Defining what constitutes a Housing Crash: US Research

[kerrynorth]

There is very little information out there in how to to define a housing crash as distinct from a slump. With the economy a recession is universally defined as 2 Qs of negative growth and with stock markets the universally accepted definition of a crash is a 20% fall from the peak. However, oddly, there is no universal definition of a housing crash which hampers us in discussing the situation in Ireland allowing people to variously claim that what we are experiencing is just a slowdown, a shallow correction or an outright crash. Take your pick!

The main reason for this the lack of true property crashes in the US. The one and only true crash was in the Great Depression when nominal values fell 25% over 1929-33. However, there is a lot of regional market data on various crashes over the last 50 years. The present slump in the US sees housing values down close to 5% from their October 2005 peak nationally. However, there are a number of regional variations with large falls circa. 15% in States such as Florida and Louisana.

However I have come across 2 research documents on this area that come to slightly different conclusions.

First of all there is one common theme that both sets of reserach come to and that is that when discussing housing values and the base from which to calculate whether a crash is taking place or not you must use nominal house values rather than the real inflation adjusted values. The rationale for this, and quite reasonably in my view, is that no one who discusses housing values adjusts the figures for inflation but thinks just in terms of nominal values.

Where the sets of research disagree is in the figure that they define constitutes a housing crash. One states boldly that a 15% fall from the peak constitutes a crash. However, a second more indepth and thoughtout research paper suggests that a 10% fall represents a crash. They present a very good logical rationale for the 10% figure. Essentially from there research in which they studied every regional housing slump in the US since the collection of data going back to the 1920's is that when a housing market falls by 10% from its peak it enters a long-term slump that can take years to come out of. They argue that a 10% fall is a tipping point of no return for a housing market for the foreseeable future. The logic being that a 10% fall cripples confidence leading to further falls and a long-term adversion to property investment and purchasing among the public.

This is of interest to us here in Ireland as one Construction Industry economist came out during the week of the election (and got no coverage as a result) that housing would fall 5% this year and a further 10% in 2008. This if accurate means that the Irish market will enter crash territory at some point in 2008 at which point it will remain in crash territory for a number of years! How deep that crash will go is anybody's guess?

[Eirenua]

Interesting, and it certainly seems to be the trend, with a slowdown in house construction. I cant see prices dropping by 15% then again I'm no George Lee.

[constitutionus]

well the houses across the road from me are down 50k than what they were going for 8 months ago and thats a 15% drop so who knows !

[Sidewinder]

I'm not sure about the logic there. Just because the herd don't have the wit to distinguish between real and nominal values in the housing market, professional economists should just forget about doing their job properly? Meh.

If house prices had ground to a halt, nominally, in May 2006, and then stayed at those nominal levels for the next 10 years with inflation running at 4-5% p.a., then those houses are really worth half what was paid for them, whether the eejit who bought them thinks they are worth the same or not.

[Coles]

Interesting stuff Kerrynorth... I'm inclined to agree with Sidewinder that 'real'/inflation adjusted house prices give a more accurate picture, but then again I suppose an argument could be made that other issues should also be taken into account (wage growth, affordability etc...).

Daftwatch has reached a plateau, which I believe can be explained as the end of the selling season. The Autumn could well be a blood bath, with 2 more rate rises before Christmas.

[powderfinger]

Would also depend on the level and grade (prime or sub prime }of debt leveraged.For example,With 50% of African American mortgages in the U.S. being in the sub prime,adjustable rate market a downturn in the property market of 10% would lead to repossesions for up to 20% of that population.As Warren Buffet said ,"It's only when the tide is out that you can see who's swimming naked."

[kerrynorth]

I'm not sure about the logic there. Just because the herd don't have the wit to distinguish between real and nominal values in the housing market, professional economists should just forget about doing their job properly? Meh.

The logic rests in psychology really. 99% of people will not think of housing values in real inflation adjusted terms. The research is just reflecting reality in this respect and if a 10% fall in nominal values is a consistent wall from their research at which the property market hits a crash then you have to just go with it!

If you look at it the other way around if high inflation leads to a corresponding increase in nominal values then the psychology of the property owners will be that they are increasingly asset rich leading to a wealth effect although in real terms they are in the same place.

[patslatt]

There is very little information out there in how to to define a housing crash as distinct from a slump. With the economy a recession is universally defined as 2 Qs of negative growth and with stock markets the universally accepted definition of a crash is a 20% fall from the peak. However, oddly, there is no universal definition of a housing crash which hampers us in discussing the situation in Ireland allowing people to variously claim that what we are experiencing is just a slowdown, a shallow correction or an outright crash. Take your pick!

The main reason for this the lack of true property crashes in the US. The one and only true crash was in the Great Depression when nominal values fell 25% over 1929-33. However, there is a lot of regional market data on various crashes over the last 50 years. The present slump in the US sees housing values down close to 5% from their October 2005 peak nationally. However, there are a number of regional variations with large falls circa. 15% in States such as Florida and Louisana.

However I have come across 2 research documents on this area that come to slightly different conclusions.

First of all there is one common theme that both sets of reserach come to and that is that when discussing housing values and the base from which to calculate whether a crash is taking place or not you must use nominal house values rather than the real inflation adjusted values. The rationale for this, and quite reasonably in my view, is that no one who discusses housing values adjusts the figures for inflation but thinks just in terms of nominal values.

Where the sets of research disagree is in the figure that they define constitutes a housing crash. One states boldly that a 15% fall from the peak constitutes a crash. However, a second more indepth and thoughtout research paper suggests that a 10% fall represents a crash. They present a very good logical rationale for the 10% figure. Essentially from there research in which they studied every regional housing slump in the US since the collection of data going back to the 1920's is that when a housing market falls by 10% from its peak it enters a long-term slump that can take years to come out of. They argue that a 10% fall is a tipping point of no return for a housing market for the foreseeable future. The logic being that a 10% fall cripples confidence leading to further falls and a long-term adversion to property investment and purchasing among the public.

This is of interest to us here in Ireland as one Construction Industry economist came out during the week of the election (and got no coverage as a result) that housing would fall 5% this year and a further 10% in 2008. This if accurate means that the Irish market will enter crash territory at some point in 2008 at which point it will remain in crash territory for a number of years! How deep that crash will go is anybody's guess?

The price comparisons are usually distorted at major turning points in housing markets. In a sudden upturn in prices and volumes,the average quality of houses for sale will be less than it was before the upturn as lower quality houses sell more strongly. The opposite is through in a slump in house prices,as the quality of houses sold improves on average. If house prices drop by,say,5%, one needs to add an estimated percentage to that for quality improvement, which depends mostly on the drop in sales volumes. If volumes drop 20%, that figure can be estimated at 5 to 10% maybe,meaning a quality adjusted price drop of 5% + (5 to 10%) = 10% to 15%.

On the other hand, in Dublin the bubble in the most expensive houses worth millions was the first to start deflating and sales dropped.

The best price comparisons are offered by sales of highly standardised housing units in modern subdivisions where the comparisons are like with like.

[CookieMonster]

A housing crash is whatever the Sunday Independent say it is.