Sustainable Growth of MSME
India has a large 7.9 million registered Micro, Small, and Medium Enterprises (MSMEs) base. The MSME Industry segment contributes 33% of the Country’s GDP and generates over 120 million jobs across industries and regions in the country contributing towards wealth creation at the grass-root level.
A large proportion of the MSMEs that are playing a major catalytic role in driving sustainable growth in the country are thin-file MSMEs. Thin-file MSMEs are small businesses that have little to no credit history. These businesses are either starting out or existing but have never taken a loan and have thin files.
Here are some of the reasons why lending to this underserved segment of the market now presents a good opportunity for lenders:
- MSMEs contribute 45% of the total exports in the country and the world is looking towards India to provide an alternative manufacturing base after the uncertainty in the Chinese market
- The government recognizes the importance of MSMEs in order to achieve the ambitious 5 trillion economy by 2025. It has introduced several guarantee mechanisms to de-risk lending to the MSMEs. Lenders now have an added incentive to lend to these businesses
- The Indian MSMEs are rapidly adopting digital payments over cash, with 72% of payments done through the digital mode compared with 28% of cash transactions. The rise in digital payments allows for alternative ways to assess the creditworthiness of these businesses
- Due to the advent of digital and the improvement in understanding of digital tools, digital lending has been able to significantly bring down lending costs and also provide larger distribution channels
Financial institutions have a great opportunity to invest in these thin-file MSMEs to diversify their lending portfolios and cater to the need of this sector for rapid growth.
Why aren’t Financial Institutions jumping on this lucrative MSME opportunity?
A pool of opportunities exists with thin-file MSMEs as the sector is growing rapidly. However, assessing and evaluating credit limits has been a challenge for financial services, digital lending groups, NBFCs, and the BFSI sector.
Many thin-file MSMEs don’t have enough data such as audited balance sheets, P&L statements, projected turnover, and acceptable credit scores, making underwriting difficult and resulting in costly credit. Therefore, most financial institutions and banks are hesitant to provide credit to small businesses as they feel firms are non-transparent regarding their financial conditions, resulting in FIs and banks asking huge collaterals for credit repayment, which in turn leaves a huge credit gap between MSMEs and financial institutions.
Recently, in the wake of rapid digital adoption across sectors, businesses have started leaving a trail of data that is invaluable. In traditional assessment, lenders deleverage themselves through asset collateralization. In a sector that is asset-light, informational collateral or the alternate data trail of the small business can serve as a very good alternative to assess businesses.
Alternate Data: A New Approach
Alternative data refers to information about a business that is readily available in digitized form but is ‘alternative’ to conventional methods such as documented credit history. This data can be used as a surrogate in order to analyze the creditworthiness of a business. This data is largely classified as structured and unstructured data. Some examples of structured data include utility bill payments, mobile phone bill receipts, rental information, GST filing trends, Employment patterns, News sentiment analysis, etc. Unstructured data could include social media behaviors, emails, text and messaging files, etc.
Considering the plethora of such alternate data sources that could be used to measure the ability to repay or the willingness to repay, we are presented with the challenge to be able to make sense of such huge amounts of information. Seemingly impossible to do manually and in near real-time, machine learning and artificial intelligence come to our rescue.
Alternate data is a modern and efficient approach to risk-based credit approval strategy for underwriting. With alternate data, financial institutions can have a much more holistic view of business creditworthiness based on business cash flow, and financial structure including balance sheet, P&L, and consent-based data such as GST filing, ITR filings, and Employment trends of the company.
The sparse availability of data for thin-file businesses makes the traditional way of risk analysis an incomplete and tedious task. Augmenting available data with alternative data to then mine insights is proving to be a game-changer for risk analysis of thin-file customers.
However, the data provided by small businesses could be more structured and should be converted into meaningful insights that can be easily interpreted for credit evaluation. The traditional way of arranging data sets manually and assessing business health is time-consuming and might lead to inaccuracy.
Here the AI/ML algorithm comes into play. The ML algorithms are continually self-improving feedback-driven systems that can be trained to interpret unstructured data to derive meaningful insights, furthermore delivering actionable intelligence and early alerts. These AI/ML-driven risk models can provide an increased level of accuracy and predictability as compared to the traditional ways of credit risk analysis.
How Alternate Data Can Help Small Businesses
Small businesses pose a risk for lenders due to a lack of formal credit information. To offset these inherent risks, lenders charge a higher interest rate, often higher than what informal capital providers charge. As a result, small business owners are trapped in an unending cycle of borrowing from informal sources, therefore never gaining access to formal capital.
Alternate data can play a key part in breaking this vicious circle. Business owners can leverage AI/ML technology to pass on insights about their business to lenders if they open up their alternate data to data companies. This allows for a faster understanding of the borrower’s ability to repay, which is now possible for lenders.
The type of insights that can be generated by AI/ML can be considered akin to business risk scores. These scores can help inform lenders about the borrower’s creditworthiness. Let’s look deeper into it in the next section.
Introducing Crediwatch Trust Score
Crediwatch Trust Score is an alternative score to provide more accurate risk pricing and metrics to thin file MSMEs leveraging existing public data sources as well as upcoming infrastructures such as Account Aggregators and other alternative data/ open finance platforms. The Trust Score has an 8-point rating scale as shown below. 
This AI-powered model enables straight-through processing of loans for digital lending
use-cases for banks, NBFCs, and other fintech as well as for BNPL use-cases of B2B marketplaces.
The same can be used as a primary input for underwriting unrated businesses for existing credit
products of the banks, supplier/ distributor risk management by large corporations, et
Crediwatch understands the data sources available in the ecosystem and how the same can be leveraged to derive insights on the credit-worthiness of a business to meet the needs of existing credit policies at banks, NBFCs, and Fintechs as well as how AI/ ML can unlock value for lenders by formulating new credit products suitable for each business segment.
The utilization of Trust Score is a good way to
- Improve the quality and accuracy of digital lending
- Empower lenders to make better financial decisions
- Optimize workflows between a lender and borrower
- A good alternative for credit risk scoring
