1. Introduction
Account management is part of the overall account life cycle, which begins once an account has been approved and covers all aspects of managing an account. One or more of the following decision functions are typically found in most credit organisations:
• Authorisations
• Credit limit management
• Overlimit collections
• Delinquent collections
• Repricing
• Re-issue management
• Marketing communication
• Repeat business policies

For the purpose of this study, the credit limit management policy will be discussed using a study based on a revolving product within a retail credit organisation.

1.1    Credit Limit Management
1.2    
Application management typically combines the applicant’s affordability and application/ credit bureau score to assign credit limits according to groups of applications who display similar risk. Once an application is approved, it enters the account management life cycle. After a pre-defined period, the credit management policy will assess each account to decide whether it qualifies for credit limit changes. Each limit change is based on the individual’s behaviour in using and managing the account. Increases are generally granted to the more profitable, low risk accounts, and decreases are made to unprofitable or high-risk accounts.
The main objectives of credit limit management are:
• To generate customer satisfaction, by increasing limits of profitable accounts that require additional credit.
• To maximise profitability through increased spend resulting from the limit increase.
• To minimise bad debt exposure by not allowing unaffordable credit Limits

In practice credit limit management is controlled by policies that attempt to deliver the organisations objectives. These policies are periodically reviewed in order to identify further opportunities to increase revenues and reduce costs. These reviews generally take the form of tests.

2. Creating Tests To Beat The Existing Policy
The existing policies that currently operate on the portfolio have proven ability in producing a profitable mix. Each new policy that is tested will try to produce a more profitable portfolio, but testing various policies in unknown areas could have a negative impact on the overall profitability of the organisation. For this reason multiple separate tests will be run, over a defined period, using a small, representative sample from the portfolio. This ensures that any possible negative impact is kept to a minimum, but at the same time the business remains competitive and profitable by testing revised policies.

Each test differs from the existing policy by one decision criteria. This ensures that once the results are known, the factor that contributed to the increase or decrease in profitability is known. In this study the risk predictor used was a behaviour score.

3. Adjusting the Credit Limit Policy For Low Scoring Accounts
3.1 Objective of the Test
The main objective for the test applied in this study was to increase customer spend by granting increases to the lower behaviour scoring ranges.

Many retailers are testing to accurately define the highest risk level at which profitable credit limit increases can be made. It is common that the lower the behaviour score, the more credit hungry and higher utilised a customer is. In this area, there is an identified need for more credit, but increases may push the customer beyond what can be afforded, resulting in the account- becoming delinquent.

Each organisation will constantly try to identify the score break at which it is no longer beneficial to grant customers credit limit increases. This study was designed to test granting credit limit increases to customers who due to their lower behaviour scores did not previously qualify for an increase in limit.

3.2 Definition of the New Test
The following credit limit policies control how limits are increased or decreased in the existing credit limit policy:
• Accounts that have been on the books longer than six months may qualify every six months for further increases
• Accounts will qualify if they have a behaviour score where the odds of repayment are acceptable
• Accounts must be up to date with payments to qualify for a credit limit change

The test contained the identical policies except that accounts with a behaviour score that is below the existing cut-off also qualified for a credit limit increase.

3.3 Measurements
The results were cumulatively recorded for each quarter and compared against a sample representing the existing credit policy.

3.4 Time Frame
Many organisations do not take into account seasonality when measuring the success of tests. Seasonality can cause peaks in the income recorded over holiday periods as well as in delinquency over January and February. This test was measured using profitability indicators over a twelve–month period allowing the test to mature and meaningful results to be gathered.

3.5 Results
The results have been divided into four reporting periods showing the status of
the two samples after each period. The results were graphed using a bar chart,
which plots the percentage difference between the two samples.

Each graph plots three key measures based on profitability:
• Cost
All costs associated with managing the account such as account management, marketing, and delinquency and write-off costs
• Income
Income, made up of revenue generated by interest and fees, combined with the sales margin of the goods purchased
• Cost: Income Ratio
Account profitability is expressed as a ratio of cost incurred to income generated. This is calculated by dividing the cost value by the income value to calculate a ratio. A value greater than 1, indicates that for every one Rand generated, more than one Rand cost was incurred creating an unprofitable situation.

An alternate to the cost: income ratio would have been to measure the absolute profit value calculated by subtracting the cost from the income.

This graph shows the difference in performance between the two strategies. The vertical (y-axis) axis displays positive and negative numbers, depending on whether the measures compared better or worse against the existing policy.
• If the income measurement is positive, it shows that the test is generating additional income.
• If the cost is lower it reflects that the test is outperforming the existing policy by reducing cost
• When combining the cost: income ratio, if the relative difference is negative it reflects that the test is more profitable.

Each quarter will measure the cumulative results from when the test was put in place i.e. over 3, 6, 9, and 12 months

3.5.1 Quarter One Results
The cumulative results as at the end of the first quarter were calculated using profitability reports. The results were extremely promising, as the new test out-performed the existing policy on each measure.

The results indicated that the new test:
• Was generating substantially lower costs than the existing policy.
• Increased income marginally by 0.66%
• Was more profitable with an improvement in the cost: income ratio of 5.3%

3.5.2 Quarter Two Results
The test was then measured using the cumulative results over the last six months. More income was generated due to further increased spend. The accounts are costing the organisation less than the existing policy with a resulting cost: income ratio just under a percent, indicating that the test has out performed the existing policy.

The conclusion at this stage was that the new policy of increasing the credit limits to the lower behaviour scoring accounts would generate additional income for the business due to the increased spend and overall created a more profitable portfolio.

3.5.3 Quarter Three Results
The cumulative results after the third quarter showed a totally different picture than the prior six months.

Income was still better, indicating that the additional credit was being utilised. Costs had dramatically increased, to the point where the overall cost: income ratio was now worse than the existing policy.

On further analysis of each of the cost variables a major change had occurred in the collections area. Many accounts who had received the increased limits had taken up the additional credit and had become delinquent and were now costing the organisation in terms of managing collection activity and higher projected write off.

3.5.4 Quarter Four Results
The cumulative results after the last quarter showed similar results as the third quarter with costs further rising, due to bad debt or accounts being charged off due to non-payment.

The results indicated that the new test was generating more income but at the expense of higher costs due to bad debt provisioning. The resulting cost: income ratio is now showing that the new test is not as profitable as the existing policy.

4. Summary
In this test the retailer was very aggressive in extending the credit limits in the lower behaviour ranges, where although additional income was generated, the increase in delinquency and subsequent bad debt provisioning outweighed this gain. The principle of testing in the lower behaviour ranges where credit limits increases have previously not been granted will apply to any organisation, although each may have its own interpretation of a ‘low’ behaviour score.

In a new test the organisation would choose to be less aggressive and create a test using scores between the existing policy and the test conducted in this study.

One of the important findings in this study was that the minimum time required to draw accurate conclusions for this test was 8-9 months. Conclusions drawn before this time would have resulted in the changes being implemented across the whole book. This would have had serious financial implications for the organisation.

In summary, this test would not have been implemented due to the higher delinquency costs, but the organisation now has objective proof of the impact of increasing credit limits at these behaviour ranges.

About the Contributor
PIC Solutions is the leading specialist credit risk solutions company in the EMEA region. With offices in Cape Town, Dubai, Johannesburg, Manama and Nairobi, we deliver integrated analytics, consulting and software solutions to over 150 companies in 30+ countries. We work worldwide with organizations to improve performance, drive strategies and enhance profitability. Analytics | Consulting | Software