Can We Predict The Future For House Prices?
Should we be worried about the fall in House prices? Can we predict the future?
Why we cannot read the future from the past
There are many factors that can affect house prices over a year or two and much of that we cannot predict. Therefore, the falls that are happening now should not necessarily affect our buying behaviour or the prices we pay for houses.
To start with, we do not know how consumer, homebuyer and employee sentiment will change in the near future.
We also do not know what will happen with the economy and interest rates. Previous falls in house prices may partly reflect what we think will happen next but, what we think will happen and what does happen are frequently very different.
According to one house price index, we have just experienced three quarters of house price falls in a row. House prices are now 1.6% lower than a year ago, according to Halifax quarterly data, at an average £164,000.
At times like this, you will often read the opinion that a one-month move in house prices does not mean much, but that three-month or longer periods are "more meaningful". This is certainly true in the sense that house prices are likely to change more over a longer time frame - so the price movement is naturally more meaningful to anyone trying to buy or sell a property at that time.
However, for forecasting purposes it is no more meaningful than flipping a coin.
Looking back over Halifax's records, which began in 1983, it's clear to see that even a nine-month fall in prices still tells us absolutely nothing about the future. Half of the time, when house prices have fallen for three quarters, they have ended up lower one year later, while for the other half, house prices have been higher.
Other figures back this up
Nationwide data which some believe is more accurate and in any case goes back further, almost 60 years to 1952.
According to Nationwide, we have suffered just two quarters of falls recently, not three. Historically, we can see that house prices have fallen for two quarters in a row on 16 occasions.
On 10 of those occasions, house prices have ended up lower one year later. That makes the probability of a further fall just under 63%.
Six heads out of 10
However, we can always twist and turn data to support any argument and we need to be careful not to do that accidentally. With just 16 periods to work with, and overlapping ones at that, we do not have much data to rely on, in the same way that if you toss a coin 16 times and it comes up heads 10 times it would not be surprising.
What we have here with our limited house price history demonstrates the complete opposite: since the result is not an extreme one, we can be confident that house prices are not predestined to do anything just because they have fallen for a while. Hence, we would do well not to let the previous house price direction influence our forecasts.
Size doesn't matter
Considering not just the direction but the size of the fall, thinking that perhaps a bigger drop over two quarters might be more likely to lead to a fall over the following year. This was not the case.
The average fall over two quarters was the same for both outcomes: when prices were lower one year later, the two-quarter fall had averaged 5%, but it was also 5% when prices rose.
Once again, then, we have no evidence that past house price moves will help us to predict the future.
Forget charts and forecasts
Considering the dreadful track record of forecasters, professional or otherwise, it makes sense to consider more your own personal circumstances, and less what everyone else thinks and what direction the house-price chart has just turned.
Many events that are yet to affect house prices cannot be reflected in the current price or in recent price movements, and thus we cannot use them to forecast the future. You would do just as well to flip a coin.