FIS Blog

Using Statistical Modeling to Better Understand the Value of Debt

Mike Kresse | Friday, February 5, 2016

As all debt buyers and debt collections teams are aware, not all debt is equal. The value of some debts are more than others. Some debts are more likely to be repaid than others. When a debt buyer bids on a portfolio of debt, it is important to evaluate the portfolio carefully in order to determine the right price. If debt buyers bid too much, they risk losing money – but if they bid too little, they risk losing the bid.

The challenges do not end there. Once a portfolio has been purchased, the debt buyer may task its own collection team with recouping the debt, or may appoint a collection agency to recoup the debt on its behalf. Some debts may be sold on and some warehoused.

In order to achieve a positive ROI on the portfolio, the purchaser needs an effective collections team and that team needs to understand where to direct its collections efforts. One would think collections teams should spend their time collecting on an account that owes $900 versus one that owes $500. But that is not always the case or the best solution. The balance of the debt should not be the only consideration as a $500 account might be more likely to pay than the $900 or settle for a greater amount.

How then can debt buyers and their collectors figure out which accounts to focus on? The two most important considerations are the value of the debt and the likelihood of payment. Where the latter is concerned, many debt buyers rely on credit bureau data – but this approach has a number of shortcomings. Credit bureau data can be expensive and may not be fully up to date or available on the accounts in question.

Moreover, legal concerns may arise if collection agencies do not have permissible purpose to use the data provided by bureaus in this way. Some would say that this scenario arises if the debtor has not actively initiated a credit transaction and applies to many types of debt, for example municipal debts such as traffic tickets, parking tickets and some types of healthcare debt.

In order to overcome both of the challenges – choosing the right price for a portfolio and deciding how best to approach debt recovery – debt buyers and collectors are increasingly turning to solutions based on statistical modelling which can help them identify the best bid amount. Statistical models can then be used to develop a prioritization strategy by helping to predict which accounts are likely to pay or to stay dormant.


Tagged in: Credit and Collections, Statistical Modeling, Corporate Solutions

Contact us

Learn how FIS’ institutional and corporate solutions can help you increase your competitive edge.

Contact us