Debt Relief Considerations After Covid-19 – Don’t Panic!

by Jashan Augustine

The global financial crisis in 2008 left many scarred over time. The crisis did provide us with many important lessons and data points, which if used properly can help us manage collections and debt relief for the pandemic effects that we are facing now. For example, one key collections observation taken from 2008 related to the average time a consumer returned to financial good. FICO research revealed the following:

Average time to return to financial good:

Two years before financial crisis – 30 months
Two years after financial crisis – 9 months

This was a positive sign, showing that the customers who rolled into collections during the global financial crisis were actually good customers, with a short-term payment problem. This implies that a similar trend could be followed during the COVID-19 pandemic. It would therefore be important to then identify and value these “short-term pain” customers, as it will ensure a lifetime of loyalty once the effects subside.

The Situation in South Africa

Understanding customer behavior is now more important than ever and this relates very well to the debt relief practices that have been put into play by many organizations in South Africa. At the outset, complying with the recommendations set at the BASA (Banking Association of South Africa) level, most banks introduced payment relief schemes. Most of these strategies were reactive and formed part of the tactical phase of COVID-19 relief as the country went into lockdown.

These strategies followed a similar theme of allowing up to 3 months of payment holiday to qualifying customers. They aimed at providing cashflow relief due to the slowdown of the economy and inability for many to enter the workforce during this time.

The slowdown of an economy already in early recession brought on further reductions (275 basis pts) in the repo rates by the SARB – the largest reduction seen since the 2008 crisis, with an anticipation of an even deeper recession than before on the horizon. For savings customers and some investors this was an unwelcome sign, but for consumers in debt it further supported the cashflow from the payment relief programs by providing additional interest relief.

That’s a lot of relief for the consumer, and in order for organizations to protect and understand their customers during these programs, we provided the following advice on changes in their risk and operational strategies.

Understanding Customers Under COVID-19

Collect and leverage as much data as possible during this period.
- Understand the type of customer taking the plans
  - Reason for taking the plan?
  - Are they permanently or temporarily affected?
  - Were they previously in collections/delinquency?
- What leading indicators are in place to segment the risk of these customers and what probability of default do you expect thereafter?
  - Which scores can be used to understand these customers better?
  - Is there any data that can be leveraged to enhance data captured during this time?
Given the pressure on the workforce, leverage as many channels as possible.
- Risk segmentation and time-based activity are critical to effectively segment and manage the critical accounts to be worked.
- Use self-service channels as much as possible to manage costs and workload effectively.
Adapt forbearance offerings to align with payment holiday strategies.
- Ensure that affordability calculations are in place.
- Define early warning customer (permanently affected individuals) actions.
Actively review risk strategies, operational execution and the reliance on IT changes.
- Also ensure alignment between the teams to prevent any operational negation or wasted effort.

Hopefully many organizations can tick quite a few boxes above, or have adequately prepared themselves for the future if these have not been considered. Now as we move into phase 2 of payment relief, a range of further considerations need to be made.

Assessing Your Analytics and Data

We are now in, hopefully, a more proactive phase, as there has been time to absorb and learn from the various organizations and activities both locally and globally. The lingering questions that need to be considered now are:

Assess risk models. Can the scores/predictive models applied pre-COVID-19 still be used?
- There are different thoughts on the topic, but it comes down to the following points:
  - What underlying data is the score made up of? Is it Internal only or internal + bureau data?
  - How dependent or sensitive is it to payment recency characteristics?
  - Has the model been stress-tested previously?
- The bureau scores are generally meant to be more robust due to additional data on the credit profiles of multi-banked consumers. It will be important to use these scores as much as possible to segment your customer base.
- In the absence of bureau scores, internal scores can still guide the risk segmentation. In general, most of the scores used in collections during the period will rank-order and segment risk accordingly. They are meant to identify and react to stress/distress and now it’s become the new normal. One should expect a downward shift in odds, which may mean adjusting the banding at the same risk level or changing the cut-off if applied in decision tree segmentation for the treatment paths.
- Further to the above, it would be worth adding an additional tree segmentation for the different types of customers — pure collections, payment relief temporarily affected, payment relief permanently affected, etc.
Identify new data. Since scores generally relate to credit profiles and this data is limited and affected by the payment relief schemes, what additional data can be applied to accurately assess the next steps for a customer journey, especially if the scores are not providing desired results?
- FICO and OpenWrks have provided a new solution in the UK market, which assesses affordability using open banking data (transactional data, spend patterns, last paycheck) with ongoing updates on customer situation (furlough, reduced hours, redundancy, BAU). Analytical insights can be used to segment customers and deploy highly tailored treatments that will help each of them repay and go back as quickly as possible to financial wellbeing.
- This may seem a long way for our market – but is it really? Alternative data has been collected in the market for a long time, with some credit bureaus having many unique data bases. Further to that, as part of the segmentation, most banks have transactional data on their primary clients, which generally make up at least 75% of their base.
- Using this data will allow you to assess a customer’s situation better and lead you to the next best actions.

Operational Issues

There is a lot of opportunity to apply additional segmentation to your data-driven strategies. The bright side is that most of the data is there and most of the processing power to analyze and utilize it exists. Aligning the operational strategies and being agile in operations is going to be the challenging part.

Our recommended approach is:

Understand your capacity plan.
- Do you have the right skills in the right place?
- Are there specialized agents to deal with good customers who are permanently affected?
Challenge your current use of channel.
- Continue to test and learn different scenarios.
- Apply feedback loops and re-test!
- Evaluate the benefits received from different channels.
- Move towards self-service delivered via omnichannel orchestration.

In my previous post, I mentioned the possibility of collections analysts being uninspired. Well, all is not lost now – it’s just the start of being in the most active and changing environment for the foreseeable future. Don’t panic!

For more information, check out these previous blog posts on debt collection in the pandemic: