Loans Data Management
Loans – How to use Data to improve your portfolio health?
Retail lending is all about volumes. We refer to Car Loans, Home Loans, Auto Loans, Credit Cards, and consumer loans etc. as retail loans only because of the volume impact they bring about. The sheer volume of these loans and card facilities, make it absolutely essential for the organisation to have an efficient data management and analytics systems and processes.
More often than not, organisations tend to either ignore or underplay or provide step-motherly treatment to data capture and maintenance. While everyone takes pride in writing voluminous process flow diagrams, when it comes to data capture, there is absolutely nothing to guide the person on how and what to capture on a particular given field in the system. This leads to inconsistencies in the data capture and impedes effective analysis of the same. We give hereunder some key steps in Data Maintenance that if taken in the right earnest, will go a long way in helping the organisation manage its loan portfolio well. These steps are :
- Data Capture
- Data Entry
- Data Storage
- Data Analysis
Data Capture Templates
While there could be systems available with most organisations to store data, the first level of data capture still happens with the help of physical file documents only. It is best to clearly define a template and the formats in which the data must be captured from the physical file. The different heads of data in the template, should match with the data capture structure in the system. The formats of data allowed in the template should also match with the format of the fields in the system. For e.g. the date format for 21st January 2015 could be, 21/01/15 or 01/21/15 or 21Jan15 to any other. Once it is thus standardised, all entries in the physical and the system should follow the same format. So also for all other fields.
If the data capture format is standardised, then the task of data entry becomes that much easier. All data points must be entered. If there is no data available for any field, the options for entering (or choosing from dropdowns) “Not Applicable” must be provided for. While entering in the system, wherever possible, system validations should be provided. For example pincode field must have 6 digits or if there is a “Private Limited” in the name of the borrower (as in ABC Private Limited), the constitution of the applicant should not allow any other entry than “Private Limited” etc. This would help improve the quality of data. For critical fields, an option of dual entries with a ‘Maker’ and ‘Checker’ steps should be done so that they are double checked. The data entry software should be tailormade to ask the Checker to reenter the datapoint. For example, if the date of birth is entered by Maker, when the file flows to Checker, the field should be blank and the Checker should reenter the Date of Birth. The system at the backend should check if the DOB entered both by maker and checker tally.
In financial services, much of the transaction data that is stored in the system is dynamic. The loan outstanding keeps changing on a daily basis as the principal amortizes. From a data analytics perspective, it is important to keep a regular back up of the data at fixed intervals so that the figures as on that particular day are available for analysis at a later date. Month end data is absolutely critical. Therefore it must be made a practice to store the month-end EOD data for all the accounts in the system. Thus if on July 2015, one wants to know the outstanding amounts for all the accounts as on 31st January 2015, the same can be retrieved from the back data stored. One could argue that the data on a daily basis can also be stored and retrieved from the system when required. That may be possible, but it would then need advanced functionalities to be built into the system, which most systems would not have.
Data Analysis is the most critical piece of the Data Management. You have put processes in place to capture, enter and store the data. So you are now sitting on a huge amount of database. But unless you know how to analyse this data, you are losing significantly on the available opportunity. The analytics officer in the organisation should be familiar with the concepts of segmentation cuts, contribution analysis, static pool analysis / Vintage MOB curves, Flow rates analysis, Resolution Rates analysis, Was-Is analysis etc. to be able to make the best of the data that is available. Data analysis is not simple tabulation of the data to show how much of business each centre or each channel has sourced. Right approach to data analysis can help you identify how to harness learnings from your past performance and use that to modulate your credit policies for the future. Discussions on credit policies between the business and the risk teams become that much more easier when done on the backdrop of the data. Data analytics can help in cost reduction, loss reduction and enhancement of revenue, if approached with the right intent and spirit.
Sineedge Consulting can help your organisation get the most out of your data. If you feel, your current process infrastructure is not strong enough to enable capture of data, we can help your organisation put that together. Please write to us at [email protected], for any specific questions you may have or send an sms to +91 9910112704 or +91 9871040392. We will surely call you back.
Contact Sineedge Consulting for Process Reviews, Data Analytics, Due Diligence, Portfolio Audits, Training, Policy Formulation and a host of other specialised services provided. Banks, SMEs, Start-ups, NBFCs, HFCs, Educational Institutions, Private Equity firms etc. are welcome to connect with us.
Photo Courtesy: Bionicteaching / Creative Commons