Most people I speak to think that UK house prices are ridiculously high, especially in London.
Admittedly, this is not a novel idea. However, as a very minor student of bubbles I think it’s worth thinking through the possible implications.
So in the rest of this post I want to outline a) why I think we’re in a house price bubble and b) what the implications might be for the future of house prices in the UK.
Bubble alert: The UK house price to earnings ratio is far above average
According to the Halifax House Price Index, the average seasonally adjusted house price in the UK is £217,400. That’s 5.7-times the average UK homebuyer earnings of £38,200.
Since 1983, Halifax’s UK house price to homebuyer earnings ratio has averaged 4.2 and so the current value of 5.7 is some 35% above that historic norm.
More importantly, the only time in that whole 33-year period that the UK house price to earnings ratio was higher than it is today was between February and October 2007.
So the only time in the last 33 years that house prices have been more expensive relative to earnings than they are today was during the very peak of the peak of the last housing bubble.
To make this picture a bit clearer, the chart below shows how the average UK house price has changed over that period, along with the range of values it might reasonably have taken in any given year:
Here’s a quick description of what that chart shows:
- The black line – The average house price in each year
- The red zone – Where the average house price would have been if houses were historically expensive, i.e. if the PE ratio had been between 5.5 and 6
- The yellow zone – Where the average house price would have been if houses were at historically average valuations, i.e. if the PE ratio was between 3.8 and 4.5
- The green zone – Where the average house price would have been if houses were cheap, i.e. if the PE ratio had been between 3 and 3.3
The house price rainbow heads upwards over the years as homeowner earnings increase. As a result, what constitutes an expensive or cheap house price also increases.
As you can see, whenever the house price to earnings ratio gets anywhere near the red zone, house prices fall back towards more normal levels.
This is because the factors that drive house prices up to expensive levels (such as low interest rates, a booming economy, relaxed mortgage lending criteria, lack of supply or irrational exuberance about future house prices) do not last forever.
The economy is cyclical and the forces that push house prices up faster than earnings are ultimately temporary and reversible.
So in the long-run the house price to earnings ratio tends to revert back to its long-term average, given enough time.
And that insight is the basis of my ten-year house price forecast:
Expected capital gains from UK housing are zero over the next ten years
In the long-run I think the most likely outcome for the house price to earnings ratio is that it will revert to its long-term average, which is currently 4.2.
If that were to happen tomorrow then the average house price would immediately fall to around £160,600, a decline of some £56,900 or 26%. This is what I refer to as “fair value”.
Of course that isn’t going to happen tomorrow as the housing market is very illiquid. Illiquidity means that house prices move much more slowly than prices in liquid markets such as the stock market.
On a positive note, slow changes in house prices can actually help reduce downside risk. If it takes a few years for house prices to fall, then during that time earnings will typically increase.
So the reduction in the house price to earnings ratio is likely to come from both falling house prices and rising earnings, which means that house prices won’t have to fall so far to reach historically average levels.
Here are a few assumptions for the next ten years:
- Inflation – Runs at 2% a year (the Bank of England‘s target)
- Earnings growth – 1.5% a year above inflation (a historically average figure)
- House price to earnings ratio – Declines, reducing the gap between its current value and its long-term average value by half each year (a historically reasonable rate of mean reversion)
This reasonable set of assumptions leads to the following chart which shows my forecast for the “expected value“ of average UK house prices over the next ten years:
Again, a quick description:
- The black line – The expected future average UK house price
- The red zone – Where the average house price would be if it continued to be at historically expensive levels
- The yellow zone – Where the average house price would be at historically normal levels
- The green zone – Where the average house price would be at historically cheap levels
As before, the house price rainbow goes upwards over time as homeowner earnings increase.
Despite those increasing earnings, the forecast is still for declining house prices over the next few years as the PE ratio reverts back to average levels.
However, the average house price is not forecast to fall all the way to its current fair value of around £160,000 because earnings, and therefore fair value, are forecast to rise each year.
So by 2020 the average house price is forecast to have fallen almost back to fair value, but by that stage fair value is forecast to have risen to almost £185,000 thanks to inflation and real earnings growth.
As a result the risk of a dramatic house price crash is somewhat reduced, but the outlook is still not good.
As the forecast chart shows, there is a very serious chance that average UK house prices will stay flat for the next decade.
Or perhaps I should say “another” decade, given the lack of overall price increases since 2007 (unless you’re in London, where prices are currently well above the previous highs of 2007).
And of course flat prices would mean that real (inflation-adjusted) prices would decline over that period.
To put it into numbers, this simple model has:
- Average homebuyer earnings at £54,000 in 2026 (which sounds like a lot but won’t feel like a lot in 2026)
- Average house prices at £228,000, just slightly higher than where they are today
The future is uncertain, but the risks are real
Obviously I don’t actually know what’s going to happen to house prices over the next ten years, just like I don’t know what’s going to happen if I drive down the motorway at 155mph.
But in both cases it’s possible to make an informed judgement and forecast of the range of possible outcomes.
Some possible outcomes might be good (house prices could stay high and I might make it to my destination in record time and in one piece) while others might be less good (there could be a massive house price crash and a massive car crash).
My conclusion is that in both cases the expected outcome, i.e. the average of all possible outcomes, does not look particularly good*. As a consequence I will not be re-entering the housing market or driving down the motorway at 155mph anytime soon.
*When I say “good” I mean from an asset class returns point of view. Personally I think house price mean reversion would be a positive outcome for the UK as I fail to see how excessively high house prices are good for most people or the economy as a whole.