e-news

Pre Delinquency – currently up to date

---

Early Collection – 0-30 days delinquent

Mid Collections – 31-90 days delinquent

Late Collections – 90+ days (prior to write off)

---

Litigation - Post write-off

Debt Sale - Post write-off

Each of these sections requires different risk segmentation and different operational action paths, as well as leveraging different data assets.

Pre Delinquency
Sometimes referred to a ‘triage’ this phase is often left outside the realms of collections, and inhabits the customer management world, where line assignments and cross sell drive much of the activity.
In recent times Pre delinquency has received more focus, with major lenders leveraging customer management engines to actively contact ‘x’% of their riskier customers proactively
This is not only seen as responsible lending in the eyes of the regulators, but also reaps significant benefits when reducing both bad debt and provisioning levels.

While internal data can be used to model and predict a customer’s likelihood of rolling into delinquency, a far stronger and more robust method combines this internal data with external data.

This external data highlights where customers are experiencing difficulty elsewhere, along with other key stress indicator – such as indebtedness.

E.g. Customer is up-to-date both internally and externally, however utilisation has moved from 10% to 90% in the last 3 months.

This is an indicator that the individual is under stress and a customer service call can help understand if the customer has undergone a major life event (unemployment) and thus result in active preventative action and encourage debt advice from internal or external debt counselling services

Early Collections
Once in collections there used to be two schools of thought, proactively chase the debt early, or leave it and then work the debt that rolls into later stages of delinquency. In a previously benign lending environment both systems had merit dependant on product and customer base.

However, with recent economic turmoil, only the most confident lender would leave all their early stage collections to self cure.

There are several effective ways of assessing risk at an early stage and using data within a dedicated collections decision engine can help look at both propensity to roll further delinquent and propensity to pay models.

Combining this modelling can help segment customers into:
1	Those to leave to low cost actions (SMS, Email, Letter) as they are likely to cure with minor prompting – e.g. irregular payers who dip in and out of collections previously having never been worse than 1 payment down.
2	Those who should get more aggressive action (fast track to litigation, early debt sale, or active dialler program) e.g. high indebtedness elsewhere, or multiple delinquent accounts.

Also worth bearing in mind as a corner stone to some simple segmentation is a Balance at Risk value. This combines the current outstanding balance with one or more risk models. The value produced help give a clearer representation of the value that could be lost.

A simple example of this is:

Account 1 has a balance of $1000
Account 2 has a balance of $500

So traditionally, on a simple balance perspective account 1 would be worked harder and more resource dedicated to it.

However factor in risk models

Account 1 has a 10% chance of rolling to write-off
Account  2 has a 50% chance of rolling to write-off

As a result Account 1 you are likely to lose $100,
while Account 2 you would lose $250.

By combining the two elements together it changes the shape of which account receives more collection activity and the actions that would be undertaken – this is a simple and fundamental technique to use in collections and ensure resource is appropriately focused when and where it would be most effective.

Mid Collections
Once accounts roll onto this status of delinquency there is still plenty that can be done to bring the account back into order while ensuring resource is most appropriately used.
Segmentation and risk profiling is generally not as split and granular as in Early delinquency but there should still be different strategies assigned based on exception types, external data, and quite significantly internal data.

Internal data will reflect both previous actions on the account so far and also performance history (broken promises to pay, aggregated history – L3m, L6m) all of which should be used to dictate how accounts should be treated going forward.

An example how this can be used to drive contact strategies are:

Initially see if the customer was contacted in the previous month, and if so have they subsequently broken a promise to pay.

These two key bits of information tell a great deal about the customer:

• Firstly that you were able to get a right party contact or not?
• Secondly if contact was made did the person state they would pay you?
  

If they have broken this promise to pay then this can drive another contact with the customer and tailor the conversation to why the promise was broken (if a customer has previously dipped in and out of collections with multiple broken promises then this should also influence if you give them another option to set up a promise to pay, or if you fast track them to litigation or debt sale – thus allowing resource to be allocated on more relevant customers).

