The Micro, Small and Medium Enterprises (MSME) sector in India was on a streak – with close to 6 crore 34 lakh enterprises raking in over 28 percent of the nation’s GDP, the sector was one of the strongest drivers for economic growth and employment in the country. That was until COVID-19 put the brakes on the dream run.
Following the outbreak, it was reported that 60 percent of MSMEs claimed recovery owing to restricted working capital and lack of manpower, according to a survey conducted by the Reserve Bank of India (RBI). Furthermore, RBI identified MSMEs among the five most adversely affected sectors.
The availability of unsecured capital loans has always been a tricky business – on one hand, it is one of the key drivers to economic growth; but on the other, it severely impacts Non-Banking Financial Companies (NBFCs) on accord of bad debts.
Unlike most banks and NBFCs, Lendingkart’s business model takes a different approach to assessing risk – the company does not assess credit-worthiness based solely on an applicant’s credit history and financial records but instead relies more on machine learning models that analyses data from numerous data sources comprising of as many as 10,000 distinct variables.
CIO India talks to Manish Bhatia – President of Technology, Analytics and Capabilities at Lendingkart to get a read on how the company plays the balancing act by helping MSMEs with capital loans, at the same time, keeping bad debts to a minimum.
How has the COVID-19 outbreak impacted MSMEs and what has Lendingkart done to cater to their lending needs?
Upstream firms are mainly affected by labour shortage while downstream firms face more serious challenges related to supply constraints and consumer demand. We are learning that MSMEs are quite resilient to such shocks and have stepped up their efforts to adopt more digital mechanisms to keep their businesses going – making it more inclusive and sustainable.
It is critical to evaluate accurate creditworthiness of our customers during these testing times, so appropriate changes were required to be incorporated in our credit scoring and underwriting models.
Lendingkart extended all support to its customers in addressing their concerns and requests by prompt interventions like applying the moratorium feature and query sections on our customer-facing apps.
The collection strategy was also re-worked to arrest the delinquent behaviours of loans by a combination of penal waivers and payment collection in parts throughout the period so that it becomes affordable for the customers.
NBFCs, like traditional banks, have to worry about bad debts and Non-Performing Assets (NPAs). What is LendingKart’s approach to manage financial risk and what role does technology play in it?
Lendingkart has been catering to credit-deprived MSMEs who are not serviced by traditional institutions, and we have built credit underwriting models and technology capabilities with a vision to scale up and make finance available to remote geographies.
To achieve the objective of addressing the credit gap, we evaluate various factors apart from credit bureau scores and financial documentation, as relying on the latter affects their credit evaluation.
We focused on evaluating and monitoring early indicators from market-wide stress scenarios of individuals and combined variables. Accordingly, we built additional checks and due diligence on the identified parameters. This approach also enabled us to offer better options to our existing customer base, for whom credit lines were maturing.
Managing financial risk is a top prerogative for CTOs in the BFSI as well as NBFC sectors. What is the technological approach CTOs must take to evaluate risk better?
Technology will be able to manage financial risk if it is built around customer behaviour. Being agile and adaptable to new challenges is key to providing a suitable response to the challenges and opportunities that are being faced currently.
Risks can be managed once you understand the customer better and then have the technological and analytical capabilities to support mitigation of associated risks.
We, at Lendingkart, have been continuously trying to learn and improve the model at a faster rate based on easy and reliable access to information. This involves solving the problem by developing alternative mechanisms to understand MSME requirements based on bank data, investment data, purchase patterns on e-commerce websites, social media, as well as utility data in absence of credit records and financial statements.
Coupling all the available digital data with machine learning makes it easier to assess risk.
Manish, could you share with our readers how the company has been able to keep its predictability models accurate – especially after the disruption that followed the nationwide lockdown?
Lendingkart has always been proactive in evaluating the early warning triggers and formulating mitigation strategies to address the probable risks.
Three types of statistical tests are run on the model to evaluate the model predictability and dependability on a sample from time to time. At the start of the lockdown period, these scenarios were simulated for testing and the projections were made based on results of the test.
The risk intensity and continuity plan have been designed accordingly for the lockdown period and post lockdown recovery period. This approach coupled with our internal efforts has gone a long way in arresting any unpredictable or surprising outcomes. Additionally, our internal teams were trained and made aware of upcoming challenges.
RBI signalled the emergence of open banking by proposing the adoption of Account Aggregators. What does this mean for technology leaders in the NBFC space?
Account Aggregators addresses the need to have an open data sharing infrastructure in place where the data could be extracted from the financial information provided directly through explicit customer consent, via a safe and secure platform. This can not only ensure instant access to data required for underwriting but also eliminate data submission frauds.
The Account Aggregator ecosystem will ensure flow of data from banks, insurance companies, mutual funds, stock broking, pension funds, GSTN and income tax authorities to financial information users instantly. This will be extended to other industries such as healthcare, telecom and other use cases.
After a consent-based mechanism opening up for personal data (Aadhaar, DigiLocker) and exchange of value (UPI), unlocking financial data will be the final step to achieving complete open banking in India.
To wrap up, what will be your top five risk management takeaways for CIOs?
- Risk is a reality for leaders and managers, irrespective of the industry sector or size they operate in. Risk can be healthy if it is calculated and a continuity plan is in place in case the calculations fail to work.
- In my experience, a robust mechanism needs to be in place for risk management which includes identifying the risk areas, their assessment and measurement, having a mitigation or continuity plan in place, and regular monitoring to have any deviations being highlighted at early stages.
- To have the best plan in place for risk areas, it is essential to deep-dive and get into specifics and formulate the risk control framework.
- Segregation of risk areas in controllable, non-controllable, and high/low impact helps in aligning the priority for objectives. Controllable and high impact ones are the first ones to be addressed. The identified scenarios can be covered and the team can sign on to structure actionable tasks accordingly. Leadership plays a big role in emphasizing the importance of the risk management framework to their teams and creating an awareness of the same.
- A balanced approach has to be taken between taking risks and reducing them. You have to be in it to win it. And finally, by-stander analyses should be taken with a pinch of salt!
Copyright © 2020 IDG Communications, Inc.