By Rodney Dickens*
Since early-2009 low mortgage interest rates have disguised the extent to which New Zealand has a housing affordability problem.
Interest rates should remain low while the financial crisis continues to fester, but at some stage in the next five years or so the crisis will have abated enough to allow interest rates to return to more normal levels.
The time it will take for the financial crisis to abate significantly is directly relevant to the length of the fuse on the housing affordability time-bomb.
This Raving quantifies the extent of the underlying housing affordability problem and how much low mortgage interest rates are disguising the problem.
I don’t expect mortgage interest rates to increase significantly any time soon, but when the financial crisis eventually abates there is a real risk that interest rates will increase significantly.
To take this into account this Raving presents three scenarios that provide some valuable insights (that is, that house prices are unlikely to keep increasing more than incomes; that house prices are more than normally vulnerable to increases in mortgage interest rates).
It also warns of the bear trap awaiting unwary home buyers who, encouraged by the recent experience of low mortgage interest rates, gear to the max.
Quantifying the underlying housing affordability problem and the interest rate disguise
The REINZ stratified median house price increased from $114,000 based on the average for the three months to March 1992 to $382,000 based on the average for the three months to July 2012, a 235% increase (black line in the chart).
Three month averages are used to smooth the volatility in the monthly median prices, while the NZ stratified median price released by REINZ is a pretty good proxy for the national average house price.
In isolation, the 235% increase is meaningless. To have meaning it needs to be compared to what happened to the prices of goods and services in general, as measured by the Consumers Price Index (CPI) and to incomes.
The green line shows how much the national average house price would have increased since 1992 if it had increased in line with prices in general as measured by the CPI. It would have increased to just over $180,000 instead of $382,000 or a 58% increase instead of a 235% increase.
Since 1992 the national average house price has increased 4x more than prices in general, which should ring warning bells.
But in countries with growing populations, like New Zealand, it is normal for house prices to increase more than prices in general, just as incomes generally increase more than prices in general.
The blue line in the chart above shows what would have happened to the national average house price if it had increased in line with incomes since 1992. Average hourly earnings, including overtime, from the Quarterly Employment Survey, are used as the income measure. If house prices had increased in line with average hourly earnings the national average house price would have increased to just over $212,000 rather than $382,000, an 86% increase rather than the 235% increase that occurred.
Since 1992 the national average house price has increased 2.7x more than incomes.
Now this really should start to ring warning bells.
However, looking at the relationship between house prices and incomes in isolation is misleading when mortgage interest rates have changed hugely since 1992. The blue line in the above chart shows that the average mortgage interest rate charged by the major banks is currently 5.88% versus 9.3% in January 1992 (right scale).
The black line shows the impact of both changes in the national average house price and changes in the average mortgage interest rate on the affordability of buying the average national house for the average employee.
The black line assumes the following:
(1) in each month since January 1992 the average income earner buys the average house;
(2) he/she borrows 80% of the value of the house:
(3) he/she pays the average mortgage interest rate applying at the time.
The black line shows the resulting annual interest cost expressed as a % of the average employee’s gross annual income. In early-1992 the average employee had to spend 29% of his/her annual gross income to pay the annual interest on the mortgage if he/she bought the average house with 80% debt.
Most recently, the average employee would have to spend 34% of his/her annual gross income to pay the annual interest on the mortgage if he/she bought the average house with 80% debt. So in interest-cost terms there isn’t currently a major housing affordability problem. Certainly not compared to the situation in 2007 when the average employee had to spend over 60% of his/her annual gross income to pay the annual interest if he/she bought the average house with 80% debt. But the moral isn’t that there isn’t an underlying affordability issue (and two negatives do make a positive in this instance).
The moral is that the financial crisis, which is largely behind the low interest rates, is helping disguise the underlying problem.
This chart provides another way of quantifying the underlying housing affordability problem that is currently being disguised by super-low interest rates and will continue to be disguised for some time. There are two lines in the chart, but they move in perfect synchrony because they each use income as the denominator and because proportional scales are used in the chart.
The black line, right scale, shows the national average house price as a multiple of the average employee’s annual gross income.
In early-1992 the average house cost 4x the average employee’s annual gross income. At the peak in 2007 the average house cost 8.6x the average employee’s gross income and most recently it cost 7.2x the average employee’s annual gross income.
There is still a massive underlying housing affordability problem (i.e. there is still a ticking time bomb). But it probably won’t raise its ugly head again until mortgage interest rates eventually increase to more normal levels.
The thick, light blue line hiding behind the black line shows the same thing but from a different perspective. It assumes the average employee buys the average house each month and it shows the deposit (20% of the value of the house) as a percentage of his/her annual gross income (left scale).
In early-1992 the 20% deposit was equal to 80% of the average employee’s annual gross income. At the peak of house prices in 2007 this had increased to 171% and most recently was 144%. Assuming the average employee pays a 25% tax rate and can save 10% of his/her after-tax income per annum, the time needed to save a 20% deposit to buy the average house has increased from 10.7 years in early-1992 to 19.2 years currently.
At the peak of prices in 2007 it would have taken 22.8 years. Super-low mortgage interest rates are largely disguising the housing affordability problem in terms of the interest outlay for the average employee buying the average house. But this doesn’t overcome the dramatically larger hurdle would-be first home buyers now face when it comes to accumulating a deposit. The specifics I used above won’t be relevant to all would-be first home buyers, but they help quantify how dramatically things have changed over the last 20 years and especially the last 10 years in terms of how long it now takes would-be first home buyers to accumulate deposits. This applies irrespective of whether the target for would-be first home buyers is a 20% deposit or a 10% deposit.
