Interest rates and Sydney property values - how correlation analysis teaches us not to worry
An interest rate rise makes it harder to pay off a mortgage, but does it have any effect on house prices?
Do rising interest rates mean decreasing property prices? If history is any guide, not at all. In fact, analysis of our data reveals that interest rates have no effect on the capital growth of property at all.
At first glance, this claim seems difficult to believe. How can interest rates have no effect on property capital growth rates? We'll come back to that question after answering a more fundamental one - how do we know that there's no effect?
The answer lies in the statistical concept of correlation. Correlation measures, roughly speaking, the extent to which two factors move together. For example, one would expect the number of ice cream cone sales that take place at a beach to be strongly correlated with the day's maximum temperature. In other words, when the temperature is high one would expect ice cream sales to be high.
...an interest rate rise this year... a drop in sales next year. But there's no corresponding drop in sale prices. Why not?
In the above example, the correlation comes about because one factor (the high temperature) causes the other factor (the high ice cream sales). This is not necessarily always the case. For example, one would expect that ice cream sales in Sydney would also correlate strongly with, say, scarf sales in London. It would be absurd to suggest that high ice cream sales in Sydney cause high scarf sales in London - or vice versa. The correlation arises due to the fact that Sydney and London are on opposite sides of the planet and summer here is winter there.
So, how do we use correlation to show that interest rates have no effect on house prices?
We simply measure the relationship between house price inflation and interest rates to see if high interest rate periods correspond to low house price inflation periods and vice versa. (This is what is known as negative correlation - when one factor is high, the other tends to be low.)
The table below shows the correlation between house price inflation and interest rates lagged a number of years (the lag simply means that instead of comparing house price inflation rates to interest rates this year, we compare it to one, two or more years ago. This allows for the possibility of interest rates having a delayed effect on house price inflation.). The data used covers over 30 years.
| Number of Lag Years | Correlation |
|---|---|
| 0 | 10.44% |
| 1 | 12.61% |
| 2 | 13.34% |
| 3 | 3.51% |
| 4 | -1.87% | 5 | 11.29% |
Correlation Table
So, what's this table telling us? Pretty much what was said in the opening paragraph. The correlation numbers in the right-hand column are low and statistically insignificant. In other words, there is no discernible relationship between interest rates and house price inflation.
(Roughly speaking, the correlation number represents the amount of uncertainty that can be removed in guessing the value of one factor, given that you know the other factor. For example, if house price inflation and interest rates had been correlated 100% (perfect correlation), then that means that knowing the interest rate would remove 100% of the uncertainty about house price inflation. That is, you would be able to predict house price inflation perfectly. If the correlation was 0%, knowing interest rates would remove 0% uncertainty about house price inflation. That is, interest rates would tell you nothing about house price inflation. The table shows figures that are much closer to 0% than 100% and, for all intents and purposes, interest rates have no impact on house price inflation.)
A person looking at the graph of interest rates and house price inflation shown below might notice rising interest rates in 1974, 1982 and 1989 and corresponding low house price inflation in following years and claim a pattern. But then, interest rates also rose in 1970, 1976 and 1985 and each time were followed by accelerating or steady house price inflation. The correlation calculations don't lie. Any pattern 'spotted' by the human eye is simply a byproduct of the human tendency to spot patterns in just about anything. That's why psychiatrists invented ink blot tests.
So we know that interest rates have no impact on house price inflation. How do we reconcile this fact with the intuitive feeling that interest rates and the housing market are closely related?
The answer, we believe, lies in the number of property sales. While we don't have quite as much historical data on the number of property sales as we do on capital growth, we do possess enough data to calculate some rough correlation numbers.
| Number of Lag Years | Correlation |
|---|---|
| 0 | -32.63% |
| 1 | -44.45% |
| 2 | -17.66% |
| 3 | 9.35% |
| 4 | -2.01% |
| 5 | -9.90% |
We see here much higher negative correlation. When interest rates are high, the tendency is for the number of property sales to be low and vice versa. The effect is most significant with a one year lag. That is, an interest rate rise this year corresponds most strongly to a drop in sales next year. But there's no corresponding drop in sale prices. Why not? The demand for property has dropped, why aren't prices dropping?
Because supply has dropped also. For home-owners, particularly owner-occupiers, there is a strong resistance to selling a property for less than what was paid for it. There is little incentive for an owner-occupier to sell a property for less than that which was paid - they will still have to live somewhere, and still have to pay to do so, so why add a loss on the property market to their burden? Owner-occupiers will generally do whatever it takes to meet their mortgage payments. Budgets will be tightened, extra jobs will be sought, etc.
Investors are a slightly different story. A loss can be taken on an investment property without sacrificing the investor's place of residence. If interest rates increase, then the viability of maintaining the investment property will usually be reconsidered. However, the self-sustaining nature of the housing market also has an impact. Lower numbers of property purchasers implies a higher number of renters, which in turn implies higher demand for rental properties and the potential for the investor to increase their rent. In many cases, this rent increase will compensate for the increase in interest rates, removing the source of the original impulse to consider selling.
Obviously, each individual case varies. Some people simply cannot make the increased mortgage repayments and will have to sell. The general trend however, historically speaking, has been for people to do everything they can to hold their properties until the market can pay them at least what they paid for it.
Will this historical trend hold in the current round of rising interest rates? Or will the Sydney property market see price decreases as a result of the interest rate increases?
We usually expect the future to perform similarly to the past, unless there has been some fundamental change in circumstances. One could argue that current lending practices have placed borrowers in a more leveraged position than they have historically been and this may make the housing market more unstable than it has historically been.
The counter-argument to that is that if borrowers are more leveraged than they have been historically then there is arguably more incentive for them to not sell at a lower price than that which they paid.
Overall, one can reasonably say that Sydney property values are as safe now as they've ever been. And a property investor need not lose money on their investments - provided they are in a position to decide when to sell. As long as their financial structurings are such that they are not forced into a sale, the current rising interest rate climate need be no more than a temporary glitch.
What now?
You might like to share this article with other people:
add to digg | post this to del.icio.us | print this
If you’ve enjoyed this article, you can click on one of the tags above
to find articles on similar topics, or look at our
main articles page to see what else we
have written.