sasactivity.blogspot.com >>

SAS Help, SAS Activity, sas installation, sas certification, sas base preparation, sas interview, sas prep guide, sas tutorial, SAS/BASE, SAS/MACROS,SAS/SQL, PROC SQL, Macros, SAS/STAT, SAS/BI, SAS/DI, sas, Enterprise Guide, Enterprise Miner, Information Delivery Portal, SAS/ACCESS, SAS/GRAPH, SAS/Warehouse Administrator, SAS GUIDE ★★★★★ #1

Thursday, July 26, 2012

Gmid

Analytics Case Study: Recovery Strategies on Written off Credit Card Portfolio



Objective


To help prioritize accounts for collection efforts by identifying the ones having the highest recovery potential, thereby optimally allocating collection efforts and maximizing recoveries.

Background and Challenges


Due to economic recession and increased default rates in credit card industry, written-off portfolios are piling up every day. Banks are digging deeper into their books of written-off customers to hunt for recovery potential.

The bank wanted to have a recovery propensity scorecard on their written-off portfolio to help them run a targeted recovery campaign.

Our Approach


The data provided by the bank was mainly profile rich. In order to identify the customer behavior, transaction data and calling history are also needed. We started by profiling of the customers. We used past one-year data for our analysis. The bank provided us the complete charge off portfolio of last one year along with the details of the settlement cases.

Various Good/Bad scenarios were considered based on the payment propensity and settlement propensity of the entire debt pool. The final scorecard evaluated the payment propensity amongst fresh charge-offs.

(i). Good: Success will defined by the event of any payment (Rs. 2000 or more) within the next 6 months of charge-off
(ii). Indeterminate: If the account makes payment less than Rs. 2000/-
(iii). Bad: If the account makes no payment

Different statistical techniques were used to identify the important attributes that help in identifying the potential settlement cases. Some of the fields that came important for the predictive model were-

(i). Number Of Times The Account Has Been In Bucket 5 (120-150 DPD) in The Last 6 Months
(ii). Difference Between Last Purchase And Last Payment
(iii). If the account makes no payment
(iv). Time Since Last Payment

Results and Implementation


Using our analysis, each customer was given a score and the portfolio was divided into ten segments of customers with varying propensities to pay. The bank then designed design their recovery campaigns for each of these segments with customers in higher segments are targeted with higher intensities and optimal settlement offers. This has allowed the bank to tap 80 % of their potential customers in 20 % of the segments thus saving on costs with improved focus.

The exercise is carried on 25th of every month and the revised segments are attempted for recovery in the subsequent months.


Posted by ankit at 5:52 PM
Email ThisBlogThis!Share to XShare to FacebookShare to Pinterest

No comments:

Post a Comment

Note: Only a member of this blog may post a comment.

Newer Post Older Post Home
Subscribe to: Post Comments (Atom)

Total Pageviews

Powered By Blogger

Infolinks In Text Ads

Pages

  • Home
  • SAS Installation
  • About SAS
  • SAS Libraries Every SAS file is stored in a SAS li...
  • SAS Data Sets
  • Proc Step Vs. Data Step
  • Datalines or cards statement
  • Import files (CSV,EXCEL)
  • Proc Content
  • Proc Print:

Followers

Search This Blog

Blog Archive

About Me

My photo
ankit
View my complete profile

Labels

  • advanced data analysis services (1)
  • business analysis companies india (1)
  • Business Continuity plan (1)
  • customer analytics india (1)
  • customer loyalty solutions india (1)
  • customer retention solutions india (1)
  • Data Analysis (1)
  • data analysis India (1)
  • data analysis services (1)
  • data analysis services companies (1)
  • Data Center Disaster recovery plan (1)
  • data modeling companies india (1)
  • Disaster Recovery plan (1)
  • every 10 observation (1)
  • every 2nd observation (1)
  • every 5 observation (1)
  • mis reporting solutions india (1)
  • monthly reporting automation solutions india (1)
  • monthly reporting solutions india (1)
  • outsource chartinf and reporting solutions india (1)
  • outsource mis automation solutions (1)
  • outsourcing data analysis to India (1)
  • Predictive analytics (2)
  • predictive customer loyalty solutions india (1)
  • predictive data analysis (1)
  • predictive data analysis companies indida (1)
  • predictive modeling companies india (1)
  • Predictive Modelling (1)
  • R every nth observation (1)
  • retail banking analysis solutions (1)
  • retail customer loyalty solutions india (1)
  • risk analysis services india (1)
  • risk analytics solutions india (1)
  • risk analytics services india (1)
  • risk assessment plan template (1)
  • risk management solutions india (1)
  • SAS Data Analyse uitbesteding (1)
  • sas every nth observation (1)
  • SAS Voorspellende analytics uitbesteden (1)
  • SAS voorspellende data-analyse outsourcing (1)
  • SAS Voorspellende Modelling uitbesteding (1)
  • telecom customer loyalty solution india (1)
  • 分析論 アウトソーシング、予測モデリング、予測分析は、インドへのオフショア分析、分析オフショア、顧客ロイヤルティ、顧客維持を外部委託 (1)
Picture Window theme. Powered by Blogger.