Share
We need a more accurate measure of house price expectations based on an agreed set of official indices as these are a key part of any investment process and relying on a myriad of surveys is risky
In my opinion
GARY STYLES STRATEGY, RISK AND ECONOMICS DIRECTOR HOMETRACK
The recent wobble in house price and consumer confidence figures has put stress testing back into sharp focus. Just as many were preparing for the gradual return to 2007 house price peak levels, perhaps we should be revisiting the stress tests we are applying.
The standard stress test produced by the Financial Services Authority is only applicable at the highest level of aggregation and is more about the overall move in the average house price than the extreme individual price movements experienced in even the most modest of downturns.
The first table below compares the two standard FSA scenarios that applied both this year and last year.
The scenarios are broadly similar, with 2010 showing a more modest total house price fall, which is slightly offset by a higher unemployment peak of 13.3%.
But any scenario should also contain a detailed interest rate and spread forecast combined with a closer analysis of consumer and cost inflation.
The interest rate view is probably one of the most important assumptions for any lender and whether we are talking about a 5% base rate or 0.5%, it will have a meaningful impact on any stress test produced.
As with most scenarios, there are a variety of combinations of the main economic variables that can generate stress for banks and financial institutions and it is difficult to imagine producing these without an interest rate and margin assumption.
There are several other ways the standard stress could be enhanced to become more realistic and dynamic. The blanket assumption that all house prices fall by 36% is unrealistic as the experience will vary considerably by property type, location and socio-demographic classification. If we use the 2008/09 experience as a benchmark, we can see that house prices fell by around 18% peak to trough, but that this fall varied from 7% to 33% for around 95% of UK properties.
One of my biggest concerns at the moment is getting accurate and appropriate information to the market
Using a similar spread for the FSA stress test would imply that price falls would vary from around 13% to 49% - i.e. around the mean of -36%.
I would argue that the implications of a fall in house value of 13% would be somewhat different from a decline of 49%. I think it is safe to assume that the risk ratings for these individual properties will also be different. The spread in this asset performance combined with variation in credit grade and concentration factors makes the scenario far more complex - and for this reason more realistic.
The current thin trading volumes and erratic moves in reported house prices look set to be with us for some time to come. The importance of having accurate current valuations on all assets has increased - the signals coming from the market may give a misleading impression as to the performance of the underlying assets on your particular balance sheet.
Regional, sectoral and socio-demographic variation will increase in the coming months as any economic recovery is unlikely to be uniform or indeed predictable at a regional and sectoral level.
One of my biggest concerns remains expectation formation and the importance of getting accurate and appropriate information to the market. As was seen in the second half of 2009, the reported large rises in house prices led to improved consumer confidence. But the increases in prices were far from uniform and looked in the main part like a mini-bounce back from the over-correction in the indices reported during 2008.
What we need is a more accurate measure of house price expectations based on an agreed set of official indices. These expectations are an essential part of any customer and investor investment process. Relying on the myriad of indices and surveys to set expectations is a high-risk strategy. For example, at the start of 2009 the Reuters survey of expectation for house prices was for prices to fall 14% largely based on the experience of 2008. But the actual outcome was that prices rose by 3% to 6%, depending on the indices used.
The need for a regular and consistent assessment of house price expectations by region and sector has never been greater and this should be of equal importance to policymakers, consumers and lenders. Anything we can do to flatten the extreme movements in expectations and reduce noise levels can only encourage a healthier and more stable market in the medium term.
