First impressions, as we all know, are what count. When people are introduced to someone new, they quickly pigeonhole them by looking at their dress sense, listening to their accent, and watching their mannerisms. This is then refined through a series of what are, on the surface, purely incidental questions such as ‘What do you do?’or ‘Where do you live?’.
This is all well and good in general, but what if this person is looking for someone to lend them money? Is the way someone looks a fair way to judge whether or not they are a good risk? Probably not. Mortgage lenders understand this and use a variety of techniques to work out if they are the sort of person they would be prepared to lend money to.
Approximately half of the mortgage industry now uses credit scoring techniques to quickly assess whether an applicant is fit to pass its lending criteria. The other half, however, still uses manual underwriting procedures to assess applicants. This is often used by smaller lenders, as it can be the most cost-effective tool for organisations whose volume of business does not warrant the high cost of developing credit scorecards. These lenders carry out their own, manual assessment of the mortgage application based on data that the borrower is asked to provide ‘ usually bank statements, payslips and employer references.
However, forthcoming changes to the regulation of mortgage providers could have major implications for this. At the moment, lenders put aside an amount of capital to offset the risk of the loans on their books. But, under the European ruling proposed by the Basel 2 Capital Accord, the way that this amount of capital is calculated will change from January 2004. From that time, it will be necessary to collect more extensive and stringent data on borrowers to be able to obtain a more accurate assessment of the risk involved.
With this increased accuracy comes the benefit of a reduction in the amount of capital reserves required, which in turn will free up money that may be used to offer even lower mortgage rates. However, another result is that, in less than 12 months time, nearly all lenders will be forced to use credit scoring. One of the implications for brokers is that they will find it harder to get the ‘less desirable’ borrowers through the system, even if theirs is a legitimate case.
If, as predicted, credit scoring is inevitable, it is well worth brokers understanding in more detail how it works in practice, to be able to best advise clients and help make their mortgage applications run as smoothly as possible.
So what is credit scoring and how does it work? At its most basic level credit scoring is a technique used by financial institutions to help them assess the level of risk involved in extending credit facilities to someone. It is consistent and speeds up the application process. Based on this calculation (the credit score) lenders will decide whether to approve the mortgage loan. Every application involves a certain level of repayment risk for the lender, no matter how reliable or responsible the person is. Calculating a credit score in each case allows the lender to assess every application they receive fairly, using the same criteria.
However, criteria change from lender to lender, depending on its lending policies and scoring systems. This means that one lender may accept an application but another may not.
The credit score is calculated from a mixture of the information provided on the credit application form, data provided by a credit reference agency and details of any existing relationship the applicant may have with the lender.
So how does credit scoring work? When a lender develops a scorecard it needs to decide what information it would like to assess. The questions it asks are known as ‘characteristics’ and the answers it obtains are known as ‘attributes’. Lenders analyse their existing borrower base to identify characteristics that are most likely to predict the future performance of accounts. For example, some lenders may find that the younger the customer, the greater the risk that they will go into arrears or default. If their existing borrower base shows this trait, it is considered likely that new applicants will follow suit, therefore ‘age’ is considered to be a very predictive characteristic in their application scorecard.
When a lender has identified the characteristics it would like to use, it needs to decide how many points each answer (or attribute) will be given. Again statistical analysis of its existing borrower base is used. Using the example above, the different attributes within the age characteristic would be banded into groups and points allocated to each band. Again, as an example, where a lender has proven that younger applicants are a greater risk, statistically they are allocated a lower score than older applicants. For an age band range of 18’24, this would typically give a score of 5, for 25’34: 12, for 35’44: 16, for 45’54: 20, and for those aged 55 and over: 22.
Once a score has been calculated for each characteristic, they are added together to calculate a total credit score for that applicant. This is then compared to a ‘cut-off’ score, which is set by the lender and reviewed periodically to reflect the level of risk the lender is prepared to take at that point in time. Its targets, desire for market share and economic conditions all influence its cut-off.
If the total credit score is above the cut-off, the application is likely to be approved, whereas if the score falls below the cut off it is likely to be declined. Applications that fall around the cut-off score may be referred to an underwriting team for review. The decision will probably be delayed for a few days while they reach a decision using all the information available to them.
The characteristics used, scores given and cut- off score are dependent upon the type of business the lender wishes to accept. For example, those lending on high income multiples use different score cards to lenders who specialise in sub-prime lending.
The role of the credit reference agencies is also very important. The data they hold typically includes:
• Electoral roll information.
• Public information such as County Court Judgments (CCJ’s) and bankruptcies.
• Payment history of other credit commitments (such as mortgages, loans and credit cards).
• The amount currently owed to other creditors.
• The number of recent credit applications made.
Most lenders find that the information available from a credit reference agency is very predictive and helps it to identify applicants who are already over-committed or who have current or historical payment problems. However, new rules will allow people to opt out of credit reference agencies’ and lenders’ rights to the automatic use of their financial partner’s data, allowing them to be assessed in their own right.
However, most lenders would say that the majority of complaints about credit reference agencies stem from people who are starting out on the credit ladder, meaning the credit reference agencies do not have any information on them.
Michelle Hodgson, head of consumer affairs at Leeds-based credit reference agency, Callcredit, a subsidiary of Skipton Building Society, says: ‘When a lender assesses an application, it does not just check the files of a credit reference agency for any negative information, such as CCJs or defaults, it is also looking for positive confirmation that the applicant manages their existing credit well and pays their debts back on time. This gives some assurance that they will also pay them back. If an applicant does not have any existing credit agreements, it is difficult for the lender to get the assurance it is looking for and it may decline the application. Of course not every lender adopts this approach. If they did, then no one would ever be accepted for credit the first time they applied for it.
‘There are steps people can take to get on the first rung of the credit ladder. Most people have a current account that is credited with their salary or wages on a regular basis. The bank or building society they are already with is therefore more likely to offer them further credit facilities than a lender they have never been to before.’
However, the use of credit scoring is not without its pitfalls and lenders must now adhere to the Guide to Credit Scoring issued in 2000 by the Office of Fair Trading. This requires lenders using credit scoring to offer an appeals process. When receiving an appeal the lender must conduct a realistic review of the decision, taking all available information into account. This is an opportunity for the broker and/or applicant to provide the lender with reasons why the application should in fact be accepted.
Looking ahead, more and more companies will be turning to credit scoring to help in their lending decisions, with the result that, although manual assessment may virtually disappear, consistency will be applied and the application process will be speeded up.
Credit scoring is set to become more prevalent as more data becomes available to lenders under Basel 2 Accord proposals.
Credit scoring does provide an indication of the risk involved but will not cater for individual circumstances.
Without a credit history, getting credit can be a problem.