Andrew Weller of Equifax Predictive Sciences explores the new landscape of credit scoring in the mortgage industry as borrower affordability becomes even more of an issue for lenders
At the end of January we heard that the
Of course, we can learn lessons from the last couple of years. The Equifax Markets report, which looks at payment performance and arrears trends in a number of key consumer markets, provides a very useful insight into behaviours that can provide pointers for the future. For example, fin the fourth quarter of 2009, mortgage arrear rates for both three and six months behind on payments certainly levelled off, which is good news after the dramatic and rapid rise in the number of home owners unable to pay in 2008. But with continued pressure on households from an increasing unemployment trend and a base rate that can really only go in one direction, arrears in this sector could still have a pretty significant impact in 2010.
We also can’t ignore the proposed Mortgage Market Review, with lenders expected to place more emphasis on affordability in order to be seen to be treating customers fairly.
On both fronts, there is clearly a fundamental role for credit scoring. But are today’s credit scores up to the huge challenge that the sector now faces?
Only as a good as the data
A credit score is only as reliable as the data used to build it, and the performance it predicts. And the challenge now is to ensure that any credit score uses the right data for the job it needs to do. So what should mortgage lenders be looking for in the data sources for their credit scores?
Recent events in the credit market suggest that more details than are currently gathered are required to ensure the right decision is reached in as many cases as possible.
Which is better – experience or youth?
Historical data has long been regarded as vital to building an accurate credit score and therefore making the best decision regarding an application for credit. Certainly stability of data is extremely important when examining the reliability of scores through time.
But many consumer behaviours have changed significantly over the last year to 18 months and this could change the effectiveness of credit scores that are built on historical credit data alone.
Credit scores are now being built on even larger samples and with more recent data. Yet this can be a challenge when looking at the mortgage arena where volumes are lower and monitoring performance on accounts requires longer periods of time.
For both existing and newly developed credit scores, suitable monitoring should be put in place to not only inform lenders of changes, but also to provide information to enable companies to act with speed and confidence where changes do take place. A combination of ongoing monitoring and utilisation of accurate data will benefit lenders significantly by showing clearly which prospects should be avoided, and which are worth the slight risk.
What’s a bad risk?
But that does raise the question of what is a bad risk.
What has also had an impact on the effectiveness of scorecards over the last couple of years is the introduction of new types of credit data. For example, the advent of new data from telecommunications companies and current accounts has changed the credit profile of many thousands of consumers – but that changing profile doesn’t necessarily mean an increased risk, particularly to mortgage lenders.
These new data sources generally represent smaller value credit. So, the addition of this data to existing consumer credit files means that there are now many consumers in the
This could potentially mean that the consumer wasn’t necessarily treated fairly – and the lender missed out on a new customer opportunity. And that brings us onto the subject of regulation. Indeed, it’s hard to think about credit scoring in the mortgage sector without making reference to the Turner review.
Recession, review, regulation
In any time of economic stress, difficulty or indeed recession, new regulations are brought in to address the original cause of the problem. In the current market, mortgage lending appears to be shouldering much of the blame and it is therefore to be expected that regulation will focus on this industry.
Changes have already been occurring over the past two years throughout the industry when it comes to lending decisions. The checks undergone and criteria customers must meet are now much more stringent than prior to the credit crisis. But a potential consequence of the downturn and subsequent regulatory reviews is that subsets of the population could end up being excluded from potential mortgage offers despite being able to repay commitments.
For every individual taking a certain secured loan with set attributes for example age, income, employment status, expenditure, etc, that subsequently went into arrears, there could be another 10 individuals with the same profile who repaid without any problems. It is therefore clear that a detailed, accurate and comprehensive report must be created for each application to allow the resulting scores and decisions to correctly reflect their suitability for a loan, and the risk they pose to the lender.
More data
Without access to more data sources and verification, we may well find ourselves struggling again.
Data and circumstances currently taken into account in the decision making process include income tax, secured and unsecured loan repayments, council tax, utility bills, other household bills, child costs, regional variations, LTV ratios, credit score and other expenditure measures listed in numerous documents on the subject.
However, even this substantial list is not granular enough; we need to think about savings as well. It would also be helpful to break down the above into individual costs, rather than, for example, grouping together Sky TV, gas, electricity, water, telephone line, mobile phone, service charges on flats and many more, into one ‘utilities’ bracket.
