Woolly risk analysis is hastening a housing crash

Borrowers have moved from a mortgage feast to a famine so bad that the government has stepped in with a £50bn support package. But there's a better way to ease the situation, argues Steve Boyle.

CIO50 2008: Top 10

The UK's leading CIOs revealed…

1.Robin Dargue Royal Mail

2.David Lister Royal Bank of Scotland

3.Neil Cameron Unilever

4.Catherine Doran Network Rail

5.John Suffolk UK government

6.Gordon Lovell-Read Siemens UK

7.Paul Coby British Airways

8.Tania Howarth Birds Eye Iglo Group

9.Simon Post Carphone Warehouse

10.Ben Wishart Whitbread

Lending institutions need not stop lending. Nor do they have to rely on government handouts. They have only to ensure their risk assessment systems are sophisticated enough to provide real-time feedback on potential defaults.

Real-time data systems enable total flexibility in a lending institution's response to the at-risk portions of its mortgage portfolio and let it be proactive customer by customer. Such systems could be the difference between a housing market slowdown and a crash.

What started off as a liquidity crisis post Northern Rock is now in real danger of developing into a credit crisis, creating a significant downturn in lending.

Although the effects of the global sub-prime lending crisis were seen as likely to bite the UK housing market, there has been very little clear evidence of the depth and breadth of the problem - until now.

The Nationwide Building Society reported that UK house prices have recorded their largest monthly fall since 1991 and have fallen by 2.5 per cent during May.

Also, Bradford & Bingley has hit the news and now faces a shareholder revolt over its £400m bailout plan.

Standard & Poor gave the first glimpse of the looming problems by reporting "more than a fifth of UK homebuyers have a chequered credit history and have fallen behind on their mortgage payments". And even those who have portfolios with top-quality ratings have seen a statistically significant rise in delinquencies.

Mortgage borrowers coming off fixed-rate terms are now seeing their monthly interest payments jump by up to £400 per month. As little as a year ago they might have been able to remortgage - today they may well see their home repossessed.

With all this doom and gloom, a housing market slump seems inevitable and a crash even possible. But does it have to be this way?

The problem is that over-cautious and unnecessary risk management - in the face of much better information - runs the risk of fuelling a crisis by putting mortgages beyond the reach of first-time buyers. It also makes upward movement in the housing market too difficult, too expensive and more tellingly, undermines general confidence.

While there is no one single remedy for curing the risk-adversity of lenders at present, there are a number of measures that can be taken to help stabilise a lending institution's existing mortgage portfolio. And a stabilised, risk-averse portfolio would contribute toward conditions that would allow a lending institution to continue to lend with confidence. These include:

• Implementation of the optional, challenging retail internal ratings basis approach as part of the Basel II capital adequacy requirements. This approach requires lenders to stress-test their mortgage portfolios against certain economic, financial and business plan events and even takes into account human and behavioural shifts. The sub-prime crisis is exactly the sort of sustained profit shock built into the stress-test requirements. Whereas most lenders test their ability to withstand single shocks, such as a rise in interest rates, some have adopted multi-variable shock models to handle a combination of circumstances. Only very few retail lenders have adopted comprehensive and consistent stress-testing across their entire organisations.
• Re-evaluation of risk, based on housing market conditions: most traditional risk models were built when prices were buoyant with little prospect of a fall.
• Deep data-mining techniques can help to improve significantly the old statistical models of loss-forecasting - which is key to the appropriate pricing of risk in lending decisions. These techniques would also help to build greater insight through further categorisation of consumers and their likelihood of default, according to risk drivers and especially, new product types.
• Performance reporting of loans in their initial stages: performance data from the past three years indicates that defaults are manifesting themselves earlier than in the 1990s. This fact will help in the application of new qualitative judgements based on sound, risk-based assessment and evidenced by corresponding documentation.

Evidence-based risk management is a key principle of real-time data systems. The concept says that if you cannot provide evidence to support your actions, they are based on judgement alone. And judgement without access to the facts will not do because you are not getting the absolute truth.

Ultimately, the only way that an organisation can protect its success and its shareholders' interests is if they adopt a model based on 100 per cent truth and insist on being shown the evidence.

Lending organisations must be able to prove they have understood the risks and are actively managing them according to the agreed controls.

By applying greater sophistication to their risk management and lending decisions, banks and building societies can have the confidence to approve more loan applications.

This approach means going way beyond simple credit scoring and not loading interest rates for potential borrowers who are perhaps less risky than the traditional models implied.

There is no doubt that the influence of sub-prime mortgage lending will continue to feature strongly in UK lenders' mortgage portfolio for the foreseeable future.

But by displaying increasing sophistication in their risk management and lending decisions, lenders will play their part in preventing a crisis in the housing market.

Steve Boyle is chief executive officer at Sutherland Consulting.