Mōrena
It's been a hectic few weeks 🤩🌟🚀
I've attended two conferences, and they were excellent; I have been networking and meeting all sorts of fantastic people doing incredible things, and we've been accepted into the Ministry of Awesome's Founder Catalyst Program.
Broadly, the people I've been chatting to fall into two categories. In the first category, people know very little about credit risk per se. However, they do know about the Credit Reporters, They understand that all credit providers need to assess credit risk at key parts of the credit risk life cycle. They intuitively understand that at the moment, we're seeing increasing arrears, driven by a combination of interest rate increases and cost of living pressures. They understand that resources are scarce and that training new recruits takes time. They also understand that this could lead to adverse impacts on credit providers, especially if things continue to track as they are.
In the second category are those people who have a much deeper understanding of credit risk. They either work in the industry, or have worked in the industry, and understand the concepts of scoring and credit risk assessment. Even within this group, there are degrees of understanding of what scorecards are and how they may be beneficial.
But the fact remains, as I have been meeting with so many people, I've been on the receiving end of many questions. I love being asked questions - it helps me better understand what the people I'm speaking with understand about credit risk. I especially love follow-up questions, as this helps me determine how much jargon I have inadvertently shared. Jargon is seldom helpful, especially when speaking to someone who has not spent 25+ years in the industry.
All this means I have been answering questions about who we are and what we do. I'm often asked some variation of this question.
How are Scores4All different from a credit bureau and the [scoring] services that they provide?
There are several pieces to the answer.
Source of Data - where does the data used in predictive models come from?
Credit Bureaux are also known as credit reporters - companies share their data with the bureaux, and the bureaux hold all the data. There are well-defined rules that govern what type of data may be shared with them and who may access that data or any scores/models or insights associated with it. The purpose of Credit Bureaux is to provide a holistic view of individuals' credit activity.
Scores4All can be considered an extension of our customers' credit risk scoring and analytics teams. Our customers send us data and we score and return their records. We only use our customers' data to score their customers. And anything we do with their data is for their benefit alone. Our purpose is to provide our customers with a collections score so that they can make the best possible decisions about how to work with their customers.
Target Variable - what is being predicted?
Credit Bureaux typically build scores that predict the likelihood of an individual reaching high levels of delinquency across several of their products over the next 6 to 12 months. These outcomes are typically not built for individual credit reporters, but rather give an overall view of the likelihood that an individual may fail to meet their payment obligations.
Credit Bureaux can, and often do, build scores to be used specifically when an application for credit is being assessed, often called Account Origination Scores. They also build Account Management Scores, typically often used in account management strategies such as credit limit management or collections.
Bureaux may also build scores specifically to be used within Collections strategies.
In all cases, the target variable, or the outcome that the score is predicting, is not specific to any one credit reporter, rather the score is a function of customer behaviour across several products and credit reporters.
At Scores4All, we build our models based on each of our customer's data. Our Collections Models predict the likelihood that our customer's customers will meet their payment obligations within the next billing period. Our models are calibrated specifically to our customers' credit risk profile, products, industry and region. This means that the credit risk scores that we deliver to our customers are designed specifically for their portfolio.
Development Frequency - how often are predictive models trained and implemented?
At Scores4All, we continuously train and evaluate our models. And then, we implement the best, most predictive model. All this happens seamlessly in the background, ensuring our models are continuously trained and deployed to deliver the best results.
Hopefully, this gives you a better idea of how Scores4All Collections Scores differ from Credit Scores provided by the Credit Reporters.
What other questions do you have? Drop them in the comments, I'd love to know what you're wondering about.
For a confidential discussion on how we can help you and your organisation implement scorecards, please email us or book a 30 min online meeting with us.
Comments