Affordability to arrears
Verification of expenditure is arguably as important as verifying an individual’s income, as consumers generally underestimate their expenditure, and therefore misjudge what is affordable. And with mortgage availability predicted to continue rising, although the economy remains less than certain, it is reasonable to expect further increases in arrears due to changes in what is and isn’t affordable for consumers.
As a result, lenders will have to start placing more emphasis on affordability calculations when underwriting new loans. This has particular importance on certain types of products or to certain consumer profiles. And this might change the way we consider credit scoring for secured lending in general.
For example, having carried out a comprehensive affordability check that includes verified gross income, income tax, national insurance, council tax, utility bills, unsecured debt commitments, secured debt commitments, pension contributions, child care, food and drink, transport, etc, to arrive at a consumer’s free disposable income, how much additional power would a credit risk score provide in predicting arrears?
It is, therefore, important to understand the place a credit score is likely to have in the lending process going forward. The Mortgage Market Review places a lot of emphasis on calculating a consumer’s disposable income, and verifying information provided. The challenge here is not only that consumers may underestimate their expenditure commitments, but also the possibility of their providing erroneous information, or life events meaning situations change quite rapidly.
For many individuals a disposable income calculation could be subject to much change even over the course of one year. Indeed, experience from credit scoring through the last two decades shows that whilst recent information tends to contribute highly to models, information that is older still adds a huge amount of value. The older information often provides a more accurate picture of a consumer’s attitude to debt through time.
So another question then arises in relation to the Mortgage Market Review. When considering affordability assessments should we not be looking at longer running averages both in terms of income and expenditure for a more accurate picture of the individual and what is affordable for them?
Wider data sharing
Income must be verified and I believe verification of any information provided by the consumer will play a bigger part than ever in the future of risk assessment and reducing fraud. And in the area of income verification and automated verification, the question arises as to where source data is held that would enable automated checks of both income and expenditure to take place.
The credit industry is used to sharing data, but in light of the Mortgage Market Review, it might be that additional data sharing user groups are created. For credit scoring purposes it could also be necessary to share more information for the development of new scorecards.
In addition, the Credit Reference Agencies must be able to start sourcing other data in order to provide the most accurate intelligence for credit decisions. This additional data would allow Credit Reference Agencies to verify information provided by a customer much more easily, thereby allowing more instant, suitable decisions to be made on an automated basis. And it could have immense benefit in supporting the wide variety of products in the market.
For example, the information shared could be aggregated, pooled into homogeneous groups and, of course, anonymised to protect confidentiality. This, in turn, could enable specific scores to be created, such as for buy-to-let mortgages, remortgages, first-time mortgages, interest only mortgages, and equity release mortgages. Then, taking this a step further, additional splits could be made to address different types of consumer groups – the credit impaired, self employed, etc.
Change is good
I firmly believe credit scoring will play a big part in the origination of new secured lending, especially when thinking about consumers and particular behaviour patterns over time. Any affordability assessment will have to include credit scoring.
More data will need to be used in assessment and the focus on verification of consumer supplied information will increase. Lenders must be mindful of not making assessments purely using recent data, but should also employ ongoing monitoring solutions to stay on top of changing conditions. The individual’s credit history must also be taken into account as data going back in time can help explain behaviour patterns and give a more accurate impression of ability to pay, affordability and the risk involved in lending to this customer.
Historical data can also minimise any potential unnecessary exclusions that could result from increased regulation, thereby increasing the number of customers and in turn, profit margins.
But the most important factor is flexibility, and constant ongoing development. As the economic landscape changes, which of course it never ceases to do, credit scoring and mortgage decision-making must adapt to suit market conditions. Credit scoring techniques and the data examined twenty years ago would be extremely inaccurate for today’s society. In the way of technology and fashion, credit scoring must be ever changing to suit its environment, taking on new sources, new expertise and new techniques.
Andrew Weller is director at Equifax Predictive Sciences
Date: 5th, March, 2010
ADVICE TO READERS
While this website is checked for accuracy, we are not liable for any incorrect information included. We recommend that you make enquiries based on your own circumstances and, if necessary, take professional advice before entering into transactions.
Enter your postcode here to find out how much your property is worth, based on Land Registry data.
RSS Feed
Delicious
Digg
reddit
Facebook
StumbleUpon