GDP versus real house prices
Bursting bubbles
In the Bank of England’s draft policy statement on supplementary capital requirements, they have identified a house price to rent ratio as a core housing market indicator.
Interestingly, they have used an average of the Nationwide and Halifax house price indices which gives a more benign picture of the market than if you use the “official” ONS house price index.
It really highlights how, in this low turnover market, there’s not even a clear picture of what national house prices are doing
Distribution of annual real house price growth (and falls)
Combining the Nationwide house price index and ONS RPI data shows that UK house prices fell by -4.1% in real terms during 2012 continuing the trend in negative price performance this decade.
2 year price growth at a local level
Some interesting maps I’ve made showing price growth over the last two years using HM Land Registry data. The maps really highlight the geographical variations in recent price growth with only London seeing significantly positive growth.
Top tip: predicting the ONS house price index 3 months in advance
Given that the ONS house price index is based on the Regulated Mortgage Survey (it excludes cash buyers that currently make up 1/3 of the market) and is recorded at completion, it is actually quite easy to see where it’s heading in the future (within reason).
All we need to do is look at the average mortgage approval value from the Bank of England and push it out by three months (the typical time between approval and completion). The Bank of England data isn’t mix-adjusted so does show some difference in performance but overall it provides a good guide to what the ONS will eventually say three months later and so I expect the index to weaken slightly over the next three months.
Land Registry: More markets saw price falls than rises over the last 3 months
Of the 172 housing markets covered by the Land Registry across England & Wales, 98 markets saw price falls while only 74 saw price rises in Oct 2012 as measured on a three monthly basis. This compares to price falls in 90 markets last month and 72 markets in August when the balance was last firmly in favour of price rises.
The Land Registry has house prices indices for all counties and unitary authorities in England & Wales except for the City of London and Isles of Scilly due to sample sizes.
Nationwide house prices flat in Oct, point to -0.5% year end growth
Following a much less pronounced spring/summer seasonal increase in house prices this year, today’s Nationwide index shows that house prices have continued to flat-line in October with 0.1% monthly growth. The widely reported 0.6% monthly rise is based on seasonally adjusted figures and reflects the fact that prices have tended to weaken during October in recent years.
My bespoke model based on historic trends suggests that house prices will end the year at about -0.5% below their level seen in December 2011, although this has improved slightly from the -0.8% it was predicting last month.
First-time buyer deposits, loan-to-income & loan-to-value ratios
This is a follow-up to my complicated chart from 10 days ago on mortgage rates, loan-to-income ratios & affordability; which demonstrated that current mortgage lending relative to income levels is appropriate for the existing interest rate environment but gives cause for concern if rates were to return to pre-recession levels.
This chart looks at the relationship between the size of deposit and size of mortgage relative to income for first-time-buyers. As I’ve shown previously, the size of mortgage was consistently around 2.2 times income during the late 1980’s and throughout the 1990’s but from 2000 the decrease in base rates and increased competition drove loan-to-income ratios up, eventually peaking at 3.2 in 2005.
However, the impact of this increase in size of mortgage relative to income was that: with loan-to-value (L.T.V.) ratios consistently around 75-80%, the size of deposit required to become a first-time buyer almost doubled from 55% of income in 1998 to 104% in 2004.
Note; the sharp drop in both loan-to-income ratios and deposit requirements during 2005 is most probably due to a change in data sample but the market weakness seen during this period may also have contributed.
With the resumption of the boom in late 2005, loan-to-income ratios kept increasing until the credit crunch hit during 2007/08. As the credit crunch reduced the willingness and ability of lenders to offer higher L.T.V. products and with loan-to-income ratios falling only slightly, the deposit requirements for first time buyers shot up from 72% of income in 2007 to 137% in 2010.
Mortgage lending has now eased slightly in terms of L.T.V. but still remains constrained relative to long term trends. Limited higher L.T.V. products combined with new mortgages at around 3 times income mean that the average first-time buyer is now having to raise 108% of their income for a deposit. It is no surprise that the private rented sector has been booming in recent years and that first-time buyers are heavily dependent on parental assistance.
Increasing the availability of higher L.T.V. products would go some way to helping improve deposit affordability but, at current house price levels, a 90% L.T.V. mortgage would require prospective first-time buyers to borrow 3.6 times their income. This would then push mortgage repayments up to around 23% of their income which is similar to the unaffordable levels seen during 2006/07.
Mortgage rates, loan-to-income ratios & affordability
The rather complicated chart below shows the relationship between the size of the loan relative to income (L.T.I.) and the respective mortgage rate for each quarter over the last 32 years. Generally, the trend has been one of reducing rates and therefore increasing L.T.I.s as households have been able to service a larger debt with a similar proportion of their income.
