Reserve Bank analysis suggests that 'speed limits' on high loan to value lending will reduce the number of house sales by about 5% in a year

The Reserve Bank's new "speed limits" on high loan to value lending might knock annual house sales by about 5%, the RBNZ's own analysis shows.

Based on the 79,472 house sales that were reported in the year to August, this would mean about 4000 less house sales in a year.

The 5% estimate is contained in an RBNZ analytical paper: "Estimating the impacts of restrictions on high LVR lending" by the RBNZ's Chris Bloor and Chris McDonald.

The authors also concede that the LVR limits, introduced on October 1, might have an impact on new house construction, "which would be an unintended consequence of the policy".

Their modelling scenarios suggest that the level of dwelling consent issuance might fall by 40-80 a month, a 2% to 5% fall from current levels, "reflecting a historically close correlation between housing market turnover and the level of consent issuance". However, they add that, for a number of reasons, it's possible the impact won't be of that magnitude.

Volatility

But they say that, generally, LVR restrictions are "likely to result in a degree of volatility for the first few months".

"The six week notice period provided before LVR restrictions came into effect, as well as the significant stock of pre-approvals outstanding, could have caused some buyers to rush in.

"Conversely, significant uncertainty, as well as conservatism on the part of the banks in allocating the speed limit in the initial months, could result in sharp declines in activity initially. As a result, it may be a number of months until the impact of the policy can be properly gauged."

The paper's authors estimate that LVR restrictions are likely to reduce housing credit growth by 1-3 percentage points and house price inflation by 1-4 percentage points. They say that in "one specific scenario featuring a severe housing market downturn and a sharp rise in interest rates" LVR restrictions imposed for two years would reduce bank losses on their residential loan book by 10% to 15%.

"Considering the range of estimates here, we conservatively estimate that the reduction in banks’ downturn losses on their housing loan portfolios would be 10-15 percent if LVR restrictions were applied for two years immediately prior to a major fall in house prices."

Change in behaviour

However, the authors say it is possible that LVR restrictions also lead to a reduction in house price expectations and to a change in consumer behaviour. While a "small allowance" has been made for "this channel" in their estimates, "a significant change in behaviour could result in larger quantitative impact on house prices and credit growth than modelled here", they say. 

The authors give a "starting point" for their analysis as banks’ gross new mortgage flows.

In the lead-up to the announcement of LVR restrictions, new monthly mortgage flows were running at an annualised rate of about $50 billion.

Lending at LVRs of over 80% was about 30% of this figure, "so roughly $15 billion annually".

This figure represented just over 8% of the stock of outstanding housing credit of $184 billion.

The authors say that after allowing for a reduction of total housing lending volumes as a result of the LVR restriction, banks should be able to write about $4.6 billion of high LVR loans within their speed limit and a further $2.5 billion of high-LVR lending is expected to be exempted.

Reduction of $7.9 billion

"This means that banks will be required to reduce their annual high-LVR lending by about $7.9 billion in order to meet the speed limit restrictions."

They assume that around 25% of affected borrowers will be able to find alternative sources of funding to reduce their bank mortgage LVR to 80 percent (for example, by borrowing from family members). Where this money comes from outside of the intermediated financial sector, there will be a small reduction in measured credit growth.

"The larger effect on credit growth comes from the remaining $5.9 billion of borrowing stemming from housing transactions that are not likely to be completed as a result of the restrictions. This number is expected to be offset somewhat by higher equity buyers completing purchases in some cases (taking on about $2 billion in debt), and by about $1 billion of debt remaining on the balance sheets of prospective house sellers."

Bloor and McDonald say that the net effect of this is that credit growth is expected to be about $3.2 billion lower over the first year that LVR restrictions are in place.

Credit growth down

"This would lower annual housing credit growth by 1.7 percentage points in the first year."

However, the longer LVR restrictions are in place, the more likely it is that borrowers would be able to find alternative sources of funding and alternative buyers would enter the market.

"As a result, we expect that LVR restrictions would have a diminishing effect on credit growth after the first year."

The authors use the same assumptions in reaching the conclusion about a 5% drop in house sales.

"The starting point for this is the 30% share of new mortgage lending by value that is at high LVRs.

"On average, high-LVR mortgages are about 30-40% larger than low-LVR mortgages, so the number of mortgages that are high-LVR is likely to be around 24% of the total.

High LVR house sales

"Furthermore, we assume that 10-20% of housing market transactions are undertaken without a mortgage, so the share of house sales involving a high-LVR mortgage is likely to be 20-22%."

Therefore, allowing for the effect of the speed limit, the assumption that 25 percent of affected borrowers will be able to raise a larger deposit, and the effect of new entrants coming into the market, "we estimate that the net effect on house sales is likely to be a reduction of around 5%", Bloor and McDonald say.

They say that "to the extent that they take some demand pressure out of the economy", LVR restrictions will also reduce the amount of work required of monetary policy to maintain price stability (and perhaps modestly ease pressure on the tradables sector).

"The Reserve Bank’s September Monetary Policy Statement presented estimates that if LVR restrictions reduced house price inflation by 2.5 percentage points (the mid-point of the estimated range), the resulting reduction in general inflation pressure would result in 90-day interest rates that were 30 basis points lower than otherwise.

"However, while these two policies would be equivalent in terms of reducing overall inflation pressure, the more targeted nature of LVR restrictions means that they are likely to have a larger impact on house prices.

"If monetary policy were to be used to target house prices directly, the OCR would need to be increased by much more than 30 basis points to achieve the same housing market outcomes as the LVR restriction."