Equally if you haven’t contacted the customer via phone (or field agent), either due to a low cost previous action path, or no right party contact then the strategy here can help drive a more direct contact route, or to carry out some additional trace work so that contact details are updated.

Analytical models also play well in this space, where the models can help dictate if customers should be pushed onto later more aggressive collections, sold immediately to generate cash flow, or kept in a standard contact strategy as they can be rehabilitated.

Late Collections
When accounts roll on deeper into collections the emphasis is on recovering as much of the outstanding debt as possible, rather than rehabilitation which drives many contact strategies up to this point.

Therefore its now a case of selecting which higher cost contact to perform, or simply to push the account onto write off.

If there is an asset still associated with the debt then this should come into play and active measures should be taken to retrieve this, and then followed up to obtain the remaining balance value (retrieving the asset will always help minimise losses as the value of the asset will generally form the bulk of the outstanding debt)

Post Write Off Collections
As accounts are written off there are still some important decision to be made concerning what to do next with the accounts and drive that final attempt to minimise the losses. Those options are as follows –

Litigation
Does the balance warrant the expense of taking legal action to retrieve outstanding funds.
What is the likelihood of the legal action being successful – this can be modelled.
In a similar way to generating a Balance at Risk value at early stage delinquency you can also combine the balance, cost of litigation and degree of success together to get a clearer picture of which accounts should be driven down this route and which alternative action should be carried out.

Debt Sale
When litigation isn’t financially viable, the next option is to sell it to an external agency.
Previously companies have been unsophisticated with debt sale, and price at a portfolio level, rather than an account by account basis. By being more granular, using individual level pricing models, and taking into considering any additional information that has been captured, the true price of each account can be calculated.
It also means different debt types can be split out and sold to specialist agency’s (e.g. gone aways) leaving a cleaner more valuable pot of individually priced account to go to main stream agencies.

Historical Write-offs
The final option is to just keep the written-off accounts in a database, and class them as ‘rested’. Then every quarter these records can be assessed to see if they should be worked. Internally data can be used in a limited fashion – primarily based around customer level information.

For really effective assessment of this debt, then external data (Bureau) should be considered, where reductions in other credit accounts, and new borrowing can help indicate an improvement in circumstance and an opportunity to retrieve the written-off debt.

Conversely if the credit status with other lenders is declining then this could indicate it’s a prime candidate for debt sale as any recovery seems unlikely in the short to medium term.

Summary
With collections a relatively under evolved area within businesses, and the current economic climate it has never been a better time to improve your collections capability, and leverage significant quick wins that directly impact the bottom line.

The above sections are by no means an exhaustive list of collections activity and tailoring it to fit in with company policies, regulation, current infrastructure and resource is a fine art and Experian have a wealth of experience to help ensure maximum impact is generated within the collections space – adopting many of the principals laid out above.

By effectively using internal data assets there is significant benefits and segmentation that can be generated. Factor this in with aggregating this data across 3, 6 and even 12 months and you get even more power and can move into reasonably sophisticated collections strategy.

Further benefit can also be driven by bringing external data into the equation, which can add value in a variety of ways across the collections cycle.
However it is important to note that data is only as good as your ability to use it and manipulate for appropriate tasks.

In many collection environments this data is used within manual process and to provide additional information to collectors. This is very powerful and allows operators to collect far more effectively by having the information at their finger tips.

Nevertheless this information can be taken to another level, where by it helps drive sophisticated collections segmentation and risk profiling to reduce the cost of recovery, bad debt levels, and provisioning, while increasing cash flow.

Ultimately if you have a range of data available to you then you should really maximise its use and really drive value from it, whether that be purely internally information or data sourced from outside the company

All collection cases are not equal, so don’t action them in the same way.

Steve Matthews
Senior Business Consultant
Decision Analytics
Experian

.Interested in this subject? You may also be interested in...

Download the Collections Business Review factsheet »

.Read the full Case Study for Anglian Water

Visit the website to read more » Anglian Water - Consulting »

.Read more about Collections and Recoveries

Visit the website to read more »