By Rodney Dickens*
Property Investors Federation executive officer Andrew King has been advising landlords to put up rents to try and offset the negative impact of mortgage interest rate increases on cashflows and profitability.
Andrew's warning is timely because the upside risk to interest rates is real, as I outlined in the January Raving.
Unfortunately for landlords, this isn't likely to happen to any great extend because low rental inflation appears to be a by-product of low general price or CPI inflation.
This Raving identifies what appears to be a major change in how rents increase as a result of the move from high general inflation in the 1970s and 1980s to low general inflation that started in the early-1990s.
This could be seen as a legacy Dr Brash has left landlords. This legacy has implications for house sales and house prices, as covered in our Housing Prospects reports.
The breaking of the link between house prices and rents
One of the great puzzles over the last decade has been the yawning gap that has opened up between the national average house price and the national average annual gross rental income (left chart).
This can be shown equally by annual rental inflation remaining low despite periods of high house price inflation since 2000, while in earlier periods of higher house price inflation rental inflation was much higher (right chart).
The result has been a large fall in the national average gross rental yield (black line, left chart below).
Most of the fall in the rental yield occurred between 2002 and 2007. This was a period during which annual house price inflation averaged 13.6% based on the QV House Price Index and rental inflation averaged 2.7% based on the rental component of the CPI.
Over this period the average mortgage interest rate increased and the 10-year government bond yield - a measure of low risk fixed interest returns for investors - changed little (left chart below).
So over the 2002 to 2007 period there wasn't a case for investors accepting lower rental yields because interest rates were lower, although the subsequent fall in interest rates has retrospectively justified some of the fall in the rental yield.
The reason rent increases didn’t vaguely keep up with increases in house prices between 2002 and 2007 and again over the last couple of years appears to be because rental inflation has become a hostage of low general price or CPI inflation.
In the 1970s and 1980s when house price and CPI inflation were high, it made sense that rental inflation was also high, although all three experienced significant cycles. The top right chart above compares house price inflation and rental inflation, while the right chart directly above compares rental price inflation and CPI inflation.
CPI inflation has been consistently lower since the early-1990s. In the first house price boom after 1990 - between 1994 and 1996 - rental price inflation increased much as had been the case earlier.
But it would seem that by the time of the 2002 pickup in house price inflation the low inflation environment had taken over.
It would appear that landlords began to struggle to justify or achieve higher rental inflation in an environment of sustained low general inflation.
This new, inflation-constrained behaviour is reflected in CPI and rental inflation living in similar ballparks since after the government interventions resulted in rental inflation turning temporarily negative in 2001 (right chart above).
The result was that the relationship between house price and rental inflation largely broke down (top right chart on page 2).
This doesn't mean that rental inflation can't increase somewhat when house price inflation increases, but rental inflation should have started to increase already if it wasn't constrained by the low general inflation environment and low inflation expectations.
Part of the story appears to be about the constraint of low inflation expectations, which can be seen as one of Dr Brash's legacies. The left chart below compares rental inflation with the ANZ survey of CPI inflation expectations that is only available back to 1988.
If landlords and tenants have reasonably low expectations about inflation in general, in line with the findings of the ANZ survey, it makes sense that this will put a limit of sorts on how much landlords can justify increasing rents.
This can also be viewed from the perspective of rents compared to incomes.
This is another means by which lower general inflation, including lower income growth, will put a ceiling on rental inflation. The right chart compares the ratio of the national average house price to the average annual gross income (black line) with the ratio of the average national annual gross rent to the average annual gross income (blue line).
Prior to 2000 the two ratios largely moved in synchrony while since then the rent/income ratio has fallen and the house price/income ratio has skyrocketed.
While being below the peak level, the current rent/income ratio is still a bit above the average level that existed between 1978 and 2000.
The fall in the rent/income ratio since 2000 may also reflect increases in government support for low income renters, with all categories of renters included in the CPI rent component.
But getting back to Andrew's recommendation, what about the behaviour of rents relative to mortgage interest rates?
The adjacent chart shows annual rental inflation compared to the average listed mortgage rate of the major banks.
In the mid- 1990s when rent inflation increased in response to higher house price inflation it also went someway to offsetting the negative impact of interest rate increases on cashflows/profitability for landlords.
However, when the average mortgage interest rate last increased significantly - between 2003 and 2007 - between 2003 and 2007 - rental inflation didn't change much.
The 2003 to 2007 experience suggests low CPI inflation is such a large constraint landlords aren't able to achieve larger rent increases even if interest costs increase significantly.
Maybe this time will be different, but I can't see why.
If landlords aren't able to achieve significantly larger increases in rent in the face of large increases in interest costs it will impact on how many investors will buy and sell properties over the next couple of years (i.e. fewer investor buying and more selling).
It will also impact significantly on the choice between renting and buying, as it did last decade.
This means it will impact on the number of dwelling sales, the demand-supply balance in the existing housing market and house prices, as discussed in the July Housing Prospects report.
*Rodney Dickens is the managing director and chief research officer of Strategic Risk Analysis Limited.