For the past three years, UK house prices as whole have gone pretty much nowhere.
You may think that this is a good thing, especially if the 2008/9 fall wiped tens, or even hundreds, of thousands of pounds off the value of your home.
In any case, given the euro crisis, weak growth in the US and the ‘double dip’ recession in the UK, there seems to be more important things to worry about.
The trouble is house prices are still too high. My colleague Phil Oakley has argued that, although propped up by low interest rates, the housing market is “rotting from within”.
And it look as though, before this year is out, all eyes may be on housing again. The latest data from Nationwide claims that prices fell by 0.7% in July alone. More importantly, this was a 2.6% fall compared with the same time last year – the largest annual decline seen in more than three years.
Of course, Nationwide isn’t the only company measuring house price statistics. Indeed, our house-price obsessed nation produces at least seven different indices purely on prices (this piece doesn’t deal with the monthly data from the Royal Institution of Chartered Surveyors, which doesn’t measure prices directly). These price indices are produced by: Rightmove, Hometrack, Office for National Statistics, Nationwide, Halifax, Land Registry and Acadametrics
Which are the most reliable? Which are the most up-to-date? And are there any early warning signals that could give you a clue as to when another crash is about to begin? Let’s have a look.
Asking prices: Rightmove and Hometrack
The Rightmove and Hometrack indices are based on asking prices, the first stage of the home buying process. These are the prices that those hunting for homes online will see from a sample of estate agents (Hometrack) or on Rightmove’s website (Rightmove.co.uk – which covers about a third of the market).
In theory, this should make these indices the timeliest measures. They should pick up price trends long before they appear further down the road in figures based on mortgage approvals or completed sales.
However, there are also some obvious downsides. For a start, as you can see from the latest figures (below), the average prices produced by each index are very different. Presumably this is down to the sample size, but the gap is quite striking.
More generally, final prices are often very different to the initial asking price. Sellers can have unrealistic views about the value of their property. We’ve all encountered owners who think they should get as much as the house down the road, even although it’s in better condition and has a larger garden. Also, if they bought it for a higher price in the near past, it may be difficult to accept that they will lose money on the deal.
Estate agents sometimes contribute to this, using unrealistic price estimates to get people to sign up with them, only to suggest a change a few weeks later (see Phil’s piece on valuing houses for more on this).
This doesn’t just cut one way of course. If a property is in an area of high demand, then it’s possible that it might spark a bidding war, driving the final price up.
And on top of all this, there’s no guarantee that just because a house is on the market that it’ll sell. The owner may have to withdraw it, then put it on at a lower price at a later date.
Overall, according to Hometrack’s own data, the completion price usually ends up being 5-10% lower on average than the asking price. During 2008, the final discount in the North of England ended up being nearly 15%.
These measures – the time taken to sell, and the average discount from original price – might be more useful than asking prices for showing how much stress the market is under.
Latest figures
Rightmove: asking prices fell 0.7% in July to £242,097, and rose 2.3% on last year.
Hometrack: asking prices fell 0.1% in July to £151,800, and fell 0.5% on last year.
Mortgage approval stage: ONS, Nationwide and Halifax
The Office for National Statistics (who took over from the DCLG), Nationwide and Halifax (now owned by Lloyds) all produce data based on lenders’ valuations at mortgage approval stage.
The idea is that this is timelier than data based on the final selling price, but it is also near enough the end of the home-selling process to be a more realistic snapshot than the initial asking prices.
The use of lenders’ valuations also means that it is based on expert opinion, filtering out cases where a buyer (or seller) went wild.
However, some of these strengths could equally be seen as flaws. Like all compromises, collecting data in the middle of the process, means that it is less timely than asking prices, but less definite than final sales.
Basing it on lenders’ valuations assumes that they are a better judge of value than buyers and sellers. As we all know, banks lent money like drunken sailors up until 2007, overvaluing most properties.
Many have argued that they have gone too far in the opposite direction since then, which means that their valuations have erred on the side of caution. We take the point, although we’d argue that the banks are right to be more cautious.
The large gap between the ONS figures and the other two is partly down to the different methodology used – the ONS figures are not ‘mix-adjusted’, as we explain below.
Latest figures
ONS: house prices were unchanged in May at £228,000, and rose 2.3% on last year.
Nationwide: house prices fell 0.7% in July to £164,389, and fell 2.6% on last year.
Halifax: house prices went up 1% in June, to £162,417 and fell 0.5% on last year.
Completion: Acadametrics and Land Registry
The final approach is to base the data on agreed prices from the Land Registry’s records (Acadametrics’ index is based on the same data but uses a slightly different methodology, which produces a higher average price).
This is the least subjective approach, since it isn’t based on an estimate by a lender, or what a buyer might think the market can bear. It also doesn’t rely on sampling (as in the case of Hometrack), or data from one firm (in the case of Halifax, Nationwide and Rightmove).
However, the fact that the prices are only taken at the end of the process means that the underlying data is the least timely of all the indices. The Land Registry admits on its website that there is sometimes a further lag between the sale taking place and the data being recorded.
The Land Registry data also includes cash sales, which may distort their data. This is because they involve high-end properties, which are not representative of the wider market. Both indexes cover England and Wales only.
Latest figures
Acadametrics/LSL: house prices fell 0.1% in June to £224,102, and rose 3% on last year.
Land Registry: house prices rose 0.1% in June to £161,777, and fell 0.9% on last year.
What else should you be aware of when using these figures?
With the exception of Hometrack and Rightmove, all the indices are seasonally adjusted, to take account of the fact that house prices are generally weaker in the winter and stronger in the summer.
Similarly, six of the seven measures (the ONS is the odd one out) are adjusted to take account of the fact that houses are not identical.
To take a silly example: if you sell a hundred Chelsea townhouses one month, then sell a hundred terraces in Newcastle the next month, it’ll look as though prices have collapsed. But all that’s happened is you sold a load of very expensive houses one month, and a load of cheap ones the next month. That tells you nothing about the ‘average’ house price.
So the figures are ‘mix-adjusted’ to account for this. Factors taken into account include (but are not limited to); bedrooms, bathrooms, garden, location, parking rights.
However, the ONS index is not mix-adjusted. Instead, it is a simple average of the mortgage valuations in the month. The argument is that there are so many transactions that the differences in the type of property sold end up averaging out.
So which house price index is the best?
What’s reassuring is that despite the differences in methodologies, all the indices tend to produce similar results in terms of the trend of prices, if not of the actual prices recorded. As a result, many prefer to use the Nationwide’s index, simply because it has the longest track record, having started in 1952.
However, in the shorter run, the differences can be important. Given that they are based on asking prices, it is not surprising that the Rightmove and Hometrack measures tend to lead the others by up to six months.
However, during periods of volatility the Rightmove and Hometrack data become much less useful. This is because households take time to adjust their asking prices.
You could also argue that the ONS, Nationwide and Halifax data have become less useful since the big house price crash. Sales have plunged drastically, and as a result their sample sizes have fallen too. Against that backdrop, leaving out cash sales, for example, means you are missing out a proportionally more important part of the market.
So in all, if you want the most accurate picture, then the data with the longest lag – the figures from Acadametrics or the Land Registry, which are based on final sales – are probably the best to go for.
Recommended video
Tim Bennett looks at some of the most popular house price surveys and explains the differences between them, how they work, and how useful they are as a guide to house prices.
• Watch