US
IN
Co-founder of ZestMoney, Priya Sharma takes us through the good, the bad, and the new of lending in a post COVID-19 world with some sharp insights and creative Nike symbol analogies.
Introduction
Tectonic landscape shifts and developments have become a feature of the Indian fintech ecosystem. Newly baptised by the coronavirus fire, the Indian consumer lending community is emerging into a newer and more nuanced future. Who better to explore this than with our first HyperInsights guest, Priya Sharma of ZestMoney! In this hour-long conversation with our CEO Kedar and moderator Deepak, we connect the dots on consumer lending.
Early in the hour, we asked participants to rank the topics they wanted to hear about and accordingly they discussed risk assessment, loan origination, and the increased use of AI.
Risk assessment (link)
The Context
The chicken and egg problem of assessing risk without a credit score and modeling a credit score without lending is compounded during these times when a good credit score may not imply a steady income. Given that the core of what drives ZestMoney is a proprietary credit scoring engine for new to credit underwriting, Priya had some fascinating insights to share:
HyperInsights
- This will be the first major shock for Indian customers. In the West, consumer data around the 2008 crash exists. So it’s unpredictable for now.
- Understanding affordability in addition to ability is the key. One can view income statements, loan repayment, and spending patterns together.
- Behavioral signals about expenses, spending, and investment preferences will help classify and service customers better.
- Retraining and scaling Artificial Intelligence and ML is critical to gain insights on affordability and behaviour without using traditional credit data.
Loan origination (link)
The Context
Irrespective of whether you are a well funded Neo bank, a large scale or regional NBFC, or even a bank, the validity of earlier originated consumers is under question. Also, since the lockdowns mean point of sales will remain closed, lenders will have to re-imagine origination.
HyperInsights
- Alternative modeling with AI will help navigate the skewing of data caused by the RBI moratorium on EMI payments where even opportunistic customers avail the moratorium (“moral hazard”) thus making it harder for worthy customers.
- There is expected to be a Nike tick like recovery in credit demand. Maybe we’ll be back on track by December.
- The effects of the moratorium will not be reflected in traditional credit bureau data at the same time it cannot be ignored entirely.
- Loan products may need to pivot to the current context. There will be a surge in demand for home office equipment, online education, and planned surgeries (that were delayed due to the lockdown).
Increased use of Artificial intelligence (link)
The Context
AI has been a part of the ZestMoney DNA since its inception. Priya tells us why AI is an enabler for the entire lifecycle of lending.
HyperInsights
- Firstly, companies have to go fully digital to use the power of AI.*
- Consumers are more comfortable than ever with video interfaces, video KYC solutions built on AI will be a key driver for onboarding going forward.
- Real-time interactions will be all the more critical in an increasingly digital user experience.
- Onboarding, fraud prediction, and collections will be the most exciting parts of the cycle to benefit from AI.
*Coincidentally, Priya had started her career with one of UK’s largest digital transformation companies: Sapient.
Questions and recurrent themes
The end of the discussion was reserved to answer a volley of questions ranging from why lending is not just a feature to personal tips to keep going during these times.
Questions
- On the supply of credit during these times (link)
- On the role of AI in collections (link)
- On the evolution of EMI (link)
- On lending being a feature, or is it? (link)
- On remaining calm and maintaining composure during these times (link)
Recurrent themes
- There is an increased risk in lending now, and that is causing conservative behavior and not a lack of capital.
- Lending is a full-stack function and not a feature, and in times like these, this is more clear than ever. (Hat tip to Kunal Shah.)
- We can be innovative about trickier aspects of lending such as collections and reconciliation with AI.
Priya ended the hour with some really wise words, which is relevant for all of us. Don’t miss it!
We followed up our excellent introduction to HyperInsights with an interview with Akshay Mehrotra of EarlySalary, which can be viewed here!
