Table of contents
At its core, lending is the business of underwriting. Financial institutions such as banks, NBFCs often use traditional underwriting methods such as bank statement analysis and credit bureau reports that provide credit scores. However, this trend is fast evolving to incorporate alternate data underwriting in the risk intelligence workflows.
Typically, 60-65% of digital loan applications are sourced from new to credit customers. In lieu of bureau data, these customers are generally not approved leaving them underserved. Alternate data underwriting is a godsend for improving the efficiency of any lending program as well as for improving financial inclusion by helping them to include thin-file as well as new-to-credit customers in their borrower cohorts.
So what is alternate data underwriting all about and how can it help you? Let’s unpack this.
An Experian report released in July 2022 stated that in the six months prior, Buy Now, Pay Later (BNPL) transactions in India grew 21% , outperforming the 18% growth registered globally. The proof of this rapid rise is all around us - today, Indians can leverage alternative credit products such as BNPL and credit lines to pay for everything from television sets and refrigerators to jewelry and clothing.
The reasons behind this rise are many - a major one being the gaps left by the traditional banking sector. According to a CIBIL report, around 480 million adult Indians till the age of 65, representing half of the overall population in the earning segment, are 'credit unserved'. This is the case for individual Indians - and when it comes to businesses, specifically the MSME sector, the situation isn’t any better. As per an IFC report, SMEs take up a miniscule 6-7% credit share and face a credit gap of close to $1.1 trillion. The gap can be attributed, in part, to reliance on past credit history.
In the words of the CIBIL report, it is a "chicken and egg" situation when it comes to unserved consumers.
“The current reality highlights the importance of incorporating enriched credit data into the lending ecosystem, so that fewer consumers find themselves as credit unserved,"
Rajesh Kumar, Managing Director and Chief Executive Officer of CIBIL
Here’s why - borrowers need credit history in order to access credit, which, clearly, leaves first-time borrowers between a rock and a hard place. Without this history, there’s no way for banks to underwrite these borrowers, and they’re left in the lurch.
So, are they doomed to remain credit unserved forever? Not quite.
What is alternate data?
Alternate data refers to any customer data that’s not directly related to their credit conduct. While traditional data sources such as credit bureau reports or credit scores provide information about a person’s activity and repayment behavior with their previous loans, alternate data provides a view of the person’s willingness and ability to repay a loan by looking at all other sources.
Alternate data refers to the myriad sources of information that can help a lender piece together a realistic picture of a person’s credit worthiness - with or without their formal credit history. Not only does this help underwrite new-to-credit customers but also helps improve efficiency of lending to even veteran borrowers with thick histories by improving the intelligence a lender has on them.
Examples of alternate data
Utility, rental, insurance, and other bill payments history
Behavioral and transactional data
Crowd-sourced fraud databases
Alternate data used in the loan lifecycle and underwriting process
Source: Aite Group for Experian
While there is no standardized definition of what makes a ‘good’ source of alternate data, these characteristics from Oliver Wyman’s research are widely quoted:
Coverage: A new data source will ideally have broad and consistent coverage
Specificity: a data source should ideally contain specific data elements that provide part of a full picture of the borrower (e.g. on- time and late payments over a significant period of time).
Accuracy and timeliness: Data should be accurate and frequently updated and regularly verified.
Predictive power (‘signal’): Data should contain information relevant to the behavior that you’re trying to predict
Orthogonality: The data source should be additive to traditional bureau data; this means that using it will improve the predictive accuracy of any new score by improving the signal-to-noise ratio.
Regulatory compliance: Data sources must comply with existing regulations for consumer credit
Benefits of alternate data underwriting
Opens up new borrower segments
As mentioned earlier, a large part of India’s population remains either credit underserved or unserved completely, due to their lack of bureau data. However, one can be fairly certain that they possess other kinds of data such as bill payments or social media usage - thus, lenders who leverage this alternate data open themselves up to this entirely new, large credit pool.
Helps improve Gini Coefficient
The Gini Coefficient is a metric that indicates a model’s (in this case, an underwriting model’s) discriminatory power, i.e. the effectiveness of the model in differentiating between ‘bad’ borrowers, who will default in the future, and ‘good’ borrowers, who won’t. alternate data enriches underwriting models and supplements bureau data, therefore improving its efficacy and upping its predictive power. Research has shown that the difference of even a single percentage point in the Gini Coefficient can help lenders save up to USD 10 million for every USD 1 billion in loans.
Facilitates tailored products and services
Alternate data sources enrich a lender’s underwriting process to a point where they have a deep assessment of a borrower’s risk appetite and likelihood of repayment. This allows lenders to push customized credit options to these borrowers - in terms of longer tenure, optimal interest rates, and credit limits.
Enables risk-based pricing
In a risk-based pricing model, lenders determine loan interest rates based on the borrower’s risk profile. In most cases, as compared to uniform pricing, borrowers end up paying less overtime. In addition, risk-based pricing incentivizes borrowers to improve their financial health to access lower priced financial products.
Alternate data gives borrowers access to more and better-priced financial products. According to the Organisation for Economic Co-operation and Development, “in the aggregate, lending is increased, leading to greater economic growth, rising productivity and greater stocks of capital. Average interest rates drop. Poverty and income inequality are alleviated.”
Alternate data helps with a lot more than just the initial borrower assessment. FinTechs, NBFCs and even some banks now use bots to automate their debt collection processes. These processes are powered by data from several sources including social media, insurance records, device data, and more. A collection and prioritization engine leverages this alternate data to:
Segment borrowers based on their riskiness and allocate collection resources intelligently
Identify risky borrowers before their loan due date and use targeted collection strategies
Get recommended date for eNACH presentation for bounced payments.
This data-backed strategy removes the need for lenders to hire and train manpower, minimizes occurrences of borrower harassment, saves on costs, and significantly boosts debt recovery rates.
Alternate data underwriting and its challenges
While alternate data in lending is rapidly gaining popularity, it is still a relatively new practice in underwriting. There still remain concerns around regulatory protections, privacy, and misuse. Considering the subjective nature of this data, there is also a possibility that information can be doctored to seem more favorable to lenders.
However, most of these concerns can be addressed if the source of data is reliable and if the lender / technology company has strict data protection protocols in place.
The global alternate data market size was valued at USD 2.7 billion in 2021 and is expected to expand at a compound annual growth rate (CAGR) of 54.4% from 2022 to 2030. The key drivers attributed to market expansion include the significant increase in the types of alternate information sources over the last decade.
FinBox DeviceConnect - our in-device SDK for risk underwriting - leverages 5000+ parameters across millions of data points and markers to generate a FinBox Inclusion Score that helps lenders differentiate between good and risky borrowers.
It has resulted in 35% less risk for our lender partners, improved financial inclusion by catering to the NTC segment, and significantly improved approval rates. To learn more about DeviceConnect and its alternate-data based credit risk scores, get in touch here.