Banks today are employing a variety of tactics to detect and prevent application fraud before fraudsters get their hands on new credit cards or loans. Authentication processes have arisen that can validate applicants’ identities to weed out both stolen and fabricated identities, as well as candidates with questionable backgrounds that may indicate first-party fraud.

A credit assessment memorandum (CAM) is the first step in evaluating a loan request in a structured and formalised manner. It initiates the formal process of arranging and packaging a loan request for approval by bank management. Credit providers can hence make meaningful decisions on credit requests based on the information provided by a CAM.

Many technologies are available to help banks verify account ownership by querying applicants about account details or submitting micro-deposits and checking whether the funds appear. Financial institutions also utilise knowledge-based authentication (KBA), which entails asking applicants questions about their credit histories that only they would know. Average mortgage payments, automobile licence plate numbers, and salary information can all be found in credit reports, which can be used for KBA.

Credit application fraud occurs when an attacker applies for a loan or other type of credit using stolen personally identifiable information. The financial system may even be exploited to establish a synthetic identity, i.e., a fake person, which is even more difficult to detect by financial institutions.

Project and Solution Overview

Kotak Mahindra Group, one of India’s largest financial services companies, provides a diverse range of services to individuals and businesses, including commercial banking, stockbroking, mutual funds, life insurance, and investment banking. The bank wanted to bolster its credit application process by identifying fraud at the document level and cross-analysing financial documents.

Perfios, a leading Indian product technology firm, was selected by Kotak Mahindra to implement a solution that provided end-to-end credit information of their customer by cross-analysing various data sources to generate reports that aid in credit underwriting & decisioning. Perfios’ product offerings assist businesses in developing vertical applications across a variety of domains.

The report generated by the solution takes information from other Perfios utilities, such as Bank Statement Analyzer, Bureau Reports, Income Statements, tax statements, etc., and provides a 360-degree view of a customer’s financial history; thus, helping to make insightful lending decisions. The document metadata identification occurs at a single document level to find if the original document was tampered with or not.

The Cross Analysis offered by Perfios participates in the bank’s underwriting in a frictionless manner, enabling the digitisation of the process and helping the credit officer take an informed data-backed decision. The triangulated data can be customised based on the bank’s requirements.

 Implementation Process

Implementation was carried out using API integration with the bank’s core banking solution and loan origination/loan management system. The base version of CAM was made available to the client in 3 weeks from the time requirements were finalised with the bank. This included certain API & report customisations as per the bank’s policy.

 Benefits Achieved

This project provided a host of benefits for the bank. The Credit Assessment Memo (CAM) saves valuable time for personnel like credit managers by offering a report that is crisp, accurate and detects financial fraud. CAM also assists with cross-validation between different financial documents to identify fraudulent behaviour by analysing variations among documents and identifying any tampering with documents.

The Fraud Analytics solution puts together all the pieces of the puzzle: bank statements, financial statements, GST, ITR, and mobile device information to form an image of a credit applicant’s financial and potentially fraudulent behaviour. The solution also cross-verifies data across numerous sources and provides a variance percentage, highlighting any behaviour or transaction of a suspicious nature.

Other key benefits for the bank include:

• ~60% reduction in cost due to reduced man-hours for loan processing, thereby improving the quality of data for underwriting.
• ~3% of applications identified as fraudulent, helping the bank reduce risk.
• Time savings of 3-4 days for underwriting and loan processing for FIs in the retail banking segment and 5-7 days in the SME segment.

Conclusion

The amount of data generated on a global scale is increasing, with no signs of slowing down, making it difficult to detect fraud indicators. Internal controls are insufficient on their own, and manually examining all this data is very expensive and time-consuming for large global corporations. However, fraud analytics provide a fast overview of business processes as well as the option to dive into specifics. The adaptability and flexibility of Perfios’s technology enabled Kotak Mahindra Bank to determine the creditworthiness of borrowers while also helping to minimise the risk associated with loan disbursements, allowing for significantly faster and more comprehensive assessments.

Kotak Mahindra Bank worked with Perfios to deploy an end-to-end credit information and fraud analysis solution to improve the underwriting process

Key Facts

Bank: Kotak Mahindra Bank
Headquarters: India
Founded: 1985
Asset Size: USD 64 Bn (2021)
Supplier: Perfios
Headquarters: India
Founded: 2008
Solution: CAM Generation and Fraud Analytics
Timeline: Three weeks; Go-live in Feb 2021