Developing an Internal Credit Score for Businesses
Credit risk arises from the uncertainty of a borrowers failure to meet its financial obligations with a creditor. For the creditor, failure to assess and manage credit risk in an effective manner can result in significant financial loss. In this regard, it is very important that credit professionals are able to identify risk at the time of granting credit to a customer and that it is monitored and managed during the term of the relationship.
The recent development of risk evaluation methods due to advancement in technology provide credit professionals with more sophisticated tools that allow them to have easy access to more and better credit information. One of these tools is the credit scoring model.
Credit scoring models have proven to be a useful tool for banks and other financial institutions for many years, but to what extent can credit scoring models be useful for the commercial credit underwriting process?
Internal Credit Scoring Models
Internal credit scores help credit professionals determine creditworthiness of a potential customer. Credit professionals identify information and other factors from the customer that are important to them to perform a credit analysis. These variables are input into a computer model, which assigns them point values creating a scoring matrix. According to
1. Behavioral Models. These models require a significant pool of customers' records that are carefully analyzed to identify factors that can be used as predictive variables. With the use of statistical predictability, these models allow identifying the relevant tradeoffs or correlations among factors and assign statistically derived weights, which constitute the base of the scoring model.
2. Rules-Based Models. These models are based on traditional standards of credit analysis and they require a set of simplistic rules. Factors such as the business's payment history, bank and trade references, credit agency ratings (such as D&B rating), number of years in business and financial ratios are scored and weighted to produce an overall credit score.
The decision of which factors to use, and how each of these factors will be scored and weighted, is generally based on the credit executives past experience with their company, the products or services offered and the industry that the company is in.
3. Neural Models. Artificial Neural Networks (ANNs) or Neural Network Models are "distributed information-processing systems composed of many simple interconnected nodes inspired biologically by the neurons of the human brain" (Eletter, 2012).' These models employ "the parallel processing of math-based historical experiences and logic-based patterns".2 Angelini et al. (2008) "pointed out that ANNs have emerged effectively in credit scoring because of their ability to model non-linear relationships between a set of inputs and a set of outputs."3
The Pros and Cons of Credit Scoring Models
For commercial credit underwriting, the use of scoring models as a tool can be beneficial, but it can have a negative impact if it is overestimated over human judgment. Businesses often use scoring models as a platform for building an overall view of the customer, which will become a basis for forming a credit decision, like estab' lishing a payment term and amount of credit, frequency of credit monitoring, delegation of authority and approving credit limits.
Pros:
* Increase speed and reduce response time
* Quantify risk and improve management control
* Consistency, accuracy and objectivity
* Reduce personnel and credit investigation costs
* Decision support and planning tools
* Prioritization of collection activities and reduced bad debt loses
* Comply with audit mandates
* Knowledge transfer mechanism
* Ease of implementation
Cons:
* Qualitative factors and situation-specific judgment
* Dependency and lack of judgment
* Insufficient and/or heterogeneous pool of customers
* Obsolescence of historical information
* Statistical difficulties
Developing an Internal Credit Score: A Judgmental Scoring Model
A judgmental scoring model is a rules-based model that can easily be implemented because it can be developed using the company's credit policies and decision processes, and the grading scale can be developed using simple rules. In addition, it does not require the services of a costly third party to create and maintain the scoring model. The following key factors may be considered in the internal credit scoring model:
1. Traditional/Internal Credit Information. This is credit information that is non-financial data that the credit professionals would normally use when making a credit decision, such as:
* Pay history
* Trade references
* Control years
* Bank rating
* Country risk
* Collections claims and/or NSF payments reported
* Judgments and/or tax liens
* Plant capacity
* Annual purchases
* Collateral
* Competitive Index
2. Credit Agency Information. The outside ratings that the credit professional would normally consider in the credit decision process, such as:
* D&B rating/Paydex/Business Reports
* NACM score/Credit Managers' Index (CMI)
* Experian Intelliscore/Business Credit Advantage Report
*
3. Financial Statement Scores. In scoring a financial statement, the ratios should be scored through comparison to a published peer group, that are similar in the following ways:
* Same industry-as defined by the SIC code
* Same financial statement year
* Same size-as defined by size of sales or assets
Some credit professionals place a very high emphasis on cashflow statements and even suggest the need to compare cashflow ratios between companies from large to small, different industries, and all parts of the world. The following four ratios are commonly used to determine a firm's financial quality when compared to its peers:
* Liquidity ratios measure a firm's ability to meet its working capital and current obligations.
