% or more.  These practices included lenders mortgage insurance, where an additional one-off fee was added to a mortgage at drawdown to cover the higher risk of default with a high-LVR loan.  Between ten and 15 years ago, this risk was taken on by an independent insurance company, but the biggest banks have generally moved to “self-insure” over the last decade.  Kiwibank was the last of the big five banks to move to self-insurance as QBE LMI withdrew from underwriting in New Zealand in August this year.

Increasingly common over recent years have been differential interest rates for borrowers with LVRs above or below 80%.  The interest rate margins imposed by banks have ranged from 0.25 percentage points for loans with an LVR of 80-85% to 1.25 percentage points for loans with an LVR of over 90%.  Recently, however, both ANZ and BNZ have increased their margins, citing increased capital requirements for high-LVR lending as making it more expensive to undertake this lending.  In the case of ANZ, its margin on mortgages with an LVR of over 90% is now a whopping two percentage points.

On this basis, it appears that the retail banks’ approach to restricting growth in high-LVR lending will be a mixture of price-based and quantity-based rationing.  Additional fees and/or higher interest rates for mortgages with an LVR of over 80% provide some incentive for borrowers to try and raise a bigger deposit – although data from the last year suggest that existing fees and margins are not enough to quell growth sufficiently for the Reserve Bank.  So unless much larger fees or margins are imposed, it is likely that banks will be able to cherry-pick the best customers based on income and other measures of creditworthiness.  Some aspiring homeowners may find themselves unable to obtain finance from a bank at any price.

From a borrower’s perspective, though, there are a number of ways of trying to circumvent the new restrictions.  Firstly, there is the potential to source a larger deposit from other family members such as parents.  Anecdotal evidence suggests that, given the mounting housing affordability issues over the last decade, parental assistance in buying a first house has already become more common.  Banks will continue to look at income, household budgets, and evidence of saving in assessing a borrower’s creditworthiness, but getting the LVR below 80% with money from other sources could be the difference between a loan being approved or not.  Westpac’s recent introduction of its “Family Springboard” home loan options indicates the bank’s willingness to help facilitate this sort of process.

Secondly, borrowers may also look to top up their deposit with money from a non-bank financial institution (NBFI).  The Reserve Bank’s rules only apply to registered banks, leaving the second-tier lenders largely free to do what they like.  It would make little sense for a property buyer to take out their whole mortgage with an NBFI at a relatively high interest rate, but there will be an incentive for some borrowers to top up their deposit with this type of loan.

What of the housing market?

The lending figures outlined earlier suggest that there should be a sizable slowdown in credit growth with the introduction of the new LVR policy.  And there are already reports that bank pre-approvals for mortgages with LVRs of over 80% have dried up in the wake of the Reserve Bank’s policy announcement in August.

In other words, the LVR policy will reduce the potential number of house buyers in the market – if not in absolute terms, at least at any given price.  With first-home buyers hardest hit, the lower end of the market is likely to be affected first, but the reduction in demand in that segment of the market will have ripples through the broader market as existing homeowners find it more difficult to sell their properties and trade up.

The immediate effects of the drop in demand are likely to include a lull in house sales and some slowdown in house price inflation.  The effects will be most marked in Auckland, where house prices are among the highest in the country and high-LVR lending is likely to be most prevalent.

Ultimately, though, we believe the LVR policy will only have a temporary effect on housing turnover and house price inflation.  We expect many prospective first-home buyers to try and find a workaround to enable them to still purchase property.  And in the current economic environment, demand from investors and cashed-up foreign buyers is likely to replace those first-home buyers who find themselves excluded from the market.

The overriding issue that has been driving the strong house price inflation in Auckland over the last 18 months has been an undersupply of new dwellings.  Any temporary slowdown in house price growth due to regulatory changes will weaken the signals to build more new houses in the region, resulting in the undersupply of housing persisting for longer and leading to faster house price growth in future years.  Given the massive gap between underlying demand for housing and the supply of new houses, funding constraints for property development, and strengthening net migration and population growth, we expect these fundamental drivers to reassert themselves by the second half of 2014.  House sales, house price growth, and residential building activity are likely to pick up again in the second half of next year.

And the broader economy?

As we mentioned in March,3 there are a number of channels through which LVR limits may influence broader economic outcomes.  Outside of the direct effect on housing, and the flow-on effect on retail spending, there are three other key channels.

  • The potential to place credit constraints on owner-occupied building
  • The potential to constrain finance for small businesses
  • Uncertainty about “what size OCR hike” the policy is equivalent to – implying a risk that the policy may effectively tighten monetary conditions more than expected, slowing the economic recovery

The Reserve Bank is conscious of all these risks, but felt that the importance of ensuring stability in the financial system exceeded the potential economic cost of a slower recovery and/or lower near-term investment.

In our current forecasts, we have taken on the same assumptions as the Reserve Bank about the effect of LVRs on these economic outcomes beyond the direct housing market effect.  In other words, we have assumed that the effect on aggregate building activity and small business investment will be small or moderate, and that the tightening in credit conditions will be equivalent to a 30 basis point interest rate hike.

However, there is a risk that the LVR restrictions will restrict credit conditions more heavily than we have allowed for.  In this case, it would be investment and durable good spending that would show the first signs of weakness during the closing stages of 2013 – a potential outcome we will be keeping a careful eye on.

1 Graeme Wheeler’s speech accompanying the announcement of the LVR restrictions can be found at, while his editorial in the NZ Herald is at  Grant Spencer’s June speech is available at

2 See as an example of this research – note that there has since been a series of studies in other countries since with similar results.


Related Articles