Debt Management and Collection Analytics

Once the consumer commits a default on his due amount he is moved to collection. Collection team follows up with mails, notices, SMSs, calls and if required debt collector visits. All these activities involve lot of cost and eat up the profit. Therefore the need arises of the targeted calling and campaigning to maximize the profit. Consumers are divided based on their vintage in collection because their behavior differs from early delinquency stage to late delinquency stage. Early to mid vintage Collection cycle is considered as mid way between time when account enters the collection and write-off. The latter half of the period is defined as “Late Vintage Collection Cycle”.

Early to Mid Vintage Collection Cycle:

Usually the consumers who defaulted on their due between 0 to 90 days past dues are identified as Early to mid collection cycle. This varies domain to domain depending upon the risk exposure. The objective in early to mid collection cycle is to collect entire due amount at the lowest cost while retaining good consumer for future business. The companies also need to identify very risky consumers or frauds so that they could be fast forwarded to termination stage without any more risk exposure.

Once the consumer fails to clear his due more than a cutoff amount then he is considered to be in collection and passed onto collection team to recover the outstanding. The challenge in the collection process is to decide the type of consumer, best type of communication and allocate the collection resource within the cost constrain. Different consumers respond differently to different action. The company needs to consider cost, effectiveness, risk exposure and the consumer preference while deciding the most effective strategy of collection.

Other most important segment is self cure. Self cure consumers don’t need any collection call and return to 0 Days Past Due (DPD) stage or current stage by themselves. Collection system should be able to identify these consumers so that they can concentrate their collection activity on other consumers with higher risk score.

The objective of all analytical activity in this stage is to collect the outstanding arrears from the consumers and prevent the flow of consumers in deeper debts. Once the consumer reaches the late stage of collection the chances of him to be paying back debt reduces drastically. Predictive models can be used to calculate the risk scores of the consumers at this stage. The consumers with very low risk scores are the consumers who are self cure or revolvers. They will usually clear all dues with all fines and interest with minimal efforts. Usually these consumers will fall in the cluster of good consumer in the segmentation exercise done in pre delinquency stage.

The consumers with high risk scores are consumers who have real risk of rolling forward i.e. to slip deeper in debt. The debt collector need to focus on these consumers and tries to bring them back to no due status by understanding their problem and if required restructuring the debt or helping them with some debt counseling. The companies apply their best debt collector and the most effective line of communication over this portfolio. The best practice is to calculate the risk score at the time consumer enters the collection cycle and after 90 days (Mid way between entering the collection and write-off/termination)

Late Vintage Collection Cycle :

Once the consumer moves to late vintage collection cycle that is beyond 90 days past due than the chances of him turning bankrupt or getting write-off/charge-off increases tremendously. According to our experience more than 50% of this portfolio roll forward to termination or write-off stage. In early stage of delinquency the entire focus is on recovery of the full arrear from the consumer. But once the companies fail to recover full outstanding in early stage of delinquency then in the later part of delinquency the need is to maximize the recovery(might not be always the full arrears) at optimal cost. At this stage company start considering other options like outsourcing the collection activities at commission basis or selling the accounts to other company, most likely the debt collection agency (DCA).The need is to decide which accounts are more likely to make some payment and which accounts are best suited to collection agency or should be sold right away.

There is unprecedented increase in Non Performing Assets (NPA) in recent time because of economic instability. The lending business has suffered the most because people lost job and were forced to default. Given some support and counseling, these people would prefer to clear their debts. Predictive analytics could be used to identify these good consumers and can be approached by the companies with some debt counseling and if required some restructuring the debt according to consumer’s financial health.

Collection function usually treat late vintage portfolio differently from early vintage portfolio. The objective at this level is to prevent consumer from rolling over to write-odd or recovery level. The best way is to calculate payment predictor score and target the consumers with greater likelihood of payment. These consumers could be approached with most effective line of communication. At the same time the accounts with lower payment predictor score could be sold to DCAs before write-off/charge-off thus maximizing the returns. This solves the one of the major challenge to decide which accounts to pass to debt collection agencies. The beauty of the implementing the analytics solution is complete automation of the allocation process. The output score could be used to determine the likelihood of any settlement at account level and dividing the portfolio into “High Risk” , “Medium Risk” and “Low Risk ” accounts. The “High Risk: accounts could be outsourced to DCAs or could be sold to external collection agencies. At the same time “Low risk” and “Medium Risk” accounts could be retained for internal collection agencies.

Another way which is very effective and easy to implement is the “Early Warning Delinquency Scorecard”. A predictive model can be built using information about consumer, his paying and billing information. The output of the model will be ordered score such that higher the score greater the chances of the default of the consumer in next billing cycle. The score could be used directly to segment the consumer for differential treatment. The consumers with higher risk could be followed up with polite reminder calls. The output of the scores can be combined with the consumer segmentation to identify the usual revolvers or Runaway Spend or Frauds proactively.

ScoreData Advantage:

Since its inception ScoreData Corporation has worked in collection analytics across the domain and across the life cycle stages of consumer- right from pre delinquency stage to collection to recovery till skip tracing stage. We have developed in house subject matter expertise in providing collection analytics solution for different markets. We take pride in providing very robust but easy to implement solutions. . Our predictive solutions segment the consumers using their demographics, collection data and also do analysis to find the most suitable treatment for each segment. We use predictive analytics to make accurate estimates of a consumer's propensity to repay, as well as the likely amount that the consumer will repay. Our collection models help to distinguish between self-cures and potential long-term delinquent accounts to maximize the collection from the delinquent accounts while preserving valuable consumer relationship.