Future Price Predictions Using My Signature Graph Data

[Scott]

As some of you may know I put together my graph that is shown as my signature early last year. I have been doing some analysis of the figures and thought you may all find them interesting. Basically it points to falls of anywhere between 38% to 55% (not taking inflation into account!).

The average house price figures are from Nationwide and wages from measuringworth.com.

It is interesting that the average salary multiple is 4.17 prior to the current speculative bubble. Having said that this doesn't take deposits into account - it is an average of 3.75 prior to the current speculative bubble if a 10% deposit is included.

I have attached the information below as a pdf document in case it isn't aligned properly. The other pdf is the information entered. I have entered the 2008 figures for reference - these are used for peak calculations.

It makes for interesting reading.

Regards

Scott

4.54 = average salary multiple over last 37 years (1971 to 2007)

4.17 = average salary multiple from 1971 to 2000 (prior to current speculative bubble)

£109,000 = average salary multiple over last 37 years multiplied by 2007 average wage (£24k)

£100,091 = average salary multiple from 1971 to 2000 multiplied by 2007 average wage (£24k)

41% = drop required to bring average house value back down to average salary multiple over last 37 years

46% = drop required to bring average house value back to 1971 to 2000 salary multiples (pre current speculative bubble)

55% = drop required to bring average house value back to the lowest multiple figure over last 37 years

4.62 = average salary multiple over last 37 years (1971 to 2007 peak)

4.17 = average salary multiple from 1971 to 2000 (prior to current speculative bubble)

£114,741 = average salary multiple over last 37 years multiplied by current average wage (£24,840k)

£103,594 = average salary multiple from 1971 to 2000 multiplied by current average wage (£24,840k)

38% = drop required to bring average house value back down to average salary multiple over last 37 years

44% = drop required to bring average house value back to average 1971 to 2000 salary multiples (pre current speculative bubble)

55% = drop required to bring average house value back to the lowest multiple figure over last 37 years

£93,333 = house price if 3.5x multiple and 10% deposit is used

50% = drop required to bring average house value back to 3.5x multiple with 10% deposit

£111,827 = house price if 3.5x multiple and 10% deposit used and salaries rise by 5% per year over next 3 years

40% = drop required to bring average house value back to above

wages_house_prices_1971_to_2007.pdf

percentage_falls.pdf

[Nicholas Cage]

Good work, only bit I would say is to add ONS codes for data, or where it comes from. Also deposit as a % of income has shown a rapid growth not seen in other crashes, unlike income ratios.

http://www.cml.org.uk/cml/filegrab/3ML4.xls?ref=4626

Edited August 5, 2008 by maxwell

[Frank Hovis]

It does Scott, but you're mainly preaching to the converted here.

55% real terms falls?

Well, in North Cornwall I have to say "and the rest?". I'm expecting 70% real terms.

[Nicholas Cage]

IPS error

Edited August 5, 2008 by maxwell

[Scott]

Good work, only bit I would say is to add ONS codes for data, or where it comes from. Also deposit as a % of income has shown a rapid growth not seen in other crashes, unlike income ratios.

Hi Maxwell

I'll look into the ONS thing.

Deposit as a % of income has only risen due to parents stumping up the deposit for their kids and FTB's being aged higher. This should change and go back down to historic norms.

[Scott]

It does Scott, but you're mainly preaching to the converted here.

55% real terms falls?

Well, in North Cornwall I have to say "and the rest?". I'm expecting 70% real terms.

Hi Frank

I know I'm preaching to the converted, but there are many new people coming to the site often as guests who would probably like to see some stats as to where this is headed.

I have tried to predict on a nationwide basis using historic values. I agree that different areas/type of properties will have lower and higher falls. And sentiment, the economy and the credit contraction may very well make it worse this time.

[Liquid Goldfish]

There's the common argument that as interest rates are lower now that the important thing is % of income spent on mortgage rather than house prices in income multiples

I know there's a solid counter arguemnt to this but I've forgotten it - can anyone remind me simply what it is, ta

[ralphmalph]

What is the source of the data in your sig?

[Nicholas Cage]

There's the common argument that as interest rates are lower now that the important thing is % of income spent on mortgage rather than house prices in income multiples

I know there's a solid counter arguemnt to this but I've forgotten it - can anyone remind me simply what it is, ta

Disposable Income would be a better indicator for me, if you pay 50% of your income for drinking water for an extreme example, the model fails.

[Liquid Goldfish]

Disposable Income would be a better indicator for me, if you pay 50% of your income for drinking water for an extreme example, the model fails.

I know this will sound really dumb, but what is "disposable income"

[Scott]

What is the source of the data in your sig?

Nationwide site for house prices. measuringworth.com for incomes.

[Scott]

Disposable Income would be a better indicator for me, if you pay 50% of your income for drinking water for an extreme example, the model fails.

There is no way of knowing what that will be in future though. As I originally stated this is using historic data, not variable future unknown.

