Till about half a decade ago, loans were given out purely on the basis of a credit-rating system which would include an individual or business’ income/revenue, current debt levels, and how the previous loans were or are being repaid. In such a scenario, a good score would translate to a better chance at securing a loan in the near future, and at viable interest rates. However, this model ended up excluding a significant share of individuals and businesses across the country, rendering them underserved.
This is because not everyone could fit into all the parameters of this model, and many did not even have a credit score yet. For instance, while a business may have excellent records, the fact that it is only 12-15 months old would exclude it from the minimum conventional bracket of 3 years. However, the information provided by credit bureaus did not put together such a comprehensive picture. Further, over-reliance on hard collateral and high cost-to-serve limited access to larger businesses in metros and Tier 1 cities.
How could underwriting be done in a manner that would help address these pressing need-gaps by generating a more accurate credit profile? Considering how the traditional credit scores were aimed at those who were already a part of the system, how could the underserved overcome this disparity? This is where new-age Fintech companies stepped in. By leveraging emerging technologies like Big Data, and predictive analytics, these companies are steadily succeeding at tapping into the vast pools of data sets that have not been considered previously and mitigating risks to pave way for a clearer picture and resultantly, more instantaneous, accurate, and effective underwriting.
A sea of alternate data sets
The emergence of alternate data has proved to be a boon for the underserved. The vast amounts of data available and accessible to lenders today equip them with a comprehensive view of the prospective borrower, enabling them to derive new and meaningful insights. The information being taken into account has expanded from a thin stream of credit bureau data to a sea of alternate data sets that may include daily transactions, and utility and recurring payments, along with entity data, and even social media presence.
With the MSME segment rapidly formalizing and digitizing, the digital footprint established by these small businesses is being gathered by new-age lenders, who leverage an ecosystem-based approach for the same. The basic idea is to partner with different transactional platforms across various levels of the supply chain of the segments that they cater to, thereby acquiring the above-mentioned alternate data.
For instance, a small retailer registered on an e-commerce platform leaves a digital trail behind by transacting through that platform. Amidst this, a digital lender that has partnered with that platform can gather the data of those retailers to map their granular sales pattern, create more accurate borrower profiles, and further assess them for their true creditworthiness by scouring the digital data, together with the traditional scores.
Moreover, this also helps bring the underserved back in purview. In an economy like India where one of its biggest growth engines—the MSME segment—has traditionally been starved of formal credit sources, the rise of alternate data is helping the smaller businesses meet their credit demands, have some financial stability, improve their lives, grow and expand their business, and generate greater outputs as well as employment opportunities. Therefore, the increasing use of these technologies in underwriting is helping drive true inclusive growth in India.
New-age technologies have immense potential to streamline underwriting by enabling the system to handle those aspects of processes that can be implemented without any human intervention. By running the vast data sets against the set of rules that are designed to determine acceptability, this system then presents lenders with a decision along with the anomalies or discrepancies. Further, by providing such a granular level of visibility of real-time data, analytics can offer underwriters information on the more complex portions of the process.
There are specific credit risks pertaining to certain business segments, which are presented with perfect clarity through data analytics. For instance, a business belonging to the travel segment risks volatility, whereas a restaurant would have the risk of succumbing to the intense competition and going out of business. Having a predictive industry-specific approach enables digital lenders to underwrite according to the inherent risks involved in each of the segments, thus creating a comprehensive borrower profile for assessment.
Additionally, analytics drive better efficiency by taking the redundancy of certain administrative tasks such as gathering borrower reports or asset statements away, thus enabling underwriters to focus on those subjective parts of the process that require human involvement, like reaching out to the borrower to understand other contingencies that the data does not reveal. Therefore, while the decision ultimately lies with the lender, these technologies facilitate better and more insightful decision-making, as effectively and quickly as possible. This enables a seamless, instantaneous, and convenient experience not just for the lenders, but for the borrowers as well.
More accurate credit-risk models
Credit-risk analysis is essentially an assessment of the possibility of failure on the borrower’s part when it comes to repayment and the loss that it may cause to the lender. Essentially, it involves prediction, and such a task invariably has the probability of mistakes creeping in. However, new-age technologies can help mitigate the probable loss. Let us see how.
The robust adoption of ML-based tools helps in the identification of complex and non-linear patterns that exist within larger data sets, thus resulting in the determination of the most comprehensive risk-return analysis. Based on such analysis, the behavioral patterns are then mapped out to make accurate predictions. Such a model also serves as an early-warning system in those cases where current behaviour points to a risk of default, alerting both the borrower and the lender before a default actually takes place. Credit risk management, analysis, and consequent mitigation are essentially the pillars of underwriting, and emerging digital technologies like AI, ML, and data analytics help optimize each of these.
While risks will always be associated with the business of lending, new-age underwriters are increasingly striving to mitigate the same. Technology is steadily filling the gaps and addressing the loopholes that have existed in the traditional credit models, be it in terms of data, process automation, or prediction of the defaults. Adoption of advanced technologies in underwriting is not intended at reducing or eliminating elements, but to drive productivity and facilitate smarter, more insightful decision-making for the lender. As they say ‘underwriting is a fine balance of art and science’—and leveraging technological tools like AI, ML, and data analytics in the underwriting process can enhance its efficiency while helping maintain the balance.
By Alok Mittal, CEO and Co-founder, Indifi Technologies