Predictive models deliver precise estimates of, well, pretty much whatever you want to predict. In active collections, we like to predict the likelihood of default (aka charge-off), and we do it on every account that is one or more days past due. We also predict the odds-of-roll, that is, the odds an account in a given DQ bucket will “roll” either better (i.e., “cure”) or worse (er… “worsen”) in the following month. Finally, for those accounts that have already charged-off, we approach the matter from a different direction—that is, we estimate the likelihood of collectability (i.e., odds of collecting future deficiency payments). To accommodate all these desires, we have about a dozen models working in concert to produce these estimates. From there, we factor-in each accounts current principal balance and the collateral’s estimated liquidation value (if applicable). And despite all this complexity, we represent risk on the screens using a simple letter-grade between “A” through “E”. And so for the end user, these letter grades are very easy to understand. And for your business execs, you can incorporate these Risk Grades into any number of custom campaigns and agent-assignment strategies. Want to assign the toughest accounts to your top agents? No problem. Want to reduce communications on those segments of your portfolio most likely to self-cure? No problem. It’s these sorts of tools that drive efficiency within your organization, and it’s all within reach if you have the right software. For a demo of our disruptive new LMS, contact us today!