Digital lending, simply defined, is sourcing, assessing and servicing loans through digital means and data points. Digital lending, in theIndian context, aims at using digital surrogates to circumvent lack of proper documentation of income statement, balance sheet, etc.
Digital lending in India has begun to show signs of prominence and has caught the attention of traditional lenders and startups alike. MSME lending in India is the most complex of these segments and is highly complicated due to underreported income and poor accounting practices; lacklustre performance and distressed status of theIndian manufacturing and commodity sector; and the hyper growth experienced by the e-commerce segment.
Image : shutterstock
Here are five trends that could shape the future of MSME digital lending in India over the next three to five years
Digital data as a surrogates for financial data
Digital lenders are increasingly farming data points such as bank transactions, bureau performance, transactions available on accounting/CRM system, transactions with suppliers and e-commerce platforms. These data points, in various combinations, can be used to assess the top line and size of business; but using them to arrive at profitability is an entirely different matter.
While underwriting, segmenting customers based on their profitability rather than the perceived value is important. For this, relevant customer profitability parameters such as current product usage levels, industry margins, potential future earnings, and risk parameters (such as probability of default) should be considered. Overlapping customers’ digital data with their financial data and traditional ratios will prove to be critical. In short, lenders’ ability to overlay profitability matrices over the transaction data will define the winners.
Digital lending is often used in context of young entrepreneurs and businesses with low exposure to banking products. Herein lies the biggest risk, as most businesses fail in the early stages, which is why traditional lenders either stay away from them or demand high collaterals. The ignorance meted out to these businesses makes this an untapped market. Such businesses often engage with large suppliers and have complex, undefined payment cycles. Using the ecosystem, in terms of supplier data, securing repayment through suppliers and routing repayments, is the key. Lending to suppliers of a large corporate, relying on their purchase data, and routing the payments through the corporate reduces the risk, yet maintains the core of digital lending.
Operating model reorientation to position for digital transformation
Most large banks, in their attempts to go digital, often mistake digital lending for just changing credit policies and underwriting methods. Success in digital lending is about transforming the customer experience as well, and requires more agile operating models. According to a study#1, customers submit only 14 percent of loan applications through digital channels and most traditional financial institutions lack digital cross selling expertise. Some of the critical changes that will define success for lenders are:
- Take a mobile-first design approach (keeping in mind the ascent of mobile banking), which imposes a useful reductionary discipline to the User Experience (UX) design
- Executedigital strategies consistently across channels
- Develop modular products that can be assembled like LEGO bricks, enabling faster time to market, a high degree of personalisation for customers and reusability through common processes and systems
- Generate real-time offers for straight-through processing using algorithms, as against using them as advisory tools
Big data and analytics
Analytics being at the core of decision making and marketing is a big cultural shock for Indian lenders. The task ahead for data scientists is to prove their worth by constantly restructuring the models, tracking defaults on a real-time basis and helping lenders invest in early warning systems. It will be critical for decision makers to let go of the ‘human expertise is the key to credit’ approach and constantly let analytics take a ‘test and grow’ approach.
This means that there will be no more thorough underwriting in the form of human intervention, and hence potentially no more thorough understanding of the risks embedded in the transactions (because not all risk factors of a transaction can be translated into a risk model or be collected through big data). However, a balance needs to be struck here with the spread of risk modelling and big data analytics; any lender aiming at meeting its regulatory requirements (as well as risk policy covenants) should be able to do so by merely creating a committee that ‘signs off’ a high risk transaction, ensuring that all the checkboxes were ticked, as delivered by the risk software.
Fintech companies around the world are creating newer models to make lending decisions and service loans. In the US, fintech companies such as ZestFinance have moved beyond traditional risk assessment to use new sources of data in underwriting, such as whether an applicant keeps a consistent phone number or has made delayed bill payments.
Agile ‘think- on-feet’ risk management
Organisations with proactive and defensive risk management practices are likely to see through the flux when the dust settles on digital lending. The classic mistake of limiting risk management only to credit risk should be avoided and an overall approach which includes assessing macro-economic, regulatory and compliance changes, the advent of e-commerce business and increased competition, should be adopted.