Credit Scoring 101
By Gregg A. Weldon
The use of consumer credit in American society has grown exponentially over the past 50 years. What was once considered a privilege is quickly becoming a necessity in today’s technologically driven environment. Credit cards are rapidly replacing cash as the preferred manner of payment for goods and services, car prices have made purchasing new vehicles for cash an option for only the truly well-off, and home ownership (and mortgage debt) have reached an all-time high. Despite this, many of the tools critically needed to ensure a smooth and efficient credit market are only now becoming familiar to the majority of the public.
I. THE CREDIT PROCESS
Creditors, whether they be banks, auto dealers, credit card companies, mortgage lenders, etc., are faced with the difficult task of predicting who will pay them back and who will not. These creditors are faced with two potential errors in their judgement: lending money to someone who won’t pay them back, and not lending money to someone who would have paid them back. As we will see, both errors lead to some serious long-term consequences.
Traditionally, a potential borrower would go to his neighborhood bank branch and make an application for a loan. The application would ask basic questions of who this person was, his educational background, job history, where he lives and how long he has lived there, etc.. It also helped if the borrower kept his accounts at that bank. This would be followed by a short interview with a loan officer.
Based on this information, the loan officer would make a judgmental decision as to whether the applicant was a good credit risk or a bad credit risk. This was based on the 4 “C’s” of lending. “Credit” is how well the applicant paid previous loans or credit obligations. A person who had three previous cars repossessed by the bank will probably not pay their newest loan back in a sterling manner either. “Capacity” is an applicant’s ability to repay a loan. A person making $15,000 a year could not afford the monthly payments on a new Mercedes. It does no one any good to get a borrower in debt over his head. This can ruin the borrower’s credit as well as create an expensive loss, or charge-off, for the lender. “Collateral” is the item (or items) the borrower will pledge to the lender in exchange for credit. If the borrower defaults, the lender gets to take the collateral and sell it to recoup his loss. The potential loss of collateral is leverage against a borrower who may otherwise not be concerned with the lender’s well being. “Character” may be the hardest aspect to judge and yet the most important. The lender is faced with deciding whether the applicant is the type of person who takes his obligations seriously enough to pay back their debts, whatever may happen.
A lender makes these decisions based on years of training and experience. A drawback to this method is that it takes a long time and a lot of money to train loan officers so that they can meet with the public on a daily basis and make good, consistent decisions. Also, face-to-face interviews with applicants is a very time-consuming way to conduct business, better suited to small towns than larger cities with more transitional populations. Another key drawback to judgmental decisioning is the potential for bias. A lender may approve or deny based on such factors as race, gender, religion, or any other illegal reason. Most often, however, bias takes the form of turning down applicants the lender “doesn’t like”, approving applicants who may not meet credit criteria but are “nice guys”, and turning down someone who applies late Friday afternoon to avoid the paperwork.
In order to save time and assist lenders in making good decisions, the use of national credit bureau reports has become standard in the credit industry. The three major credit bureaus, Equifax, TransUnion, and Experian, receive payment performance information from all types of creditors nationwide. This information is compiled into a credit bureau report for each individual. When an applicant applies for credit, a lender can very quickly see how this person has paid all of their other debts. Over the years, as credit bureaus have become more accurate, cheaper, and more accessible, they have supplanted applications and interviews as the predominant tool for decision-making.
In the 1950’s, when national credit card companies began to solicit their services, they began to receive daily applications by the thousands rather than by the dozen. Clearly, some form of automation was needed.
II. THE ADVENT OF CREDIT SCORING
Credit scoring came into being to combat the problems with judgmental lending. The introduction of the computer into the business environment, with its ability to process data quickly, consistently, and relatively cheaply, made the development of credit scorecards inevitable. Credit scorecards are exactly what they sound like: a scorecard that assigns points for various credit factors. Positive points are given for “good” behavior and negative points are given for “bad” behavior. The fact that these scorecards look at exactly the same things that lenders do judgmentally made scoring very popular in the credit industry.
Below is a simple example of a credit scorecard:
| Variable | Range | Points |
| Constant | 500 | |
| Years on the job | 0 to 1 year | -50 |
| 2 to 3 years | 0 | |
| 4 to 7 years | 35 | |
| 8 years of more | 75 | |
| # of charged-off loans | none | 0 |
| 1 to 2 years | -60 | |
| 3 to 4 years | -100 | |
| 5 years or more | -280 | |
| # of loans paid on time | 0 to 2 years | 0 |
| 3 to 8 years | 25 | |
| 9 to14 years | 75 | |
| 15 years or more | 120 |
There are only three variables (plus the constant that all applicants get) in the scorecard. An applicant who has been on the job for 8 or more years, has no charged-off loans over time, and has 15 or more loans/credit cards that he pays on time each month will get the maximum number of points, 695. The lowest possible score in this example is 170. The higher the score, the greater the likelihood that the applicant would pay back his loan as agreed. Creditors would look at the score for each applicant and make a decision as to whether or not to approve the loan.
