Reading Time: 39 mins

TRANSFORMED RISK ASSESSMENT OF LOAN PORTFOLIOS FOR A LEADING ASIAN BANK

CASE STUDY

A CASE STUDY ON AI-DRIVEN CREDIT UNDERWRITING SOLUTION

A large Asian bank with a market capitalization of US$1.8 billion, offering a wide range of banking services, deposits, loans, saving/current accounts, wanted to improve the underwriting process and approvals rates for Overdraft (OD) and Cash Credit (CC) loan products.

A CASE STUDY ON AI-DRIVEN CREDIT UNDERWRITING SOLUTION

A large Asian bank with a market capitalization of US$1.8 billion, offering a wide range of banking services, deposits, loans, saving/current accounts, wanted to improve the underwriting process and approvals rates for Overdraft (OD) and Cash Credit (CC) loan products.

Problem Statement

Our client is one of Asia’s most prominent banks with total assets of over US$10 billion. The bank caters to a broad customer demographic group with different credit products – from overdraft loans to MSME and corporate loan products.

With a vast customer base, the bank faced tremendous pressure to keep track of all overdraft accounts and cash credit products on an account-to-account basis.

Our client manually monitored credit limits for Overdraft (OD) accounts to understand the borrower usage to increase or decrease the credit limits based on customers’ credit history. Also, the bank was manually identifying accounts that were not using overdraft limits and allocated them to potential creditworthy borrowers who were exceeding the credit limits.

But the manual allocation, reviewing, and monitoring of accounts was tiresome, error-prone, and held back the bank’s overall loan revenue.

Similarly, the bank faced difficulties in validating companies’ credit usage for cash credit products to increase or decrease the credit limits and scrutinize any personal spending from the credit provided.

Lastly, the bank’s risk assessment was solely based on the hands of a branch manager, who evaluates the customer risk appetite based on the income statement, balance sheet, and other collaterals provided. This led to higher default risk.

To overcome these challenges and reduce OD/CC delinquency percentage, boost approval rates and maximize loan-loss adjusted net interest income, the client wanted to transform their current portfolio risk assessment system with a innovative and scalable automation solution.

BUSINESS GOALS

BUSINESS GOALS

Our team of data scientists initiated the project by understanding the current end-to-end credit assessment and approval process in providing a credit limit for Overdraft and Cash Credit products. This helped our team to identify various business challenges and to define the success metric beforehand.

Post discovery phase, our team of data engineers collected all the relevant data fields required to help determine customers’ creditworthiness from different disparate systems using our Enterprise Data Bus (EDB) solution, ZIO.

After collection, ZIO’s data pipeline helped with the continuous integration of data into our home-grown AI-based Credit Underwriting solution, HALO.

By leveraging a combination of traditional financial data, historical repayment data, and alternate data, our machine learning technology identified nuanced patterns and created a unique credit underwriting model that exploits these patterns to identify risky borrowers at the time of underwriting. Also, the HALO machine learning model helped the bank with continuous risk assessment and scrutinizing approved loan portfolios in real-time.

Finally, the model was continuously trained to predict and monitor credit risk for different loan products and deployed the solution in the bank’s environment.

SOLUTION

SOLUTION

Our team of data scientists initiated the project by understanding the current end-to-end credit assessment and approval process in providing a credit limit for Overdraft and Cash Credit products. This helped our team to identify various business challenges and to define the success metric beforehand.

Post discovery phase, our team of data engineers collected all the relevant data fields required to help determine customers’ creditworthiness from different disparate systems using our Enterprise Data Bus (EDB) solution, ZIO.

After collection, ZIO’s data pipeline helped with the continuous integration of data into our home-grown AI-based Credit Underwriting solution, HALO.

By leveraging a combination of traditional financial data, historical repayment data, and alternate data, our machine learning technology identified nuanced patterns and created a unique credit underwriting model that exploits these patterns to identify risky borrowers at the time of underwriting. Also, the HALO machine learning model helped the bank with continuous risk assessment and scrutinizing approved loan portfolios in real-time.

Finally, the model was continuously trained to predict and monitor credit risk for different loan products and deployed the solution in the bank’s environment.

HOW ZUCI SYSTEMS HELPED

HOW ZUCI SYSTEMS HELPED

HOW ZUCI SYSTEMS HELPED

HOW ZUCI SYSTEMS HELPED

BUSINESS OUTCOME

0%
Transparency in the loan approval process
0%
Decrease in delinquencies
0x
Faster loan approvals
0%
increase in loan revenue
0%
Reduction in risk assessment effort (equivalent to 800 FTEs)
0 Months
ROI realized

BUSINESS OUTCOME

0%
Transparency in the loan approval process
0%
Decrease in delinquencies
0x
Faster loan approvals
0%
increase in loan revenue
0%
Reduction in risk assessment effort (equivalent to 800 FTEs)
0 Months
ROI realized

AI-BASED CREDIT UNDERWRITING SOLUTION

ENTERPRISE DATA BUS (EDB) SOLUTION

TECH STACK

TECH STACK

AI-BASED CREDIT UNDERWRITING SOLUTION

ENTERPRISE DATA BUS (EDB) SOLUTION

WANT TO MITIGATE RISK AND BIAS WITH
AI-DRIVEN CREDIT UNDERWRITING? HALO CAN HELP!

Download My Copy
Book a consultation