CREDIT SCORE BEYOND BUREAU
Alternate Data: An Overview
Part 2: A deep dive into Alternate Data and its Impact on Credit Inclusion
What is Alternate Data?
In the context of credit scoring, Alternate Data refers to the new data sources used by lenders such as Banks and NBFCs to evaluate an individual’s credit risk that is not captured within their traditional data sources such as credit bureaus, historical financial records, etc.
Why use Alternate Data?
Within the Indian context, as half of India’s adult population is credit unserved (part 1), there is an urgent need to bring more population segments within the fold of the formal credit economy.
This requires businesses to be able to assess the Creditworthiness of credit unserved customers before extending financial products to these unserved segments.
This is not easy; evaluating Creditworthiness without historical bureau data or financial data is challenging. Alternate data helps fill in this data gap and allows for robust credit scoring even in the absence of historical bureau data or financial data.
In a nutshell, Alternate Data informs credit decisions in the absence of formal bureau data.
Modalities of Alternate Data
Alternate Data is multimodal, and hence, seeks to augment the absence of bureau data robustly.
Multimodal Alternate Data
Alternate Data, by definition, exists across modalities such as:
- Geo-location: Decoding any unusual pattern and anomalies using Real-time Location data
- Geo-demography: The analysis of people by where they live based on the principle, i.e., two people living in the same neighbourhood are more likely to be similar
- App usage: Decoding any suspected or fraudulent pattern through app usage pattern
- Email: Now with the digital revolution, email has become one of the important tool for communication and at the same time an enabler for cyber crime. Therefore, validation of email becomes very critical
- Ownership: Vehicle and Property ownership status can be an indicator around ability to re-pay, especially when individual does not have any credit history
- Utility Bills: Expenditure on household consumption, such as gas, electricity and their bill payment pattern proved to be a good indicator on ability and intention of re-payment
- Telecom Data: Similarly, telecom has become a basic neccesity whose usage pattern provides a level of affluence, which proved to be a key indicator for alternate credit scoring
Using Alternate Data at CueNex
At CueNex, we leverage AI and multimodal Alternate Data to create Alternate Credit scores for the unserved and NTC populations.
This allows us to cater to Enterprise use cases such as Custom Risk Scoring, Fraud Identification, and Recovery Analytics.
Our ML Models leverage both Alternate Data as well client-provided internal datasets to enable robust risk analysis, especially for NTC and credit-unserved customers.
Overall, Alternate Data has greatly impacted our customers, often providing them a competitive edge in terms of newer growth areas in NTC/credit-unserved segments, and has enabled almost risk-free growth in these areas.
In the next blog, we will look at the various financial inclusion centric use cases and application of AI to solve those.