The BNZ-REINZ monthly survey of real estate agents that Tony Alexander writes is useful in helping assess how far we are down the track in terms of would-be first home buyers having saved the larger deposits now needed relative to incomes. In the last year it has reported a significant increase in first home buyers, which has been partly helped by Kiwisaver. To access these useful monthly reports use this link.
But I suspect that many of the first home buyers entering the market don’t realise the bear trap that awaits them down the road if they haven’t been conservative in the amount of debt they have taken on board.
The ticking time bomb of housing affordability
As discussed in our monthly economic reports, the financial crisis is a long way from over (for info on these reports visit http://www.sra.co.nz/index.php/interesting-times).
The best international research I have read on the topic suggests that when a debt crisis is eventually resolved a period of strong economic growth emerges, which was NZ’s experience in the mid-1990s after crawling out of a debt and inflation crisis. And with strong economic growth comes significant interest rate increases, as happened in NZ in the mid-1990s.
It therefore appears to be a question of when not if interest rates will return to more normal levels, although it is too early to push the panic button.
In the Housing Prospects reports we predicted that house prices would increase this year and we expect the upturn to continue into next year. Since we commenced business in 2006 we have provided clients with advance warnings of every upturn and downturn in house prices, so we are confident that it is too early to hit the panic button (see http://www.sra.co.nz/pdf/HousingProspectsSample.pdf for an assessment of our track record at forecasting house prices and housing market activity). But it is useful to preview what may happen when the veil of affordability provided by super-low interest rates is removed.
Predicting what will happen on the financial crisis front is extremely difficult, if not impossible. This means that any predictions of interest rates should be treated with lots of scepticism. So rather than predict the future for interest rates and the housing market I provide three scenarios that help highlight the implications of a significant increase in mortgage interest rates over the next five years.
All three scenarios assume that the average mortgage interest rate increases from 5.88% currently back to the average rate of 8.12% experienced since 1992 by September 2017 (see the light blue lines in the three charts on this page, right scales).
Scenario 1 (above) assumes that life goes on as normal with house prices and incomes both increasing at the average annual rates experienced since 1992 (i.e. 6.4% and 3.0%, respectively). If this occurs the average employee would need 55% of his/her gross annual income to pay the annual interest bill on a mortgage equal to 80% of the value of the average house by September 2017 versus 34% currently (top chart).
This path leads to a housing affordability time bomb.
This isn’t to say it won’t happen, at least in part. But if we head down this path for any length of time a subsequent significant fall in house prices is inevitable, just as happened after housing affordability last became a massive challenge in 2007.
Scenario 2 (above) assumes that house prices and incomes increase at 3% per annum, which is the historical average rate of growth in the measure of incomes used in this analysis.
This means that housing affordability remains the same relative to incomes as is the case currently.
But with mortgage interest rates increasing, the annual interest cost of buying the average house with 80% debt would increase to 47% of income by September 2017 versus 34% now. Again, it can’t be ruled out that we go down a path like this for a period. But the constraint of housing affordability would eventually kick in and result in house prices increasing less than incomes (that is, if mortgage interest rates eventually return to around the historical average rate then house prices will probably increase less than incomes).
Scenario 3 (above) looks at things from a different perspective. It assumes that mortgage interest rates return to average and that incomes grow at the average rate of 3%, which is the same as the other two scenarios. But it assesses what the average house price would have to do to keep the annual interest cost associated with buying the average house with 80% of debt fixed at the historical average of 34% of annual income. By chance, the current annual interest cost of buying the average house with 80% debt is equal to the historical average of 34% of income (after rounding).
For the annual interest cost to remain fixed at 34% of income, the national average house price would have to fall from $382,000 currently to $326,500 by September 2017 (i.e. a 15% fall). It is extremely unlikely that the national average house price will fall 15% over the next five years.
But Scenario 3 is useful in highlighting the vulnerability of house prices to rising mortgage interest rates because the current starting point is one in which house prices are still extremely high relative to incomes.
Scenario 3 shows that even if incomes increase at the historical average rate over the next five years it makes only a little dent in the underlying housing affordability problem, leaving house price more than normally vulnerable to increases in interest rates. All three scenarios highlight that the underlying housing affordability problem is still large.
This means it will probably take quite a bit longer than five years for some form of normality to return, while whatever occurs it won’t happen in the neat, straight-line fashion shown in the scenario charts.
The bear trap that lies ahead for unwary house buyers
People almost invariably base their expectations of the future on relatively recent experience.
Mortgage interest rates have in general been below 7% for the last three years. Consequently, I suspect that many recent home buyers will have geared up on the assumption that mortgage interest rates won’t increase much (i.e. won’t increase above 7%).
But the current situation is unusual and won’t last indefinitely.
The recent experience is also that borrowing floating or relative short-term fixed rates has worked best and reflecting this 80% of total outstanding bank mortgages were either floating or fixed for less than one year based on the latest data supplied by the Reserve Bank, while a massive 95% of outstanding mortgages were either floating or fixed for less than two years in July (as in above chart).
The focus on floating and short-term fixed mortgages may be quite rational for minimising interest costs over the next couple of years. But if this remains the case it means lots of borrowers will get an unpleasant surprise when interest rates eventually return to more normal levels.
Home buyers who gear to the max over the next couple of years on the assumption that interest rates won’t increase much face the risk of a disaster.
Again, recent experience has undue weight in the decisions people make.
This means it is almost inevitable that a reasonable portion of home buyers over the next couple of years will unwarily borrow too much and not fix for long enough.
This group risks having an unpleasant encounter with a bear trap when interest rates eventually increase to more normal levels.
* Rodney Dickens is the managing director of Strategic Risk Analysis. See more detail here.