* Profitability ratios measure management's ability to generate positive returns from utilizing all available resources.
* Leverage & Coverage ratios measure the degree of protection for suppliers of long-term funds and can also aid in judging a firm's ability to raise additional debt and its capacity to pay its liabilities on time.
* Efficiency, activity or turnover ratios provide information about management's ability to control expenses and to earn a return on the resources committed to the business.
A judgmental scoring system requires proper identification of a benchmark by which scores for each rating category can be measured to determine the risk classification. Performance of peer companies within an industry can be utilized as a form of benchmark. For example, companies within the utility industry may have an average current ratio of 2-times and can be assigned a moderate risk. A current ratio exceeding 2 may indicate a low risk and anything lower than 2 may indicate a high risk. This methodology can be applied to all categories within the internal rating score. Historical trends can be used as a benchmark. Once a benchmark is established, the risk score for each category is determined and the overall score is derived by multiplying the scores obtained in each category by the weight assigned. The overall risk score is then compared to a scale that contains a rating from high risk to highest quality, to determine the customer's credit quality according to the internal scoring system.
Other Factors to Consider in the Credit Decision Process
In commercial lending, the decision to grant credit based on a credit scoring system alone may not consider other factors that may significantly impact the credit risk determination of accounts. Judgment should be exercised by considering other factors outlined below:
Country risk
Appetite for risk
Product status
Length of payment term
One-time sales vs. ongoing sales. (While one-time sales do not have the leverage of a new order to hold to get payment, they do offer a quick exit. The risk of a one-time sale has a lower risk of default than an ongoing relationship with a customer.)
Industry
Down payments or partial payments
Willingness to pay
Age of credit and financial data used for credit scoring
Potential of customer
Competitor's action
News reports, periodical articles, etc.
Weather patterns
Credit insurance
The experience of the credit professional is crucial for assessing and pondering these factors in the credit decision.
Conclusion
The benefits of utilizing a credit score cannot be ignored especially for a business with a high volume and low dollar value customers. But if the number of customers is low or heterogeneous, and the credit amounts are high, precision in credit decision is crucial because of the potential impact on the revenue and customer relationship. It is extremely important to have an experienced credit professional who would perform the credit underwriting process. There are several ways to integrate the scoring models into the credit underwriting process that will achieve significant benefits and reduce the cons of credit scoring.
Credit scoring models can be implemented in different stages of the credit underwriting process.
* Starting point for ratings assignment. Scoring models can be used as the initial step in the underwriting process. Human judgment can then be applied to modify the model's output according to considerations outside the model, perhaps with adjustments beyond a certain threshold requiring review from a credit committee.
* Second opinion. Scoring models could be applied at a late stage in the underwriting process, after a preliminary rating has been assigned by a human credit analyst. Large discrepancies between the model and the analysts rating could be referred to a credit committee for further review.
* Underwriting "traffic controller." Scoring models could be used at the front end of the process to group loans into three categories: automatic approvals, those that require a review by a human analyst and automatic fails. Human review might be reserved for larger loans and for segments where the scoring models perform less robustly."4 In this regard, scoring models should be viewed as a tool to help underwriters uniquely focus their human skills where they will have the greatest impact.
The decision to use credit scoring or not will be depend on customer composition and other circumstances unique to the company. The decision on which credit scoring methods to use, be it judgmental or statistical, will depend on the perceived cost and benefit.
The most important reason for using the credit scoring model is that it is the only effective tool by which credit risk is quantified. The ability to quantify credit risk allows management to define the amount of risk it is able to assume, to effect changes to the company's credit policy and procedures, to communicate any changes in the company's credit risk profile and to mitigate the credit risk. 1
For commercial credit underwriting, the use of scoring models as a tool can be beneficial, but it can have a negative impact if it is overestimated over human judgment.
Some credit professionals place a very high emphasis on cash-flow statements and even suggest the need to compare cash-flow ratios between companies from large to small, different industries, and all parts of the world.
In commercial lending, the decision to grant credit based on a credit scoring system alone may not consider other factors that may significantly impact the credit risk determination of accounts.
The above text is an abbreviated version of a
1. Bekhet, H. and Kamel Eletter,
2. Gahala, Ch., Credit Management Principles and Practices, NACM, 2013, p. 136/137
3. Bekhet, H. and Kamel Eletter, S. op. cit., p. 22
4. Jameson R., editor, op cit.
Allegheny County D.A. will retry arson case from 1995 when 3 firefighters were killed
Advisor News
Annuity News
Health/Employee Benefits News
Life Insurance News