Bolstering credit with pre-delinquency management

Credit delinquency has been a serious problem for banks over the last decade, particularly between 2008 and 2011. The delinquencies have fallen sharply in the last three to four years but it is generally accepted in the industry that any deterioration in economic conditions can quickly lead to renewed problems. No surprise, then, that financial institutions are increasingly looking at proactive measures to tackle delinquency by identifying customers who are likely to default and where account restructuring may help prevent such defaults.

Pre-delinquency management (PDM) refers to the set of policies, processes and models aimed at proactively resolving potential delinquencies before they materialize and is likely to become more widely used in the future for the following reasons:

First, credit cards traditionally have displayed a higher delinquency and default rate than other retail banking asset categories due to their sensitivity to economic conditions. Second, increasing regulatory pressures and demands from consumers have led to multiple restrictions being placed on collection/recovery efforts by the credit card and mortgage industries. This makes it imperative for the banks and credit card providers to ensure a lower delinquency rate. And, finally, credit cards and EMI (splitting a purchase into Equated Monthly Installments) facilities are being increasingly used in online transactions.

Behavioral Data Analysis

The PDM solutions currently available in the market are based on statistical and trend analysis of behavioral data to discern credit drawdown patterns and repayment behavior. The delinquency forecasting methodologies in the industry cover statistical analysis, which is based on technical analysis, regression analysis and trend forecasting, as well as behavioral scoring, which focuses only on analyzing payments/draw-down patterns and largely ignores specific credit utilization patterns.

These PDM practices display several weaknesses that have resulted in customers being identified as “risky” without an objective assessment. As a result, many “good customers,” or those who do not carry a large credit balance, have never missed payments and are financially strong, have been rejected unfairly, resulting in lost opportunity for the financial institution. For this reason, the credit industry clearly requires an improved risk analysis solution.

Such a solution should ideally enable transaction-oriented behavioral analysis of credit card customers. Given the itemized transaction history on customer purchases across various channels, the bank could undertake in-depth account analysis to evaluate customer behavioral patterns. For example:

• How much is the customer spending, on average, over a period?
• How much of the spending is patterned and is consistent?
• What is the mode of spending and how does it impact account health?
• How much is he or she repaying, in what frequency and in patterns?
• How is the customer usage fluctuating and is it impacting account health?
• Is the customer arbitraging within and between the accounts?

Answers to all such questions could be gained from a diligent capture, organization, dissection, analysis and interpretation of the transactional data available with the bank.

The four critical aspects of this transactional behavior which need to be addressed by a PDM solution are: customer behavioral analysis; peer-to-peer comparative analysis; product behavior-based risk scores; and historical regression factor analysis. Therefore, the key features of an effective PDM solution should include:

• Transactional analysis of potentially delinquent accounts to assess credit orientation and credit worthiness;
• Integration of credit scoring intelligence into pre-delinquency assessment;
• Objective, intuitive and comprehensive set of practical business rules based on different transactional modes, patterns and types;
• Application of separate sets of rules for different customer classes/groups based on various customer segmentation modes/models;
• Ability to create and customize business rule sets to suit the particular credit profile and risk aspects of specific groups of customers;
• Allocation of different weights to specific business rules applied to each customer group, which offers the optimal flexibility to define delinquency assessment models according to the credit and transactional profiles of underlying customer segments. 

Additional risk insights can be made possible by merging the transactional analytics with social media data and periodic credit bureau data. 

Such a PDM solution would also enhance a bank’s business intelligence by enabling access to extensive amounts of customer data with the relevant transactional patterns, behavior trends and credit profiles.

Mr. Varma is principal consultant, Banking Risk and Compliance, with Pune, India-based Tech Mahindra. He can be reached at [email protected].