Many lenders have rigid policies about granting credit to consumers who have had derogatory events on their credit file. But are these policies justified? A recent analysis shows these consumers are not always as high a risk as we may assume.
Advanced evaluation tools can help determine likely performance on a new account and give a clearer view of risk. Glenn Waine, Director of Data Science and Solution Consulting at TransUnion Canada, takes us through a TransUnion study examining whether lenders may wish to consider adjusting their policies for this consumer segment.
When life happens …
Over the last four years, an average of 904,000 derogatory events appeared on consumers’ credit files each year1 . By law, these events — bankruptcy, charge-off, collection, judgement and repossession — can remain on file for anywhere between two to ten years from the date of the delinquency, depending on the province (bankruptcy filings that occur more than once fall off after 14 years).
It’s understandable that many consumers who have been affected by these events will start to look for credit and open new accounts once time has passed since the information on the derogatory event appeared on their credit file. This presents an opportunity for lenders to engage with these consumers and extend suitable offers.
What tends to happen, though, is that lenders implement stringent policies to mitigate what they perceive to be a high risk, which can results in limited access to credit for these consumers. But are these risks as great as they believe?
Let’s take a closer look.
Are conservative lending policies around derogatory events warranted?
To answer this question, we looked at a straightforward scenario where consumers have a derogatory event that stays on their credit file for seven years, with no other events taking place during or after that period.
Using this scenario, we looked at a test group of over 400,000 consumers2 over a four-year period, from the year before the event dropped off the file to three years after.
We compared their behaviour to that of our control group, which had no derogatory information on file at any time. The control group also had a similar distribution of risk tiers (10% of consumers were prime and above). Our analysis was across lender types.
What happens when derogatory event information drops off the credit file
Consumers with a derogatory event about to drop off their credit file are more credit-active and potentially more likely to respond to new credit offers. Their performance on new accounts indicates they may not be always as high a risk as we may expect.
Note: The drop-off date we talk about here is the date on which this information is removed from the credit file. The DIDO group (‘derogatory information drop-off’) we refer to are consumers whose credit files have been cleared of information relating to the derogatory event.
High demand for credit after drop-off date
A spike in inquiries into credit a couple of months before the drop-off date suggests that DIDO consumers are aware of the approaching drop-off.
Originations of many credit products, particularly bank cards, rise sharply in the first three months after drop-off. In fact, DIDO consumers are more than twice as likely as those in the control group to open a credit card within those first few months and almost twice as likely to do so after 12 months.
Credit score improves significantly after drop-off date, then stabilizes
The vast majority of DIDO consumers studied saw an improvement in their credit scores after the derogatory event information was removed from their files. For many, this improvement is significant. Sub-prime consumers are the most likely to see an improvement in their risk score and least likely to see it decrease.
The picture is a bit different after six months, as we see more changes to the risk scores of DIDO consumers than those in our control group. However, things do settle down for the most part, with the majority of these changes (73.0%) being between -20 and +20 points (76.9% for the control group).
As shown in the table below, DIDO risk scores are far more stable after 12 months, although they’re more likely to worsen among those in the lower risk tiers.
Score Migration Comparison, Post-score to 12 Months after Derogatory Drop Off
Differences in delinquency rates on new accounts narrower than expected
The tables below compare the account-level delinquency rates of the control group and DIDO group on newly opened accounts.
Account-level delinquency on newly opened accounts, measured at 12 months on book
While the control group generally outperforms the DIDO group, this is not the case for all risk tiers and in some cases (for example, near prime) the gap between the two is narrower than we would think.
Why the gap? It could be that consumers who have experienced a major derogatory event are less concerned with the implications of doing so again. Or, that lower access to revolving credit before the DIDO date could reduce their capacity to pay if they experience financial hardships. Looking deeper into the data, we saw that the DIDO group had slightly more open trades than the control group at the time their credit files were cleared of the derogatory event information (the drop-off date). Yet, overall, balances and limits were lower, especially in the higher tiers (prime plus and super prime).
Our hypothesis is that this may be largely due to rules that exclude DIDO consumers from marketing campaigns, automatically decline their applications for credit or allow credit at very low limits only.
Conservative to confident: revising the rules to expand the customer universe
There could be an opportunity for lenders to engage with these consumers in a different way, and it starts with revisiting policies and modifying them to reflect the current operating environment.
There are several ways to approach this. In our analysis, for example, we used just one CreditVision variable to find DIDO consumers who would have a similar level of delinquency risk as those in the control group. The level we chose was one that could be suitable for approving credit. After running the analysis, we found that 83% of consumers in the DIDO group would have been considered as having a satisfactory level of risk for credit in the study parameters. Without this analysis, though, there would be no way of determining this, and those same consumers would likely have been declined.
Advanced credit evaluation tools that give a view of the consumer’s risk trajectory over time and propensity to move to a different risk tier can also help. With this kind of insight, lenders can be better able to identify consumers whose risk levels meet their thresholds and targets.
There’s value to lenders in refining their portfolio to understand which DIDO consumers they could be doing more business with. By revisiting their policies and using additional analytical tools, they will be in a stronger position to expand access to credit for this group of consumers and in turn grow their own portfolios.