Artificial Intelligence

How can Analytics help Asset Reconstruction Companies?

The leading problem in the country right now is unforeseen volume of Non-Performing assets with the banking system. To clean the balance sheets of the banks asset reconstruction companies purchase bad assets or NPAs from banks at negotiable price. Banks auction NPA accounts through e-auction or offline bidding process. The information about the accounts for auction is received through emails or is posted on bank’s web site. There are a lot of steps that need to be followed to accomplish the process within the stipulated time. The whole process is generally managed manually by the ARCs using Excel files that leads to a lot of challenges like:

  • Information about PIMs (Preliminary Information Memorandum), legal reports present in silos.
  • Manual process of submitting EOI for each auction, shortlisting of accounts etc. is prone to errors and delay.
  • Difficult to identify accounts and securities which have already been evaluated earlier (in case an unsold account is again listed for auction)
  • Difficult to track the expenses incurred in the realization process regularly
  • Difficult to track the current status and progress of each account and security

One of the largest publicly traded asset reconstruction company in India was facing some similar issues. The company wanted to digitize the entire process in which the complete information of all accounts and securities is maintained in a central database. The firm wanted to automate following processes:

  • Scanning bank sites regularly to find new auctions.
  • Extracting required account & associated securities information from PIMs and store it into a database.
  • Linking accounts which have been put up for auction again by the bank to remove duplication of effort.
  • Extracting required information from legal and valuation report.
  • Tracking of status & progress of the account realization, post-acquisition.
  • Tracking of expenses associated with the realization process.
  • Generation of consolidated reports from the system.

The firm decided to build a data model to store account and security-related information from different sources like PIMs, Legal reports, Valuation reports. After which a tool was developed to upload the information extracted from PIM, legal reports and valuation reports along with the raw file into a single database. Also, it was necessary to remove duplicate accounts which had been put up for auction again by the bank to obtain a comprehensive and clear view of the information available. This was accomplished using a de-duplication algorithm.


Indicative process and data flow


The company also automated the process of scanning bank websites for new auctions and developed an expense tracking system associated with the project realization process. This way the company was able to improve the overall efficiency and accuracy of the overall process.

Want to know more use cases of analytics for your company? Then reach out to us at