However, there are two clear periods where the relationship breaks down; the first is from 1988 when mortgage rates increased substantially and stretched affordability until they returned to previous levels in 1992. This obviously coincided with the late 1980’s house price boom and subsequent crash.
The second period occurred from 2004 when both an increase in mortgage rates and continued increases in the size of mortgages relative to income stretched affordability until the end of 2007 when falling L.T.I.s and then falling rates substantially improved affordability. I shouldn’t need to remind you that this period also coincided with a boom and subsequent crash in house prices.
So what does that mean for the current market? As we can see from the chart, the size of new mortgages being lent relative to prevailing mortgage rates shows that the current affordability of repayments is well within a more stable area of the chart. However, if mortgage rates were to increase back to their pre-2008 levels of 5-6% but L.T.I.s remain where they are, then the chart suggests that affordability will become as stretched as it was in 2006/07.
What’s going on with the “official” house price indices?
There’s quite a bit of divergence in the results of the ONS and Land Registry indices at the moment. The chart below shows the annual change in prices for the two indices (smoothed over 3 months) and it’s clear than the ONS index is indicating much better (or less bad) performance than the Land Registry across all regions but in particular the North of England, West Midlands & Wales.
It’s important to remember that the ONS index is based on data from the Regulated Mortgage Survey and so will only reflect the mortgage dependent market whereas the Land Registry index is based on repeat sales (measuring the difference between recently sold prices and historic prices for the same property) and so should cover a wider element of the market (it does have some exclusions).
Given that mortgages currently account for 66% percent of all transactions and the ONS reckons that their sample accounts for 65-70% of all mortgaged transactions, that suggests that the ONS is only capturing data on 43-46% of the market.
Although the Land Registry index isn’t perfect and only includes those properties for which repeat sales data exists, I think it does a better job at monitoring the wider market than the ONS index.
Therefore, the above chart suggests that one possible explanation for the divergence is that mortgage lenders are focusing on select parts of the market that are experiencing better performance than the wider market and the ONS mix-adjustment process is failing to account for this.
Nationwide house prices: lowest trajectory in 3 years
With no sign of a September bounce, house prices are on their lowest trajectory in 3 years and are pointing towards -0.8% price falls by the end of the year.
Regional house prices - UK prices flat over the last two years
Only London has seen substantial growth during the last year while the recovery from the market trough in 2009 has been limited to the south of England & Wales.
ONS House Prices - doing a great job at telling us what happened three months ago
Today’s ONS house price index shows annual growth of 2% reflecting the relative market strength we saw earlier in the summer from the asking price/mortgage based indices.
I expect the index to weaken in coming months as the Olympic lull feeds through to the ONS data.
Nationwide explanation: July weakness rather than August strength
Today’s Nationwide house price index is being widely quoted as the biggest monthly jump since January 2010. However having a more detailed look at the figures takes the shine off this attention grabbing statistic.
The 1.3% monthly growth figure quoted is actually a seasonally adjusted figure and it looks like Nationwide have made their typical August adjustment by adding approx 1% to the underlying growth figure of 0.2% as August tends to be much weaker than July.
But as we can see from the chart below, house prices since March this year have been on a lower trajectory than the last two years and June/July were particularly weak as we didn’t see the run-up in prices that typically occurs during this period.
With this in mind, a more realistic explanation of the index today would be that the market has remained flat during the summer and is generally weaker than previous years.
Perhaps Nationwide also need to look at their seasonal adjustment given all the one-off events that have take place this year.
This will be one of my last blog posts at this address. I’ll soon be moving to my new home on the savills.co.uk research page (details to follow).
Comparison of mortgage lender house price indices
With today’s Lloyds HBOS index pointing towards slightly weakening prices at a national level, it is worth having a look at the two house price indices produced by mortgage lenders.
Following the 2008/09 price falls, the Nationwide index recovered strongly and since 2010 has flat-lined with house prices up 10.2% while the Lloyds HBOS (Halifax) index showed a far weaker recovery in 2009 and prices are only 2.4% above their lows in 2009.
Although the sample size and geographic bias of each index are not published, it is generally known that the Nationwide tends to have better coverage in London and the south of England while Halifax tends to have better coverage in the north of England.
The chart below, showing price changes since the bottom of the market, reflects this trend: Nationwide shows far stronger growth since 2009 in London and the south of England while Halifax shows much weaker growth in these markets and almost no recovery outside of London and the South East with negative price changes in Yorkshire & Humber, Scotland and the North of England.
At the moment both indices are showing a similar story with fluctuating seasonal price changes leading to flat/slightly falling annual growth but it will be worth watching the two indices in coming months to see how they deal with their regional biases with other housing market indicators pointing to a slowdown.