The obvious question, based on the above example, is “At what score is someone turned down for credit?” The answer is really based a many factors, but it comes down to how the individual applicant scores compared to the rest of the lender’s customers. For example, assume that two lenders, Lender A and Lender B, both use this scorecard. Lender A is a conservative, risk-averse bank whose customers average a score of 580. He has a delinquency rate (the percentage of borrowers who are not paying as agreed) of 2%. Lender B is a small, new bank, trying to attract customers with looser credit criteria and lower interest rates. Lender B has an average portfolio score of 520 and a 6% delinquency rate. An applicant who scores 475 may have better luck being approved by Lender B, who is more willing to lend to relatively riskier applicants. Lender A and lender B are both making a loan decision based on the applicant’s job stability and credit history. By translating the applicant’s history into a specific numeric score, the decision is made quickly, dispassionately, and consistently, based on the goals of the individual lender.
III. ADVANTAGES TO CREDIT SCORING
The advantages to credit scoring have proven to be enormous over time, benefiting lenders, consumers, and the economy itself. Credit scoring removes bias from the equation. The scorecard (and computer that uses it) doesn’t know the race, gender, marital status, religion, or anything else about the applicant that is not specifically in the scorecard. The scorecard is consistent, giving the same relative weights to each variable, no matter what time of the day or night the application is submitted. The score can be calculated in seconds over a modem, allowing the lender to convey the decision to the applicant immediately.
The lender has more control of credit policy when scoring is utilized. In a judgmental environment, a creditor may tell his lenders to “be a little more cautious” in making loans. To 30 lenders, this may mean 30 different things. With credit scoring, a creditor merely has to change the “cut-off”, the score at which he will approve a loan. In the above example, if Lender B, faced with the 6% delinquency rate, decides that collection costs are rising too fast, he may raise the minimum score needed to be approved from 475 to 500. Again, this change in policy is specific, immediate, and across-the-board.
By setting loan policies and cut-off scores this way, the lender can review results from these policies easier. For example, if Lender A finds that people scoring from 500 to 540 have a higher delinquency rate than they are comfortable with, but not high enough to decline credit, Lender A may want to charge applicants in this score range a higher interest rate to compensate for the higher risk. Other options include asking for more collateral, giving shorter loan terms, or requiring a larger downpayment.
Credit scoring is good for the lender in that his costs of training personnel is lower and delinquencies decline because credit scores are consistently able to more accurately predict performance than judgmental decisions. Also, fewer people are needed for the credit process, allowing lenders to branch out into other areas with new products and services. Some examples are “first time buyer” programs, student loans, and home equity loans.
Earlier in this article, the two types of errors with lending were discussed. These errors will always exist, no matter how accurate a scorecard may be. However, credit scoring is able to minimize these errors much better than judgmental lending. Credit scoring can better identify those applicants who should never have been loaned money as well as those rejected applicants who would have paid as agreed. This “swap set” is the critical factor in a creditor converting over to credit scoring. The swap set represents a lucrative opportunity for any lender. Let’s assume that a credit score can identify 50 people who a judgmental lender wanted to loan to but who will become delinquent and 50 people who the judgmental lender wanted to turn down but who would have paid well. The scorecard would have not just transferred funds from one potential borrower to another, but would have saved money as well. The money that would have used for collection purposes (repossession, foreclosure, charge-off, etc.) can now be used for additional loans to other borrowers. Credit scoring, then, not only saves money by being more accurate, it also allows the lender to expand his loan volume.
Other beneficiaries of credit scoring are the stockholders of the lending institution, the applicants who should have received the loan but, under judgmental lending, did not, federal regulators, who are able to more accurately police the credit industry, and the taxpayers, who should have fewer banks to bail out in times of financial crisis. The losers when credit scoring takes place are those borrowers who do not take their credit obligations seriously.
IV. DISADVANTAGES TO CREDIT SCORING
Credit scoring is not the Holy Grail of the credit industry. It has a number of disadvantages that have yet to be overcome. First and foremost, credit scoring usually only measures an applicant’s willingnessto repay, not his ability. A person who believes in always paying his debts but is then laid off from his job for 18 months may have a credit history that confuses ability with willingness to repay. Most lenders utilizing credit scoring are prompted to override the score approximately 5% to 10% of the time in order to compensate for this.
Second, credit scores are only as good as the data used to build them (garbage in, garbage out). Scoring models built on erroneous or skewed data will give unrealistic results. Scorecards should be validated after development and prior to implementation to make sure that they are working as they were intended. Likewise, ongoing monitoring is essential in making sure that the models continue to perform.
V. CONCLUSION
Credit scoring will continue to grow as a credit tool as computer systems become more sophisticated and the credit industry witnesses more competition for consumers’ business. Although created primarily for the processing of credit card applications in the 1950’s, credit scoring is now standard in all areas of new applicant risk lending. In addition, scorecards are used for response models (which consumers will respond to junk mail in their mailboxes), account management (increasing credit limits for existing customers), collections models (which delinquent customers should be called vs. mailed), and even insurance models (which customers are most likely to file a claim). Although viewed with trepidation by some, scorecards have proven their worth time and again, and benefit lenders as well as consumers.
Gregg Weldon is the Chief Analytical Officer of AnalyticsIQ, Inc.