Regards

Scott

[Nicholas Cage]

I know this will sound really dumb, but what is "disposable income"

Disposable income, Personal or Household, is the money left over after paying all the income taxes, Council Taxes (which changed drastically so skewing the income to house price ratio compared to last crash) and National Insurance and other realted taxes that are unavoidable from income for the employed and self employed.

Edit:

There is no way of knowing what that will be in future though. As I originally stated this is using historic data, not variable future unknown.

Yes, I am in a minority here and Extrodinary Martini and other well respected posters stick to the income house price ratio. But the effect of changing council tax from a local collection to a government source of funding skews it too far in my opinion. As well as fiscal drag for tax bands and stamp duty etc.

Edited August 5, 2008 by maxwell

[AteMoose]

As some of you may know I put together my graph that is shown as my signature early last year. I have been doing some analysis of the figures and thought you may all find them interesting. Basically it points to falls of anywhere between 38% to 55% (not taking inflation into account!).

The average house price figures are from Nationwide and wages from measuringworth.com.

It is interesting that the average salary multiple is 4.17 prior to the current speculative bubble. Having said that this doesn't take deposits into account - it is an average of 3.75 prior to the current speculative bubble if a 10% deposit is included.

I have attached the information below as a pdf document in case it isn't aligned properly. The other pdf is the information entered. I have entered the 2008 figures for reference - these are used for peak calculations.

It makes for interesting reading.

Regards

Scott

4.54 = average salary multiple over last 37 years (1971 to 2007)

4.17 = average salary multiple from 1971 to 2000 (prior to current speculative bubble)

£109,000 = average salary multiple over last 37 years multiplied by 2007 average wage (£24k)

£100,091 = average salary multiple from 1971 to 2000 multiplied by 2007 average wage (£24k)

41% = drop required to bring average house value back down to average salary multiple over last 37 years

46% = drop required to bring average house value back to 1971 to 2000 salary multiples (pre current speculative bubble)

55% = drop required to bring average house value back to the lowest multiple figure over last 37 years

4.62 = average salary multiple over last 37 years (1971 to 2007 peak)

4.17 = average salary multiple from 1971 to 2000 (prior to current speculative bubble)

£114,741 = average salary multiple over last 37 years multiplied by current average wage (£24,840k)

£103,594 = average salary multiple from 1971 to 2000 multiplied by current average wage (£24,840k)

38% = drop required to bring average house value back down to average salary multiple over last 37 years

44% = drop required to bring average house value back to average 1971 to 2000 salary multiples (pre current speculative bubble)

55% = drop required to bring average house value back to the lowest multiple figure over last 37 years

£93,333 = house price if 3.5x multiple and 10% deposit is used

50% = drop required to bring average house value back to 3.5x multiple with 10% deposit

£111,827 = house price if 3.5x multiple and 10% deposit used and salaries rise by 5% per year over next 3 years

40% = drop required to bring average house value back to above

Scott are you using average salary, or the more commonly used metric when calculating these things, average household salary?

[Mr Nice]

Scott are you using average salary, or the more commonly used metric when calculating these things, average household salary?

average household salary isn't "more commonly used," it's just a different number with different multiples.

a "quality" mortgage is usually no more than 3.5 times individual average earnings, or 2.5 times joint household incomes.

you end up with similar results no matter which metric you use.

[Scott]

Scott are you using average salary, or the more commonly used metric when calculating these things, average household salary?

Hi Moosetea

It is average salary.

Regards

Scott

[crown]

I use similar data and end up with a similar result here

http://thecrownblogspot.blogspot.com/2008/...-june-2008.html

[AteMoose]

Hi Moosetea

It is average salary.

Regards

Scott

just for clarity the both the graphs in your sig uses average household income (which is 32k?), i think that prices will fall but imho using different data skews the data down too much.

Edited August 6, 2008 by moosetea

[Scott]

just for clarity the both the graphs in your sig uses average household income (which is 32k?)

No, individual average income. So for 2007 the figure was £24k.

[AteMoose]

No, individual average income. So for 2007 the figure was £24k.

right one certainly does, it has household in the description.... I agree with average salaries however there is more to this than meets the eye.

1. Less women stay at home now than in the past, especially for working young couples. The result is average household salary is increasing faster than average salary, because average household salary is a maximum threashold on income it sets the boundary and its the sensible metric to use

2. Average salary/household salary increases by 4.4% PA according to the ONS stats, over the past hundred or so years average salary has always risen, if you expect the crash to last three years your salary will rise slightly reducing the salary multiple.

IMO If you account for future wage growth during a crash use household salary you will get a better prediction. The million dollar question is will the government print more money and create more inflation in the system (to reduce the salary multiple)

Edited August 6, 2008 by moosetea

[Extradry Martini]

As some of you may know I put together my graph that is shown as my signature early last year. I have been doing some analysis of the figures and thought you may all find them interesting. Basically it points to falls of anywhere between 38% to 55% (not taking inflation into account!).

The average house price figures are from Nationwide and wages from measuringworth.com.

It is interesting that the average salary multiple is 4.17 prior to the current speculative bubble. Having said that this doesn't take deposits into account - it is an average of 3.75 prior to the current speculative bubble if a 10% deposit is included.

I have attached the information below as a pdf document in case it isn't aligned properly. The other pdf is the information entered. I have entered the 2008 figures for reference - these are used for peak calculations.

It makes for interesting reading.

Regards

Scott

Rather than a list of numbers, why don’t you chart it? (see below)

Anyway, I believe house prices are going to fall 35-50% peak to trough in nominal terms (i.e. after inflation). The graph below (Nationwide price index / ONS average earnings) demonstrates that outside of leveraged and irrational bull markets driven by equally leveraged and irrational bull markets in credit, real estate prices oscillate between 3 and 5 times earnings. I believe that the pain is going to be so great in the housing market and the accompanying recession so severe that we will return to at least 3 times average earnings. That translates into a 35-50% nominal fall in house prices.

OK, so how do I come to a 35%-50% peak to trough fall in nominal house prices? The first step is to see where it is likely to go in earnings-adjusted terms: I believe (because I have seen it time and again) that when bubbles burst markets fall to levels well below fair value – as the famous elastic band gets over-stretched one way, it gets over-stretched the other way in reaction. To me, that means 3 times average earnings – especially given that it went there after the much milder 1989-1994 bear market. From the peak of 6.86 times earnings a fall to 3 times earnings is in the order of 56%.

The next step is to estimate average earnings growth over the coming years: The MPC has one brief – to keep inflation as close to 2% as possible. We know that they are expecting inflation to remain high (even going up to 4%) before falling back to their target. Let’s assume, conservatively, that this doesn’t happen quite in the way they expect and consumer price inflation averages as high as 3% over the coming years (interest rates would almost certainly go up in that case, but let’s assume they don’t). To CPI, we add real earnings growth: that is rarely much more than 1% in a downturn, but let’s assume its 1.5%. These three conservative assumptions (i.e. that inflation remains high, that rates don’t go up in response and that real earnings growth is higher than is normal in a recession) give us nominal earnings growth of 4.5% per year.

The last step, and probably the most difficult, is to estimate how long the bear market is going to last. There are various scenarios here, and this is what gives me the range of falls. To begin with lets look at what happens if house prices fall the same amount every year over different timescales in order to fall by 56% in earnings-adjusted terms:

Years to trough from 2007 peak:.....8.................5.................3

Earnings adjusted total price chg:...-56%..........-56%...........-56%

Earnings adjusted p.a. price chg:....-9.8%........-15.1%........-23.9%

Nominal p.a. earnings growth:.........4.5%..........4.5%............4.5%

Nominal p.a. house price chg:........-5.3%.........-10.6%........-19.4%

Nominal total house price chg:.......-35.1%.........-43%.........-47.7%

Ok, so that’s how the 35%-50% range is defined. So which is the most likely? My view is that as house prices are so stratospherically far from fair value, there will be a big fall over the first 3 years to around 3.7 times earnings(a crash phase), followed by a longer slower phase of the bear market to 3 times earnings as it will take a very, very long time before confidence returns. (By “confidence” I mean something akin to 1996 confidence, not the feverish madness of 2006). Were these phases to take 3 years each, we would get:

“Crash phase”

Years to trough from 2007 peak:......3

Earnings adjusted total price chg:..-46%

Earnings adjusted p.a. price chg:....18.6%

Nominal p.a. earnings growth:........4.5%

Nominal p.a. house price chg:......-14.1%

Nominal total house price chg:......-36.6%

“Slow phase”

Years from end of crash phase:......3

Earnings adjusted total price chg:..-19%

Earnings adjusted p.a. price chg:..-6.8%

Nominal p.a. earnings growth:........4.5%

Nominal p.a. house price chg:.......-2.3%

Nominal total house price chg:......-6.6%

The calculations for the slow phase are taken from the end of the crash phase, so we have to adjust the final number to put it in 2007 peak terms, meaning that the entire nominal house price fall over the 6 years is 40%. That, representing a reutn to 2002 prices according to the Nationwide data, is my very best guess at this stage.

What are the risks to this view? I think the major risk is to the downside - i.e. to falls below 3 times earnings. Not only have I been conservative in my high forecasts of earnings growth and inflation without rate rises, but I haven't factored in at all a big rise in unemployment which ios very possible and whihc would have a significant impact on house prices relative to earnings.

[Extradry Martini]

Okay. So 40% nominal in 6 years. If we have inflation of 3% cpi pa then we can add c.15% to the 40% to get a real correction of 55% in your view.

I am working it back from an earnings-adjusted fall anyway, but if you want it just in terms of an inflation-adjusted fall, around 51%.

Calculation is a s follows: (((0.6^(1/6) - 1) - 0.03) + 1)^6 - 1

Edited August 7, 2008 by Extradry Martini